Although the adverse effects of systemic glucocorticoids (GCs) are well known, their association with hypertension in rheumatoid arthritis (RA) has been unclear. Now, a large population-based study shows that the drugs are linked to a 17% overall increased risk for incident hypertension among patients with RA.

Further, when the researchers stratified participants by dose category, they found that doses higher than 7.5 mg were significantly associated with hypertension. Cumulative dosage was not tied to any clear pattern of risk.

The authors, led by Ruth E. Costello, a researcher at the Centre for Epidemiology Versus Arthritis in the Centre for Musculoskeletal Research at the University of Manchester (England) concluded that patients who are taking these drugs for the treatment of RA should be monitored for high blood pressure, which is an important but modifiable cardiovascular risk factor, and treated appropriately.

The results of Ms. Costello and colleagues’ study were published June 27 in Rheumatology.

“While fractures associated with these steroid drugs are well studied, hypertension is a side effect that seems to have been less well studied, and yet it is an important cardiovascular risk factor that can be managed,” Ms. Costello said in an interview.

To better understand the possible association, Ms. Costello and colleagues identified 17,760 patients who were newly diagnosed with RA between 1992 and 2019 and were included in the Clinical Practice Research Datalink, which represents about 7% of the U.K. population. None of the patients had hypertension at initial RA diagnosis. Slightly more than two-thirds were women (68.1%), and the mean age was 56.3 years.

Of those patients, 7,421 (41.8%) were prescribed GCs during postdiagnosis follow-up. Most patients (73%) were followed for at least 2 years.

Patients who used GCs were slightly older than never-users (mean age, 57.7 vs. 55.3 years), were predominantly women, had a history of smoking, and had more comorbidities.

The overall incidence rate (IR) of hypertension was 64.1 per 1,000 person-years (95% confidence interval, 62.5-65.7). There were 6,243 cases of incident hypertension over 97,547 person-years of follow-up.

Among those exposed to GCs, 1,321 patients developed hypertension, for an IR of 87.6 per 1,000 person-years. Among unexposed participants, the IR for hypertension was 59.7 per 1,000 person-years. In Cox proportional hazards modeling, GC use was associated with a 17% increased risk for hypertension (hazard ratio, 1.17; 95% CI, 1.10-1.24).

The researchers noted that 40% of GC users with hypertension were not prescribed an antihypertensive agent at any point during the study. “Whilst some may have been offered lifestyle advice, left untreated this has important implications in terms of addressing modifiable risk factors in an RA population already at increased risk of CV disease,” they wrote.

They noted that cardiovascular disease is a major driver of the elevated mortality risk seen among adults with RA compared with the general population and that recent treatment recommendations address management of cardiovascular risks in these patients.

“There are several routes by which GCs may promote cardiovascular disease, including hypertension, metabolic changes, diabetes, and weight gain. We don’t currently know the extent to which each of these individual mechanisms may be increasing cardiovascular disease,” said Ms. Costello.

“Glucocorticoids increase fluid retention and promote obesity and hypertension,” said Rajat S. Bhatt, MD, a rheumatologist at Prime Rheumatology and Memorial Hermann Katy Hospital in Richmond, Texas, who sees hypertension in GC users in his clinical practice. “So patients need to be monitored for these risk factors,” he said in an interview.

Although hypertension may be a significant factor in the increase in cardiovascular disease in the RA population, Dr. Bhatt said the major driver is likely the intrinsic inflammatory state caused by the disease itself. As to why the GC-hypertension connection has flown under the radar in RA, he added, “That specific link has been difficult to tease out since RA patients are often on multiple medications.”

In regard to the role of dosage, Dr. Bhatt said that hypertension risk increases with higher GC doses, as the U.K. study indicates, and usually subsides when patients stop using GCs.

“Whether the observed dose association is causal or influenced by the underlying disease severity, our results suggest we should be vigilant in patients on all doses of GC, especially higher doses,” Ms. Costello added.

In regard to using drugs that are less cardiotoxic than GCs, Dr. Bhatt said that there are clinical scenarios in which GC therapy is the best choice, so just switching to nonsteroidal drugs is no panacea. “All RA drugs have adverse side effects, and anyway, the goal of rheumatology treatment is always to get patients off corticosteroids as soon as possible,” he said.

Ms. Costello and colleagues noted that their results are consonant with earlier research, including a single-center, cross-sectional study in which less than 6 months’ use of prednisolone at a median dose of 7.5 mg was associated with hypertension. In a German registry study, among patients who received doses of less than 7.5 mg for less than 6 months, there were higher rates of self-reported elevations in blood pressure.

The findings are at odds, however, with a recent matched-cohort study, which also used data from the Clinical Practice Research Datalink. That study found no association between GC use and hypertension.

GCs have come under increasing scrutiny in regard to several diseases. A study published July 7 found that even short-term courses of a few days’ duration entail risks for serious adverse events.

Ms. Costello’s group says that an estimate of GC-related incident hypertension in RA should allow more informed treatment decisions and that their findings highlight the ongoing need to monitor for and address this risk.

The study was supported by the Centre for Epidemiology Versus Arthritis and by the National Institute for Health Research Manchester Biomedical Research Centre. Coauthor William G. Dixon, PhD, has received consultancy fees from Google and Bayer unrelated to this study. Dr. Bhatt has disclosed no relevant financial relationships.

SOURCE: Costello RE et al. Rheumatology. 2020 June 27. doi: 10.1093/rheumatology/keaa209.

A version of this article originally appeared on Medscape.com.

Although the adverse effects of systemic glucocorticoids (GCs) are well known, their association with hypertension in rheumatoid arthritis (RA) has been unclear. Now, a large population-based study shows that the drugs are linked to a 17% overall increased risk for incident hypertension among patients with RA.

Further, when the researchers stratified participants by dose category, they found that doses higher than 7.5 mg were significantly associated with hypertension. Cumulative dosage was not tied to any clear pattern of risk.

The authors, led by Ruth E. Costello, a researcher at the Centre for Epidemiology Versus Arthritis in the Centre for Musculoskeletal Research at the University of Manchester (England) concluded that patients who are taking these drugs for the treatment of RA should be monitored for high blood pressure, which is an important but modifiable cardiovascular risk factor, and treated appropriately.

The results of Ms. Costello and colleagues’ study were published June 27 in Rheumatology.

“While fractures associated with these steroid drugs are well studied, hypertension is a side effect that seems to have been less well studied, and yet it is an important cardiovascular risk factor that can be managed,” Ms. Costello said in an interview.

To better understand the possible association, Ms. Costello and colleagues identified 17,760 patients who were newly diagnosed with RA between 1992 and 2019 and were included in the Clinical Practice Research Datalink, which represents about 7% of the U.K. population. None of the patients had hypertension at initial RA diagnosis. Slightly more than two-thirds were women (68.1%), and the mean age was 56.3 years.

Of those patients, 7,421 (41.8%) were prescribed GCs during postdiagnosis follow-up. Most patients (73%) were followed for at least 2 years.

Patients who used GCs were slightly older than never-users (mean age, 57.7 vs. 55.3 years), were predominantly women, had a history of smoking, and had more comorbidities.

The overall incidence rate (IR) of hypertension was 64.1 per 1,000 person-years (95% confidence interval, 62.5-65.7). There were 6,243 cases of incident hypertension over 97,547 person-years of follow-up.

Among those exposed to GCs, 1,321 patients developed hypertension, for an IR of 87.6 per 1,000 person-years. Among unexposed participants, the IR for hypertension was 59.7 per 1,000 person-years. In Cox proportional hazards modeling, GC use was associated with a 17% increased risk for hypertension (hazard ratio, 1.17; 95% CI, 1.10-1.24).

The researchers noted that 40% of GC users with hypertension were not prescribed an antihypertensive agent at any point during the study. “Whilst some may have been offered lifestyle advice, left untreated this has important implications in terms of addressing modifiable risk factors in an RA population already at increased risk of CV disease,” they wrote.

They noted that cardiovascular disease is a major driver of the elevated mortality risk seen among adults with RA compared with the general population and that recent treatment recommendations address management of cardiovascular risks in these patients.

“There are several routes by which GCs may promote cardiovascular disease, including hypertension, metabolic changes, diabetes, and weight gain. We don’t currently know the extent to which each of these individual mechanisms may be increasing cardiovascular disease,” said Ms. Costello.

“Glucocorticoids increase fluid retention and promote obesity and hypertension,” said Rajat S. Bhatt, MD, a rheumatologist at Prime Rheumatology and Memorial Hermann Katy Hospital in Richmond, Texas, who sees hypertension in GC users in his clinical practice. “So patients need to be monitored for these risk factors,” he said in an interview.

Although hypertension may be a significant factor in the increase in cardiovascular disease in the RA population, Dr. Bhatt said the major driver is likely the intrinsic inflammatory state caused by the disease itself. As to why the GC-hypertension connection has flown under the radar in RA, he added, “That specific link has been difficult to tease out since RA patients are often on multiple medications.”

In regard to the role of dosage, Dr. Bhatt said that hypertension risk increases with higher GC doses, as the U.K. study indicates, and usually subsides when patients stop using GCs.

“Whether the observed dose association is causal or influenced by the underlying disease severity, our results suggest we should be vigilant in patients on all doses of GC, especially higher doses,” Ms. Costello added.

In regard to using drugs that are less cardiotoxic than GCs, Dr. Bhatt said that there are clinical scenarios in which GC therapy is the best choice, so just switching to nonsteroidal drugs is no panacea. “All RA drugs have adverse side effects, and anyway, the goal of rheumatology treatment is always to get patients off corticosteroids as soon as possible,” he said.

Ms. Costello and colleagues noted that their results are consonant with earlier research, including a single-center, cross-sectional study in which less than 6 months’ use of prednisolone at a median dose of 7.5 mg was associated with hypertension. In a German registry study, among patients who received doses of less than 7.5 mg for less than 6 months, there were higher rates of self-reported elevations in blood pressure.

The findings are at odds, however, with a recent matched-cohort study, which also used data from the Clinical Practice Research Datalink. That study found no association between GC use and hypertension.

GCs have come under increasing scrutiny in regard to several diseases. A study published July 7 found that even short-term courses of a few days’ duration entail risks for serious adverse events.

Ms. Costello’s group says that an estimate of GC-related incident hypertension in RA should allow more informed treatment decisions and that their findings highlight the ongoing need to monitor for and address this risk.

The study was supported by the Centre for Epidemiology Versus Arthritis and by the National Institute for Health Research Manchester Biomedical Research Centre. Coauthor William G. Dixon, PhD, has received consultancy fees from Google and Bayer unrelated to this study. Dr. Bhatt has disclosed no relevant financial relationships.

SOURCE: Costello RE et al. Rheumatology. 2020 June 27. doi: 10.1093/rheumatology/keaa209.

A version of this article originally appeared on Medscape.com.

Although the adverse effects of systemic glucocorticoids (GCs) are well known, their association with hypertension in rheumatoid arthritis (RA) has been unclear. Now, a large population-based study shows that the drugs are linked to a 17% overall increased risk for incident hypertension among patients with RA.

Further, when the researchers stratified participants by dose category, they found that doses higher than 7.5 mg were significantly associated with hypertension. Cumulative dosage was not tied to any clear pattern of risk.

The authors, led by Ruth E. Costello, a researcher at the Centre for Epidemiology Versus Arthritis in the Centre for Musculoskeletal Research at the University of Manchester (England) concluded that patients who are taking these drugs for the treatment of RA should be monitored for high blood pressure, which is an important but modifiable cardiovascular risk factor, and treated appropriately.

The results of Ms. Costello and colleagues’ study were published June 27 in Rheumatology.

“While fractures associated with these steroid drugs are well studied, hypertension is a side effect that seems to have been less well studied, and yet it is an important cardiovascular risk factor that can be managed,” Ms. Costello said in an interview.

To better understand the possible association, Ms. Costello and colleagues identified 17,760 patients who were newly diagnosed with RA between 1992 and 2019 and were included in the Clinical Practice Research Datalink, which represents about 7% of the U.K. population. None of the patients had hypertension at initial RA diagnosis. Slightly more than two-thirds were women (68.1%), and the mean age was 56.3 years.

Of those patients, 7,421 (41.8%) were prescribed GCs during postdiagnosis follow-up. Most patients (73%) were followed for at least 2 years.

Patients who used GCs were slightly older than never-users (mean age, 57.7 vs. 55.3 years), were predominantly women, had a history of smoking, and had more comorbidities.

The overall incidence rate (IR) of hypertension was 64.1 per 1,000 person-years (95% confidence interval, 62.5-65.7). There were 6,243 cases of incident hypertension over 97,547 person-years of follow-up.

Among those exposed to GCs, 1,321 patients developed hypertension, for an IR of 87.6 per 1,000 person-years. Among unexposed participants, the IR for hypertension was 59.7 per 1,000 person-years. In Cox proportional hazards modeling, GC use was associated with a 17% increased risk for hypertension (hazard ratio, 1.17; 95% CI, 1.10-1.24).

The researchers noted that 40% of GC users with hypertension were not prescribed an antihypertensive agent at any point during the study. “Whilst some may have been offered lifestyle advice, left untreated this has important implications in terms of addressing modifiable risk factors in an RA population already at increased risk of CV disease,” they wrote.

They noted that cardiovascular disease is a major driver of the elevated mortality risk seen among adults with RA compared with the general population and that recent treatment recommendations address management of cardiovascular risks in these patients.

“There are several routes by which GCs may promote cardiovascular disease, including hypertension, metabolic changes, diabetes, and weight gain. We don’t currently know the extent to which each of these individual mechanisms may be increasing cardiovascular disease,” said Ms. Costello.

“Glucocorticoids increase fluid retention and promote obesity and hypertension,” said Rajat S. Bhatt, MD, a rheumatologist at Prime Rheumatology and Memorial Hermann Katy Hospital in Richmond, Texas, who sees hypertension in GC users in his clinical practice. “So patients need to be monitored for these risk factors,” he said in an interview.

Although hypertension may be a significant factor in the increase in cardiovascular disease in the RA population, Dr. Bhatt said the major driver is likely the intrinsic inflammatory state caused by the disease itself. As to why the GC-hypertension connection has flown under the radar in RA, he added, “That specific link has been difficult to tease out since RA patients are often on multiple medications.”

In regard to the role of dosage, Dr. Bhatt said that hypertension risk increases with higher GC doses, as the U.K. study indicates, and usually subsides when patients stop using GCs.

“Whether the observed dose association is causal or influenced by the underlying disease severity, our results suggest we should be vigilant in patients on all doses of GC, especially higher doses,” Ms. Costello added.

In regard to using drugs that are less cardiotoxic than GCs, Dr. Bhatt said that there are clinical scenarios in which GC therapy is the best choice, so just switching to nonsteroidal drugs is no panacea. “All RA drugs have adverse side effects, and anyway, the goal of rheumatology treatment is always to get patients off corticosteroids as soon as possible,” he said.

Ms. Costello and colleagues noted that their results are consonant with earlier research, including a single-center, cross-sectional study in which less than 6 months’ use of prednisolone at a median dose of 7.5 mg was associated with hypertension. In a German registry study, among patients who received doses of less than 7.5 mg for less than 6 months, there were higher rates of self-reported elevations in blood pressure.

The findings are at odds, however, with a recent matched-cohort study, which also used data from the Clinical Practice Research Datalink. That study found no association between GC use and hypertension.

GCs have come under increasing scrutiny in regard to several diseases. A study published July 7 found that even short-term courses of a few days’ duration entail risks for serious adverse events.

Ms. Costello’s group says that an estimate of GC-related incident hypertension in RA should allow more informed treatment decisions and that their findings highlight the ongoing need to monitor for and address this risk.

The study was supported by the Centre for Epidemiology Versus Arthritis and by the National Institute for Health Research Manchester Biomedical Research Centre. Coauthor William G. Dixon, PhD, has received consultancy fees from Google and Bayer unrelated to this study. Dr. Bhatt has disclosed no relevant financial relationships.

SOURCE: Costello RE et al. Rheumatology. 2020 June 27. doi: 10.1093/rheumatology/keaa209.

A version of this article originally appeared on Medscape.com.

Proposed changes in the eligibility criteria for lung cancer screening with low-dose CT could result in a near doubling of eligibility, and could especially benefit Black persons and women, say lung cancer experts.

The new proposals include lowering the age at which screening starts from 55 to 50 years, and to reduce the smoking history from 30 to 20 pack-years.

The draft recommendation from the United States Preventive Service Task Force (USPSTF) is available for public comment until August 3.

The task force recommends that adults age 50 to 80 who have a 20 pack-year or greater smoking history and currently smoke or have quit within the last 15 years undergo annual screening for lung cancer with low-dose CT,

“In my opinion, the proposed criteria by USPSTF represent a huge step in the right direction,” Lecia Sequist, MD, director of innovation at the Mass General Cancer Center in Boston, told Medscape Medical News.

“If these are adopted and implemented, we could see the benefit of screening (measured as reduction in lung cancer mortality) go from 9.8% with current parameters up to 13% with the broader parameters,” she said. “In addition, the new criteria should reduce racial disparities in screening eligibility.”

The recommendation also earned high marks from the American Lung Association.

The USPSTF has continued its ‘B’ recommendation – allowing for coverage of the screening with no cost for many under the Affordable Care Act – and is now proposing to expand the eligibility criteria “so that even more Americans at higher risk for lung cancer can be screened,” the ALA commented.

Start screening at 50  

Lowering the minimum age of screening to 50 would likely mean that more Black individuals and women would be eligible for screening, the recommendation authors contend. The current screening age of 55 is currently recommended under guidelines issued by the American Association for Thoracic Surgery, American Cancer Society, American College of Chest Physicians, and National Comprehensive Cancer Network.

“African Americans have a higher risk of lung cancer, compared with whites, and this risk difference is more apparent at lower levels of smoking intensity,” they write.

As previously reported by Medscape Medical News, lung cancer screening in an urban, largely black cohort yielded roughly double the rates of positive screens and detected lung cancers compared with results from the National Lung Screening Trial, which enrolled mostly White individuals.

In addition, although lung cancer risk is greater for men than women who smoke, and women generally accumulate fewer pack-years than men, there is evidence to suggest that women who smoke may develop lung cancer earlier and with lower levels of exposure.

Therefore, “a strategy of screening persons ages 50 to 80 years who have at least a 20 pack-year smoking history and currently smoke or have quit within the past 15 years (A-50-80-20-15) would lead to a relative increase in the percentage of persons eligible for screening by 81% in men and by 96% in women,” the proposed recommendation states.

What’s the harm?

One of the major concerns about low-dose CT screening for lung cancer is the relatively high rate of false-positive results reported in two large scale clinical trials, the recommendation authors acknowledged.

For example, in the NLST, which was the basis for an earlier USPSTF recommendation (for annual screening of adults 55 to 80 years of age who have a 30 pack-year smoking history and currently smoke or have quit in the previous 15 years), the false-positive rates were 26.3% at baseline, 27.2% at year 1, and 15.9% at year 2.

Similarly, in the NELSON trial, results of which were published earlier this year, false-positive rates for men were 19.8% at baseline, 7.1% at year 1, 9% at year 3, and 3.9% at year 5.5 of screening, they noted.

“Yes, false-positive results are one of the things we need to think carefully about when embarking on lung screening,” Dr. Sequist told Medscape Medical News. “The potential harm of a false-positive (unnecessary scans, biopsies or even surgery) can be minimized by having a multidisciplinary team with experience working up lung nodules see patients who have a positive screening test. In fact, the American College of Radiology recommends that all lung screening programs be paired with such a team.”

Mass General has a pulmonary nodule clinic to evaluate screen-detected lung nodules, with the goal of minimizing unnecessary procedures, she noted.

Asked about the potential harm from radiation exposure, Sequist said that exposure from low-dose CT screening is fairly minimal, comparable to that from solar radiation at sea level over a 6-month period, or about the level from three cross-country airplane trips.

“While it is not zero radiation, there is very little concern that this low level of radiation would cause a cancer or damage one’s lungs,” she said.

Albert Rizzo, MD, chief medical officer of the ALA, said that the potential harms of unnecessary interventions are outweighed by the benefits of detecting lung cancer at an early stage.

“I think what has been learned over the last 5 years is that the original recommendations that were put out really allowed the overall rate of positivity well within what’s seen with mammography, for example, and the number of patients who have needlessly gone on to procedures remains very low, and the morbidity of those procedures remains low as well,” Dr. Rizzo told Medscape Medical News.

Not enough takers

Despite the clear benefits of low-dose CT screening, however, US screening rates for high-risk individuals are still very low, ranging from 12.3% in Massachusetts to a low of 0.5% in Nevada, according to a 2019 research report on the state of lung cancer from the ALA.

“For screening to be most effective, more of the high-risk population should be screened. Currently, screening rates are very low among those at high risk. This may be because of a lack of access or low awareness and knowledge among patients and providers. As rates vary tremendously between states, it is clear that more can be done to increase screening rates,” the report stated.

“I think that there are some mixed messages sent out into the population as to whether or not an individual patient should be screened,” Dr. Rizzo said.

He noted that some physicians may be reluctant to take on the nuanced risk–benefit discussion required, or may not have the time during a brief patient visit.

“It really boils down to that discussion between the physician and the patient who falls under these risk categories, to say, ‘Look, this is what these studies have found, and you fall under a category where if we find a cancer early, it’s very likely you’re going to be saved,’ as compared for waiting for it to present by itself,” he said.

Dr. Sequist and Dr. Rizzo have disclosed no relevant financial relationships.

This article first appeared on Medscape.com.

Proposed changes in the eligibility criteria for lung cancer screening with low-dose CT could result in a near doubling of eligibility, and could especially benefit Black persons and women, say lung cancer experts.

The new proposals include lowering the age at which screening starts from 55 to 50 years, and to reduce the smoking history from 30 to 20 pack-years.

The draft recommendation from the United States Preventive Service Task Force (USPSTF) is available for public comment until August 3.

The task force recommends that adults age 50 to 80 who have a 20 pack-year or greater smoking history and currently smoke or have quit within the last 15 years undergo annual screening for lung cancer with low-dose CT,

“In my opinion, the proposed criteria by USPSTF represent a huge step in the right direction,” Lecia Sequist, MD, director of innovation at the Mass General Cancer Center in Boston, told Medscape Medical News.

“If these are adopted and implemented, we could see the benefit of screening (measured as reduction in lung cancer mortality) go from 9.8% with current parameters up to 13% with the broader parameters,” she said. “In addition, the new criteria should reduce racial disparities in screening eligibility.”

The recommendation also earned high marks from the American Lung Association.

The USPSTF has continued its ‘B’ recommendation – allowing for coverage of the screening with no cost for many under the Affordable Care Act – and is now proposing to expand the eligibility criteria “so that even more Americans at higher risk for lung cancer can be screened,” the ALA commented.

Start screening at 50  

Lowering the minimum age of screening to 50 would likely mean that more Black individuals and women would be eligible for screening, the recommendation authors contend. The current screening age of 55 is currently recommended under guidelines issued by the American Association for Thoracic Surgery, American Cancer Society, American College of Chest Physicians, and National Comprehensive Cancer Network.

“African Americans have a higher risk of lung cancer, compared with whites, and this risk difference is more apparent at lower levels of smoking intensity,” they write.

As previously reported by Medscape Medical News, lung cancer screening in an urban, largely black cohort yielded roughly double the rates of positive screens and detected lung cancers compared with results from the National Lung Screening Trial, which enrolled mostly White individuals.

In addition, although lung cancer risk is greater for men than women who smoke, and women generally accumulate fewer pack-years than men, there is evidence to suggest that women who smoke may develop lung cancer earlier and with lower levels of exposure.

Therefore, “a strategy of screening persons ages 50 to 80 years who have at least a 20 pack-year smoking history and currently smoke or have quit within the past 15 years (A-50-80-20-15) would lead to a relative increase in the percentage of persons eligible for screening by 81% in men and by 96% in women,” the proposed recommendation states.

What’s the harm?

One of the major concerns about low-dose CT screening for lung cancer is the relatively high rate of false-positive results reported in two large scale clinical trials, the recommendation authors acknowledged.

For example, in the NLST, which was the basis for an earlier USPSTF recommendation (for annual screening of adults 55 to 80 years of age who have a 30 pack-year smoking history and currently smoke or have quit in the previous 15 years), the false-positive rates were 26.3% at baseline, 27.2% at year 1, and 15.9% at year 2.

Similarly, in the NELSON trial, results of which were published earlier this year, false-positive rates for men were 19.8% at baseline, 7.1% at year 1, 9% at year 3, and 3.9% at year 5.5 of screening, they noted.

“Yes, false-positive results are one of the things we need to think carefully about when embarking on lung screening,” Dr. Sequist told Medscape Medical News. “The potential harm of a false-positive (unnecessary scans, biopsies or even surgery) can be minimized by having a multidisciplinary team with experience working up lung nodules see patients who have a positive screening test. In fact, the American College of Radiology recommends that all lung screening programs be paired with such a team.”

Mass General has a pulmonary nodule clinic to evaluate screen-detected lung nodules, with the goal of minimizing unnecessary procedures, she noted.

Asked about the potential harm from radiation exposure, Sequist said that exposure from low-dose CT screening is fairly minimal, comparable to that from solar radiation at sea level over a 6-month period, or about the level from three cross-country airplane trips.

“While it is not zero radiation, there is very little concern that this low level of radiation would cause a cancer or damage one’s lungs,” she said.

Albert Rizzo, MD, chief medical officer of the ALA, said that the potential harms of unnecessary interventions are outweighed by the benefits of detecting lung cancer at an early stage.

“I think what has been learned over the last 5 years is that the original recommendations that were put out really allowed the overall rate of positivity well within what’s seen with mammography, for example, and the number of patients who have needlessly gone on to procedures remains very low, and the morbidity of those procedures remains low as well,” Dr. Rizzo told Medscape Medical News.

Not enough takers

Despite the clear benefits of low-dose CT screening, however, US screening rates for high-risk individuals are still very low, ranging from 12.3% in Massachusetts to a low of 0.5% in Nevada, according to a 2019 research report on the state of lung cancer from the ALA.

“For screening to be most effective, more of the high-risk population should be screened. Currently, screening rates are very low among those at high risk. This may be because of a lack of access or low awareness and knowledge among patients and providers. As rates vary tremendously between states, it is clear that more can be done to increase screening rates,” the report stated.

“I think that there are some mixed messages sent out into the population as to whether or not an individual patient should be screened,” Dr. Rizzo said.

He noted that some physicians may be reluctant to take on the nuanced risk–benefit discussion required, or may not have the time during a brief patient visit.

“It really boils down to that discussion between the physician and the patient who falls under these risk categories, to say, ‘Look, this is what these studies have found, and you fall under a category where if we find a cancer early, it’s very likely you’re going to be saved,’ as compared for waiting for it to present by itself,” he said.

Dr. Sequist and Dr. Rizzo have disclosed no relevant financial relationships.

This article first appeared on Medscape.com.

Proposed changes in the eligibility criteria for lung cancer screening with low-dose CT could result in a near doubling of eligibility, and could especially benefit Black persons and women, say lung cancer experts.

The new proposals include lowering the age at which screening starts from 55 to 50 years, and to reduce the smoking history from 30 to 20 pack-years.

The draft recommendation from the United States Preventive Service Task Force (USPSTF) is available for public comment until August 3.

The task force recommends that adults age 50 to 80 who have a 20 pack-year or greater smoking history and currently smoke or have quit within the last 15 years undergo annual screening for lung cancer with low-dose CT,

“In my opinion, the proposed criteria by USPSTF represent a huge step in the right direction,” Lecia Sequist, MD, director of innovation at the Mass General Cancer Center in Boston, told Medscape Medical News.

“If these are adopted and implemented, we could see the benefit of screening (measured as reduction in lung cancer mortality) go from 9.8% with current parameters up to 13% with the broader parameters,” she said. “In addition, the new criteria should reduce racial disparities in screening eligibility.”

The recommendation also earned high marks from the American Lung Association.

The USPSTF has continued its ‘B’ recommendation – allowing for coverage of the screening with no cost for many under the Affordable Care Act – and is now proposing to expand the eligibility criteria “so that even more Americans at higher risk for lung cancer can be screened,” the ALA commented.

Start screening at 50  

Lowering the minimum age of screening to 50 would likely mean that more Black individuals and women would be eligible for screening, the recommendation authors contend. The current screening age of 55 is currently recommended under guidelines issued by the American Association for Thoracic Surgery, American Cancer Society, American College of Chest Physicians, and National Comprehensive Cancer Network.

“African Americans have a higher risk of lung cancer, compared with whites, and this risk difference is more apparent at lower levels of smoking intensity,” they write.

As previously reported by Medscape Medical News, lung cancer screening in an urban, largely black cohort yielded roughly double the rates of positive screens and detected lung cancers compared with results from the National Lung Screening Trial, which enrolled mostly White individuals.

In addition, although lung cancer risk is greater for men than women who smoke, and women generally accumulate fewer pack-years than men, there is evidence to suggest that women who smoke may develop lung cancer earlier and with lower levels of exposure.

Therefore, “a strategy of screening persons ages 50 to 80 years who have at least a 20 pack-year smoking history and currently smoke or have quit within the past 15 years (A-50-80-20-15) would lead to a relative increase in the percentage of persons eligible for screening by 81% in men and by 96% in women,” the proposed recommendation states.

What’s the harm?

One of the major concerns about low-dose CT screening for lung cancer is the relatively high rate of false-positive results reported in two large scale clinical trials, the recommendation authors acknowledged.

For example, in the NLST, which was the basis for an earlier USPSTF recommendation (for annual screening of adults 55 to 80 years of age who have a 30 pack-year smoking history and currently smoke or have quit in the previous 15 years), the false-positive rates were 26.3% at baseline, 27.2% at year 1, and 15.9% at year 2.

Similarly, in the NELSON trial, results of which were published earlier this year, false-positive rates for men were 19.8% at baseline, 7.1% at year 1, 9% at year 3, and 3.9% at year 5.5 of screening, they noted.

“Yes, false-positive results are one of the things we need to think carefully about when embarking on lung screening,” Dr. Sequist told Medscape Medical News. “The potential harm of a false-positive (unnecessary scans, biopsies or even surgery) can be minimized by having a multidisciplinary team with experience working up lung nodules see patients who have a positive screening test. In fact, the American College of Radiology recommends that all lung screening programs be paired with such a team.”

Mass General has a pulmonary nodule clinic to evaluate screen-detected lung nodules, with the goal of minimizing unnecessary procedures, she noted.

Asked about the potential harm from radiation exposure, Sequist said that exposure from low-dose CT screening is fairly minimal, comparable to that from solar radiation at sea level over a 6-month period, or about the level from three cross-country airplane trips.

“While it is not zero radiation, there is very little concern that this low level of radiation would cause a cancer or damage one’s lungs,” she said.

Albert Rizzo, MD, chief medical officer of the ALA, said that the potential harms of unnecessary interventions are outweighed by the benefits of detecting lung cancer at an early stage.

“I think what has been learned over the last 5 years is that the original recommendations that were put out really allowed the overall rate of positivity well within what’s seen with mammography, for example, and the number of patients who have needlessly gone on to procedures remains very low, and the morbidity of those procedures remains low as well,” Dr. Rizzo told Medscape Medical News.

Not enough takers

Despite the clear benefits of low-dose CT screening, however, US screening rates for high-risk individuals are still very low, ranging from 12.3% in Massachusetts to a low of 0.5% in Nevada, according to a 2019 research report on the state of lung cancer from the ALA.

“For screening to be most effective, more of the high-risk population should be screened. Currently, screening rates are very low among those at high risk. This may be because of a lack of access or low awareness and knowledge among patients and providers. As rates vary tremendously between states, it is clear that more can be done to increase screening rates,” the report stated.

“I think that there are some mixed messages sent out into the population as to whether or not an individual patient should be screened,” Dr. Rizzo said.

He noted that some physicians may be reluctant to take on the nuanced risk–benefit discussion required, or may not have the time during a brief patient visit.

“It really boils down to that discussion between the physician and the patient who falls under these risk categories, to say, ‘Look, this is what these studies have found, and you fall under a category where if we find a cancer early, it’s very likely you’re going to be saved,’ as compared for waiting for it to present by itself,” he said.

Dr. Sequist and Dr. Rizzo have disclosed no relevant financial relationships.

This article first appeared on Medscape.com.

Opioid overdose deaths in the United States have more than tripled in recent years, from 6.1 deaths per 100,000 individuals in 1999 to 20.7 per 100,000 individuals in 2018.¹ Although the availability of medication-assisted treatment (MAT) has expanded over the past decade, this lifesaving treatment remains largely inaccessible to some of the most vulnerable members of our communities: opioid users facing reentry after incarceration.

Just as abstinence in the community brings a loss of tolerance to opioids, individuals who are incarcerated lose tolerance as well. Clinicians who treat patients with opioid use disorders (OUD) are accustomed to warning patients about the risk of returning to prior levels of use too quickly. Harm reduction strategies include using slowly, using with friends, and having naloxone on hand to prevent unintended overdose.

The risks of opioid use are magnified for those facing reentry; incarceration contributes to a loss of employment, social supports, and connection to care. Those changes can create an exceptionally stressful reentry period – one that places individuals at an acutely high risk of relapse and overdose. Within the first 2 years of release, an individual with a history of incarceration has a risk of death 3.5 times higher than that of someone in the general population. Within the first 2 weeks, those recently incarcerated are 129 times more likely to overdose on opioids and 12.7 times more likely to die than members of the general population.2

Treatment with MAT dramatically reduces deaths during this crucial period. In Rhode Island, for example, implementation of a corrections MAT program decreased all-cause death rates after release by an astonishing 60.5%.³ In England, large national studies have shown a similar 75% decrease in all-cause mortality within the first 4 weeks of release among individuals with OUD.⁴ In California, the counties with the highest overdose death rates are consistently those with fewer opioid treatment programs, which suggests that access to treatment is necessary to prolong the lives of those suffering from OUD.⁵ In-custody overdose deaths are quite rare, and access to MAT during incarceration has decreased in-custody deaths by 74%.⁶

Decreased opioid overdose deaths is not the only outcome of MAT. Pharmacotherapy for OUD also has been shown to increase treatment retention,⁷ reduce reincarceration,⁸ prevent communicable infections,⁹ and decrease use of other illicit substances.¹⁰ The provision of MAT also has been shown to be cost effective.¹¹

Despite those benefits, as of 2017, only 30 out of 5,100 jails and prisons in the United States provided treatment with methadone or buprenorphine.¹² When individuals on maintenance therapy are incarcerated, most correctional facilities force them to taper and discontinue those medications. This practice can cause distressing withdrawal symptoms and actively increase the risk of death for these individuals.

Concerns related to the provision of MAT, and specifically buprenorphine, in the correctional health setting often are related to diversion. Although safe administration of opioid full and partial agonists is a priority, recent literature has suggested that buprenorphine is not a medication frequently used for euphoric properties. In fact, the literature suggests that individuals using illicit buprenorphine primarily do so to treat withdrawal symptoms and that illicit use diminishes with access to formal treatment.^13,14

Another concern is that pharmacotherapy for OUD should not be used without adjunctive psychotherapies and social supports. While dual pharmacotherapy and psychotherapy is ideal, the American Society for Addiction Medicine 2020 National Practice Guidelines for the treatment of OUD state: “a patient’s decision to decline psychosocial treatment or the absence of available psychosocial treatment should not preclude or delay pharmacotherapy, with appropriate medication management.”¹⁵ Just as some patients wish to engage in mutual help or psychotherapeutic modalities only, some patients wish to engage only in psychopharmacologic interventions. Declaring one modality of treatment better, or worse, or more worthwhile is not borne out by the literature and often places clinicians’ preferences over the preferences of patients.

Individuals who suffer from substance use disorders are at high risk of incarceration, relapse, and overdose death. These patients also suffer from stigmatization from peers and health care workers alike, making the process of engaging in care incredibly burdensome. Because of the disease of addiction, many of our patients cannot envision a healthy future: a future with the potential for intimate relationships, meaningful community engagement, and a rich inner life. The provision of MAT is lifesaving and improves the chances of a successful reentry – an intuitive first step in a long, but worthwhile, journey.
 

References

1. Hedegaard H et al; National Center for Health Statistics. Drug overdose deaths in the United States, 1999–2018. NCHS Data Brief, 2020 Jan, No. 356.

2. Binswanger IA et al. N Engl J Med. 2007;356:157-65.

3. Green TC et al. JAMA Psychiatry. 2018;75(4):405-7.

4. Marsden J et al. Addiction. 2017;112(8):1408-18.

5. Joshi V and Urada D. State Targeted Response to the Opioid Crisis: California Strategic Plan. 2017 Aug 30.

6. Larney S et al. BMJ Open. 2014. doi: 10.1136/bmjopen-2013-004666.

7. Rich JD et al. Lancet. 2015;386(9991):350-9.

8. Deck D et al. J Addict Dis. 2009. 28(2):89-102.

9. MacArthur GJ et al. BMJ. 2012. doi: 10.1136/bmj.e5945.

10. Tsui J et al. J Subst Abuse Treat. 2019. 109:80-5.

11. Gisev N et al. Addiction. 2015 Dec;110(12):1975-84.

12. National Mental Health and Substance Use Policy Laboratory. “Use of Medication-Assisted Treatment for Opioid Use Disorder in Criminal Justice Settings.” HHS Publication No. PEP19-MATUSECJS. Rockville, Md.: Substance Abuse and Mental Health Services Administration, 2019.

13. Bazazi AR et al. J Addict Med. 2011;5(3):175-80.

14. Schuman-Olivier Z. et al. J Subst Abuse Treat. 2010 Jul;39(1):41-50.

15. Crotty K et al. J Addict Med. 2020;14(2)99-112.
 

Dr. Barnes is chief resident at San Mateo County Behavioral Health and Recovery Services in California. He disclosed no relevant financial relationships. Dr. Lenane is resident* at San Mateo County Behavioral Health and Recovery Services. He disclosed no relevant financial relationships. The opinions shared in this article represent the viewpoints of the authors and are not necessarily representative of the viewpoints or policies of their academic program or employer.

*This article was updated 7/9/2020.

Opioid overdose deaths in the United States have more than tripled in recent years, from 6.1 deaths per 100,000 individuals in 1999 to 20.7 per 100,000 individuals in 2018.¹ Although the availability of medication-assisted treatment (MAT) has expanded over the past decade, this lifesaving treatment remains largely inaccessible to some of the most vulnerable members of our communities: opioid users facing reentry after incarceration.

Just as abstinence in the community brings a loss of tolerance to opioids, individuals who are incarcerated lose tolerance as well. Clinicians who treat patients with opioid use disorders (OUD) are accustomed to warning patients about the risk of returning to prior levels of use too quickly. Harm reduction strategies include using slowly, using with friends, and having naloxone on hand to prevent unintended overdose.

The risks of opioid use are magnified for those facing reentry; incarceration contributes to a loss of employment, social supports, and connection to care. Those changes can create an exceptionally stressful reentry period – one that places individuals at an acutely high risk of relapse and overdose. Within the first 2 years of release, an individual with a history of incarceration has a risk of death 3.5 times higher than that of someone in the general population. Within the first 2 weeks, those recently incarcerated are 129 times more likely to overdose on opioids and 12.7 times more likely to die than members of the general population.2

Treatment with MAT dramatically reduces deaths during this crucial period. In Rhode Island, for example, implementation of a corrections MAT program decreased all-cause death rates after release by an astonishing 60.5%.³ In England, large national studies have shown a similar 75% decrease in all-cause mortality within the first 4 weeks of release among individuals with OUD.⁴ In California, the counties with the highest overdose death rates are consistently those with fewer opioid treatment programs, which suggests that access to treatment is necessary to prolong the lives of those suffering from OUD.⁵ In-custody overdose deaths are quite rare, and access to MAT during incarceration has decreased in-custody deaths by 74%.⁶

Decreased opioid overdose deaths is not the only outcome of MAT. Pharmacotherapy for OUD also has been shown to increase treatment retention,⁷ reduce reincarceration,⁸ prevent communicable infections,⁹ and decrease use of other illicit substances.¹⁰ The provision of MAT also has been shown to be cost effective.¹¹

Despite those benefits, as of 2017, only 30 out of 5,100 jails and prisons in the United States provided treatment with methadone or buprenorphine.¹² When individuals on maintenance therapy are incarcerated, most correctional facilities force them to taper and discontinue those medications. This practice can cause distressing withdrawal symptoms and actively increase the risk of death for these individuals.

Concerns related to the provision of MAT, and specifically buprenorphine, in the correctional health setting often are related to diversion. Although safe administration of opioid full and partial agonists is a priority, recent literature has suggested that buprenorphine is not a medication frequently used for euphoric properties. In fact, the literature suggests that individuals using illicit buprenorphine primarily do so to treat withdrawal symptoms and that illicit use diminishes with access to formal treatment.^13,14

Another concern is that pharmacotherapy for OUD should not be used without adjunctive psychotherapies and social supports. While dual pharmacotherapy and psychotherapy is ideal, the American Society for Addiction Medicine 2020 National Practice Guidelines for the treatment of OUD state: “a patient’s decision to decline psychosocial treatment or the absence of available psychosocial treatment should not preclude or delay pharmacotherapy, with appropriate medication management.”¹⁵ Just as some patients wish to engage in mutual help or psychotherapeutic modalities only, some patients wish to engage only in psychopharmacologic interventions. Declaring one modality of treatment better, or worse, or more worthwhile is not borne out by the literature and often places clinicians’ preferences over the preferences of patients.

Individuals who suffer from substance use disorders are at high risk of incarceration, relapse, and overdose death. These patients also suffer from stigmatization from peers and health care workers alike, making the process of engaging in care incredibly burdensome. Because of the disease of addiction, many of our patients cannot envision a healthy future: a future with the potential for intimate relationships, meaningful community engagement, and a rich inner life. The provision of MAT is lifesaving and improves the chances of a successful reentry – an intuitive first step in a long, but worthwhile, journey.
 

References

1. Hedegaard H et al; National Center for Health Statistics. Drug overdose deaths in the United States, 1999–2018. NCHS Data Brief, 2020 Jan, No. 356.

2. Binswanger IA et al. N Engl J Med. 2007;356:157-65.

3. Green TC et al. JAMA Psychiatry. 2018;75(4):405-7.

4. Marsden J et al. Addiction. 2017;112(8):1408-18.

5. Joshi V and Urada D. State Targeted Response to the Opioid Crisis: California Strategic Plan. 2017 Aug 30.

6. Larney S et al. BMJ Open. 2014. doi: 10.1136/bmjopen-2013-004666.

7. Rich JD et al. Lancet. 2015;386(9991):350-9.

8. Deck D et al. J Addict Dis. 2009. 28(2):89-102.

9. MacArthur GJ et al. BMJ. 2012. doi: 10.1136/bmj.e5945.

10. Tsui J et al. J Subst Abuse Treat. 2019. 109:80-5.

11. Gisev N et al. Addiction. 2015 Dec;110(12):1975-84.

12. National Mental Health and Substance Use Policy Laboratory. “Use of Medication-Assisted Treatment for Opioid Use Disorder in Criminal Justice Settings.” HHS Publication No. PEP19-MATUSECJS. Rockville, Md.: Substance Abuse and Mental Health Services Administration, 2019.

13. Bazazi AR et al. J Addict Med. 2011;5(3):175-80.

14. Schuman-Olivier Z. et al. J Subst Abuse Treat. 2010 Jul;39(1):41-50.

15. Crotty K et al. J Addict Med. 2020;14(2)99-112.
 

Dr. Barnes is chief resident at San Mateo County Behavioral Health and Recovery Services in California. He disclosed no relevant financial relationships. Dr. Lenane is resident* at San Mateo County Behavioral Health and Recovery Services. He disclosed no relevant financial relationships. The opinions shared in this article represent the viewpoints of the authors and are not necessarily representative of the viewpoints or policies of their academic program or employer.

*This article was updated 7/9/2020.

Opioid overdose deaths in the United States have more than tripled in recent years, from 6.1 deaths per 100,000 individuals in 1999 to 20.7 per 100,000 individuals in 2018.¹ Although the availability of medication-assisted treatment (MAT) has expanded over the past decade, this lifesaving treatment remains largely inaccessible to some of the most vulnerable members of our communities: opioid users facing reentry after incarceration.

Just as abstinence in the community brings a loss of tolerance to opioids, individuals who are incarcerated lose tolerance as well. Clinicians who treat patients with opioid use disorders (OUD) are accustomed to warning patients about the risk of returning to prior levels of use too quickly. Harm reduction strategies include using slowly, using with friends, and having naloxone on hand to prevent unintended overdose.

The risks of opioid use are magnified for those facing reentry; incarceration contributes to a loss of employment, social supports, and connection to care. Those changes can create an exceptionally stressful reentry period – one that places individuals at an acutely high risk of relapse and overdose. Within the first 2 years of release, an individual with a history of incarceration has a risk of death 3.5 times higher than that of someone in the general population. Within the first 2 weeks, those recently incarcerated are 129 times more likely to overdose on opioids and 12.7 times more likely to die than members of the general population.2

Treatment with MAT dramatically reduces deaths during this crucial period. In Rhode Island, for example, implementation of a corrections MAT program decreased all-cause death rates after release by an astonishing 60.5%.³ In England, large national studies have shown a similar 75% decrease in all-cause mortality within the first 4 weeks of release among individuals with OUD.⁴ In California, the counties with the highest overdose death rates are consistently those with fewer opioid treatment programs, which suggests that access to treatment is necessary to prolong the lives of those suffering from OUD.⁵ In-custody overdose deaths are quite rare, and access to MAT during incarceration has decreased in-custody deaths by 74%.⁶

Decreased opioid overdose deaths is not the only outcome of MAT. Pharmacotherapy for OUD also has been shown to increase treatment retention,⁷ reduce reincarceration,⁸ prevent communicable infections,⁹ and decrease use of other illicit substances.¹⁰ The provision of MAT also has been shown to be cost effective.¹¹

Despite those benefits, as of 2017, only 30 out of 5,100 jails and prisons in the United States provided treatment with methadone or buprenorphine.¹² When individuals on maintenance therapy are incarcerated, most correctional facilities force them to taper and discontinue those medications. This practice can cause distressing withdrawal symptoms and actively increase the risk of death for these individuals.

Concerns related to the provision of MAT, and specifically buprenorphine, in the correctional health setting often are related to diversion. Although safe administration of opioid full and partial agonists is a priority, recent literature has suggested that buprenorphine is not a medication frequently used for euphoric properties. In fact, the literature suggests that individuals using illicit buprenorphine primarily do so to treat withdrawal symptoms and that illicit use diminishes with access to formal treatment.^13,14

Another concern is that pharmacotherapy for OUD should not be used without adjunctive psychotherapies and social supports. While dual pharmacotherapy and psychotherapy is ideal, the American Society for Addiction Medicine 2020 National Practice Guidelines for the treatment of OUD state: “a patient’s decision to decline psychosocial treatment or the absence of available psychosocial treatment should not preclude or delay pharmacotherapy, with appropriate medication management.”¹⁵ Just as some patients wish to engage in mutual help or psychotherapeutic modalities only, some patients wish to engage only in psychopharmacologic interventions. Declaring one modality of treatment better, or worse, or more worthwhile is not borne out by the literature and often places clinicians’ preferences over the preferences of patients.

Individuals who suffer from substance use disorders are at high risk of incarceration, relapse, and overdose death. These patients also suffer from stigmatization from peers and health care workers alike, making the process of engaging in care incredibly burdensome. Because of the disease of addiction, many of our patients cannot envision a healthy future: a future with the potential for intimate relationships, meaningful community engagement, and a rich inner life. The provision of MAT is lifesaving and improves the chances of a successful reentry – an intuitive first step in a long, but worthwhile, journey.
 

References

1. Hedegaard H et al; National Center for Health Statistics. Drug overdose deaths in the United States, 1999–2018. NCHS Data Brief, 2020 Jan, No. 356.

2. Binswanger IA et al. N Engl J Med. 2007;356:157-65.

3. Green TC et al. JAMA Psychiatry. 2018;75(4):405-7.

4. Marsden J et al. Addiction. 2017;112(8):1408-18.

5. Joshi V and Urada D. State Targeted Response to the Opioid Crisis: California Strategic Plan. 2017 Aug 30.

6. Larney S et al. BMJ Open. 2014. doi: 10.1136/bmjopen-2013-004666.

7. Rich JD et al. Lancet. 2015;386(9991):350-9.

8. Deck D et al. J Addict Dis. 2009. 28(2):89-102.

9. MacArthur GJ et al. BMJ. 2012. doi: 10.1136/bmj.e5945.

10. Tsui J et al. J Subst Abuse Treat. 2019. 109:80-5.

11. Gisev N et al. Addiction. 2015 Dec;110(12):1975-84.

12. National Mental Health and Substance Use Policy Laboratory. “Use of Medication-Assisted Treatment for Opioid Use Disorder in Criminal Justice Settings.” HHS Publication No. PEP19-MATUSECJS. Rockville, Md.: Substance Abuse and Mental Health Services Administration, 2019.

13. Bazazi AR et al. J Addict Med. 2011;5(3):175-80.

14. Schuman-Olivier Z. et al. J Subst Abuse Treat. 2010 Jul;39(1):41-50.

15. Crotty K et al. J Addict Med. 2020;14(2)99-112.
 

Dr. Barnes is chief resident at San Mateo County Behavioral Health and Recovery Services in California. He disclosed no relevant financial relationships. Dr. Lenane is resident* at San Mateo County Behavioral Health and Recovery Services. He disclosed no relevant financial relationships. The opinions shared in this article represent the viewpoints of the authors and are not necessarily representative of the viewpoints or policies of their academic program or employer.

*This article was updated 7/9/2020.

Transitioning from parenteral bortezomib-based induction to all-oral ixazomib-lenalidomide-dexamethasone therapy increased proteasome inhibitor (PI)–based treatment adherence and duration, according to early results from a clinical trial designed to include patients representing the real-world U.S. multiple myeloma population.

The US MM-6 study was designed to evaluate a novel in-class therapy (iCT) transitioning approach from intravenous to oral treatment in the community-based setting with the aims of increasing PI-based treatment duration and adherence, maintaining health-related quality of life (HRQoL), and improving outcomes in a representative, real-world, community population of multiple myeloma patients, according to Sudhir Manda, MD, of Arizona Oncology/U.S. Oncology Research, Tucson, and colleagues.

Dr. Manda and colleagues reported on the early results of the US MM-6 trial (NCT03173092), which is a community-based, real-world, open-label, single-arm, phase 4 study of adult multiple myeloma patients who do not meet transplant-eligibility criteria, or for whom transplant would be delayed for 2 years or more, and who are receiving first-line bortezomib-based induction. All patients in the study had no evidence of progressive disease after three treatment cycles.

By the data cutoff for the reported analysis, 84 patients had been treated. The patients had a median age of 73 years; 49% were men; 15% black/African American; 10% Hispanic/Latino. A total of 62% of the patients remain on therapy, with a mean duration of total PI therapy of 10.1 months and of ixazomib-lenalidomide-dexamethasone (ixazomib-Rd) of 7.3 months.

The overall response rate was 62% (complete response, 4%; very good partial response, 25%; partial response, 33%) after bortezomib-based induction and 70% (complete response, 26%; very good partial response, 29%; partial response, 15%) after induction to all-oral ixazomib-Rd.

“The use of this novel iCT approach from parenteral bortezomib-based to oral ixazomib-based therapy facilitates long-term PI-based treatment that is well tolerated in real-world, nontransplant [newly diagnosed multiple myeloma] patients,” according to Dr. Manda and colleagues. In addition, “preliminary findings indicate that the iCT approach results in promising efficacy and high medication adherence, with no adverse impact on patients’ HRQoL or treatment satisfaction.”

The study was sponsored by Millennium Pharmaceuticals. Four of the authors are employees of Millennium Pharmaceuticals and several authors disclosed relationships with various pharmaceutical companies, including Millennium Pharmaceuticals.

mlesney@mdedge.com

SOURCE: Manda S et al. Clin Lymphoma Myeloma Leuk. 2020 Jun 30. doi: 10.1016/j.clml.2020.06.024.

Transitioning from parenteral bortezomib-based induction to all-oral ixazomib-lenalidomide-dexamethasone therapy increased proteasome inhibitor (PI)–based treatment adherence and duration, according to early results from a clinical trial designed to include patients representing the real-world U.S. multiple myeloma population.

The US MM-6 study was designed to evaluate a novel in-class therapy (iCT) transitioning approach from intravenous to oral treatment in the community-based setting with the aims of increasing PI-based treatment duration and adherence, maintaining health-related quality of life (HRQoL), and improving outcomes in a representative, real-world, community population of multiple myeloma patients, according to Sudhir Manda, MD, of Arizona Oncology/U.S. Oncology Research, Tucson, and colleagues.

Dr. Manda and colleagues reported on the early results of the US MM-6 trial (NCT03173092), which is a community-based, real-world, open-label, single-arm, phase 4 study of adult multiple myeloma patients who do not meet transplant-eligibility criteria, or for whom transplant would be delayed for 2 years or more, and who are receiving first-line bortezomib-based induction. All patients in the study had no evidence of progressive disease after three treatment cycles.

By the data cutoff for the reported analysis, 84 patients had been treated. The patients had a median age of 73 years; 49% were men; 15% black/African American; 10% Hispanic/Latino. A total of 62% of the patients remain on therapy, with a mean duration of total PI therapy of 10.1 months and of ixazomib-lenalidomide-dexamethasone (ixazomib-Rd) of 7.3 months.

The overall response rate was 62% (complete response, 4%; very good partial response, 25%; partial response, 33%) after bortezomib-based induction and 70% (complete response, 26%; very good partial response, 29%; partial response, 15%) after induction to all-oral ixazomib-Rd.

“The use of this novel iCT approach from parenteral bortezomib-based to oral ixazomib-based therapy facilitates long-term PI-based treatment that is well tolerated in real-world, nontransplant [newly diagnosed multiple myeloma] patients,” according to Dr. Manda and colleagues. In addition, “preliminary findings indicate that the iCT approach results in promising efficacy and high medication adherence, with no adverse impact on patients’ HRQoL or treatment satisfaction.”

The study was sponsored by Millennium Pharmaceuticals. Four of the authors are employees of Millennium Pharmaceuticals and several authors disclosed relationships with various pharmaceutical companies, including Millennium Pharmaceuticals.

mlesney@mdedge.com

SOURCE: Manda S et al. Clin Lymphoma Myeloma Leuk. 2020 Jun 30. doi: 10.1016/j.clml.2020.06.024.

Transitioning from parenteral bortezomib-based induction to all-oral ixazomib-lenalidomide-dexamethasone therapy increased proteasome inhibitor (PI)–based treatment adherence and duration, according to early results from a clinical trial designed to include patients representing the real-world U.S. multiple myeloma population.

The US MM-6 study was designed to evaluate a novel in-class therapy (iCT) transitioning approach from intravenous to oral treatment in the community-based setting with the aims of increasing PI-based treatment duration and adherence, maintaining health-related quality of life (HRQoL), and improving outcomes in a representative, real-world, community population of multiple myeloma patients, according to Sudhir Manda, MD, of Arizona Oncology/U.S. Oncology Research, Tucson, and colleagues.

Dr. Manda and colleagues reported on the early results of the US MM-6 trial (NCT03173092), which is a community-based, real-world, open-label, single-arm, phase 4 study of adult multiple myeloma patients who do not meet transplant-eligibility criteria, or for whom transplant would be delayed for 2 years or more, and who are receiving first-line bortezomib-based induction. All patients in the study had no evidence of progressive disease after three treatment cycles.

By the data cutoff for the reported analysis, 84 patients had been treated. The patients had a median age of 73 years; 49% were men; 15% black/African American; 10% Hispanic/Latino. A total of 62% of the patients remain on therapy, with a mean duration of total PI therapy of 10.1 months and of ixazomib-lenalidomide-dexamethasone (ixazomib-Rd) of 7.3 months.

The overall response rate was 62% (complete response, 4%; very good partial response, 25%; partial response, 33%) after bortezomib-based induction and 70% (complete response, 26%; very good partial response, 29%; partial response, 15%) after induction to all-oral ixazomib-Rd.

“The use of this novel iCT approach from parenteral bortezomib-based to oral ixazomib-based therapy facilitates long-term PI-based treatment that is well tolerated in real-world, nontransplant [newly diagnosed multiple myeloma] patients,” according to Dr. Manda and colleagues. In addition, “preliminary findings indicate that the iCT approach results in promising efficacy and high medication adherence, with no adverse impact on patients’ HRQoL or treatment satisfaction.”

The study was sponsored by Millennium Pharmaceuticals. Four of the authors are employees of Millennium Pharmaceuticals and several authors disclosed relationships with various pharmaceutical companies, including Millennium Pharmaceuticals.

mlesney@mdedge.com

SOURCE: Manda S et al. Clin Lymphoma Myeloma Leuk. 2020 Jun 30. doi: 10.1016/j.clml.2020.06.024.

Extended use of corticosteroids for chronic inflammatory conditions puts patients at risk for serious adverse events (AEs), including cardiovascular disease, osteoporosis, cataracts, and diabetes. Now, a growing body of evidence suggests that even short bursts of these drugs are associated with serious risks.

Most recently, a population-based study of more than 2.6 million people found that taking corticosteroids for 14 days or less was associated with a substantially greater risk for gastrointestinal (GI) bleeding, sepsis, and heart failure, particularly within the first 30 days after therapy.

In the study, Tsung-Chieh Yao, MD, PhD, a professor in the division of allergy, asthma, and rheumatology in the department of pediatrics at Chang Gung Memorial Hospital in Taoyuan, Taiwan, and colleagues used a self-controlled case series to analyze data from Taiwan’s National Health Insurance Research Database of medical claims. They compared patients’ conditions in the period from 5 to 90 days before treatment to conditions from the periods from 5 to 30 days and from 31 to 90 days after therapy.

With a median duration of 3 days of treatment, the incidence rate ratios (IRRs) were 1.80 (95% confidence interval, 1.75-1.84) for GI bleeding, 1.99 (95% CI, 1.70-2.32) for sepsis, and 2.37 (95% CI, 2.13-2.63) for heart failure.

Given the findings, physicians should weigh the benefits against the risks of rare but potentially serious consequences of these anti-inflammatory drugs, according to the authors.

“After initiating patients on oral steroid bursts, physicians should be on the lookout for these severe adverse events, particularly within the first month after initiation of steroid therapy,” Dr. Yao said in an interview.

The findings were published online July 6 in Annals of Internal Medicine.

Of the 15,859,129 adult Asians in the Taiwanese database, the study included 2,623,327 adults aged 20-64 years who received single steroid bursts (14 days or less) between Jan. 1, 2013, and Dec. 31, 2015.

Almost 60% of the indications were for skin disorders, such as eczema and urticaria, and for respiratory tract infections, such as sinusitis and acute pharyngitis. Among specialties, dermatology, otolaryngology, family practice, internal medicine, and pediatrics accounted for 88% of prescriptions.

“Our findings are important for physicians and guideline developers because short-term use of oral corticosteroids is common and the real-world safety of this approach remains unclear,” the authors wrote. They acknowledged that the database did not provide information on such potential confounders as disease severity and lifestyle factors, nor did it include children and vulnerable individuals, which may limit the generalizability of the results.

The findings echo those of a 2017 cohort study conducted by researchers at the University of Michigan in Ann Arbor. That study, by Akbar K. Waljee, MD, assistant professor of gastroenterology, University of Michigan, Ann Arbor, and colleagues, included data on more than 1.5 million privately insured U.S. adults. The researchers included somewhat longer steroid bursts of up to 30 days’ duration and found that use of the drugs was associated with a greater than fivefold increased risk for sepsis, a more than threefold increased risk for venous thromboembolism, and a nearly twofold increased risk for fracture within 30 days of starting treatment.

Furthermore, the elevated risk persisted at prednisone-equivalent doses of less than 20 mg/d (IRR, 4.02 for sepsis, 3.61 for venous thromboembolism, and 1.83 for fracture; all P < .001).

The U.S. study also found that during the 3-year period from 2012 to 2014, more than 20% of patients were prescribed short-term oral corticosteroids.

“Both studies indicate that these short-term regimens are more common in the real world than was previously thought and are not risk free,” Dr. Yao said.

Recognition that corticosteroids are associated with adverse events has been building for decades, according to the authors of an editorial that accompanies the new study.

“However, we commonly use short corticosteroid ‘bursts’ for minor ailments despite a lack of evidence for meaningful benefit. We are now learning that bursts as short as 3 days may increase risk for serious AEs, even in young and healthy people,” wrote editorialists Beth I. Wallace, MD, of the Center for Clinical Management Research at the VA Ann Arbor Healthcare System and the Institute for Healthcare Policy and Innovation at Michigan Medicine, Ann Arbor, and Dr. Waljee, who led the 2017 study.

Dr. Wallace and Dr. Waljee drew parallels between corticosteroid bursts and other short-term regimens, such as of antibiotics and opiates, in which prescriber preference and sometimes patient pressure play a role. “All of these treatments have well-defined indications but can cause net harm when used. We can thus conceive of a corticosteroid stewardship model of targeted interventions that aims to reduce inappropriate prescribing,” they wrote.

In an interview, Dr. Wallace, a rheumatologist who prescribes oral steroids fairly frequently, noted that the Taiwan study is the first to investigate steroid bursts. “Up till now, these very short courses have flown under the radar. Clinicians very commonly prescribe short courses to help relieve symptoms of self-limited conditions like bronchitis, and we assume that because the exposure duration is short, the risks are low, especially for patients who are otherwise healthy.”

She warned that the data in the current study indicate that these short bursts – even at the lower end of the 1- to 2-week courses American physicians prescribe most often – carry small but real increases in risk for serious AEs. “And these increases were seen in young, healthy people, not just in people with preexisting conditions,” she said. “So, we might need to start thinking harder about how we are prescribing even these very short courses of steroids and try to use steroids only when their meaningful benefits really outweigh the risk.”

She noted that a patient with a chronic inflammatory condition such as rheumatoid arthritis may benefit substantially from short-term steroids to treat a disease flare. In that specific case, the benefits of short-term steroids may outweigh the risks, Dr. Wallace said.

But not everyone thinks a new strategy is needed. For Whitney A. High, MD, associate professor of dermatology and pathology at the University of Colorado at Denver, Aurora, the overprescribing of short-term corticosteroids is not a problem, and dermatologists are already exercising caution.

Bruce Jancin/MDedge News

“I only prescribe these drugs short term to, at a guess, about 1 in 40 patients and only when a patient is miserable and quality of life is being seriously affected,” he said in an interview. “And that’s something that can’t be measured in a database study like the one from Taiwan but only in a risk-benefit analysis,” he said.

Furthermore, dermatologists have other drugs and technologies in their armamentarium, including topical steroids with occlusion or with wet wraps, phototherapy, phosphodiesterase inhibitors, calcipotriene, methotrexate and other immunosuppressive agents, and biologics. “In fact, many of these agents are specifically referred to as steroid-sparing,” Dr. High said.

Nor does he experience much pressure from patients to prescribe these drugs. “While occasionally I may encounter a patient who places pressure on me for oral steroids, it’s probably not nearly as frequently as providers in other fields are pressured to prescribe antibiotics or narcotics,” he said.

According to the Taiwanese researchers, the next step is to conduct more studies, including clinical trials, to determine optimal use of corticosteroids by monitoring adverse events. In the meantime, for practitioners such as Dr. Wallace and Dr. High, there is ample evidence from several recent studies of the harms of short-term corticosteroids, whereas the benefits for patients with self-limiting conditions remain uncertain. “This and other studies like it quite appropriately remind providers to avoid oral steroids when they’re not necessary and to seek alternatives where possible,” Dr. High said.

The study was supported by the National Health Research Institutes of Taiwan, the Ministry of Science and Technology of Taiwan, the Chang Gung Medical Foundation, and the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (NIH). Dr. Yao has disclosed no relevant financial relationships. Dr. Wu has received grants from GlaxoSmithKline outside the submitted work. The editorialists and Dr. High have disclosed no relevant financial relationships. Dr. Wallace received an NIH grant during the writing of the editorial.
 

A version of this article originally appeared on Medscape.com.

Extended use of corticosteroids for chronic inflammatory conditions puts patients at risk for serious adverse events (AEs), including cardiovascular disease, osteoporosis, cataracts, and diabetes. Now, a growing body of evidence suggests that even short bursts of these drugs are associated with serious risks.

Most recently, a population-based study of more than 2.6 million people found that taking corticosteroids for 14 days or less was associated with a substantially greater risk for gastrointestinal (GI) bleeding, sepsis, and heart failure, particularly within the first 30 days after therapy.

In the study, Tsung-Chieh Yao, MD, PhD, a professor in the division of allergy, asthma, and rheumatology in the department of pediatrics at Chang Gung Memorial Hospital in Taoyuan, Taiwan, and colleagues used a self-controlled case series to analyze data from Taiwan’s National Health Insurance Research Database of medical claims. They compared patients’ conditions in the period from 5 to 90 days before treatment to conditions from the periods from 5 to 30 days and from 31 to 90 days after therapy.

With a median duration of 3 days of treatment, the incidence rate ratios (IRRs) were 1.80 (95% confidence interval, 1.75-1.84) for GI bleeding, 1.99 (95% CI, 1.70-2.32) for sepsis, and 2.37 (95% CI, 2.13-2.63) for heart failure.

Given the findings, physicians should weigh the benefits against the risks of rare but potentially serious consequences of these anti-inflammatory drugs, according to the authors.

“After initiating patients on oral steroid bursts, physicians should be on the lookout for these severe adverse events, particularly within the first month after initiation of steroid therapy,” Dr. Yao said in an interview.

The findings were published online July 6 in Annals of Internal Medicine.

Of the 15,859,129 adult Asians in the Taiwanese database, the study included 2,623,327 adults aged 20-64 years who received single steroid bursts (14 days or less) between Jan. 1, 2013, and Dec. 31, 2015.

Almost 60% of the indications were for skin disorders, such as eczema and urticaria, and for respiratory tract infections, such as sinusitis and acute pharyngitis. Among specialties, dermatology, otolaryngology, family practice, internal medicine, and pediatrics accounted for 88% of prescriptions.

“Our findings are important for physicians and guideline developers because short-term use of oral corticosteroids is common and the real-world safety of this approach remains unclear,” the authors wrote. They acknowledged that the database did not provide information on such potential confounders as disease severity and lifestyle factors, nor did it include children and vulnerable individuals, which may limit the generalizability of the results.

The findings echo those of a 2017 cohort study conducted by researchers at the University of Michigan in Ann Arbor. That study, by Akbar K. Waljee, MD, assistant professor of gastroenterology, University of Michigan, Ann Arbor, and colleagues, included data on more than 1.5 million privately insured U.S. adults. The researchers included somewhat longer steroid bursts of up to 30 days’ duration and found that use of the drugs was associated with a greater than fivefold increased risk for sepsis, a more than threefold increased risk for venous thromboembolism, and a nearly twofold increased risk for fracture within 30 days of starting treatment.

Furthermore, the elevated risk persisted at prednisone-equivalent doses of less than 20 mg/d (IRR, 4.02 for sepsis, 3.61 for venous thromboembolism, and 1.83 for fracture; all P < .001).

The U.S. study also found that during the 3-year period from 2012 to 2014, more than 20% of patients were prescribed short-term oral corticosteroids.

“Both studies indicate that these short-term regimens are more common in the real world than was previously thought and are not risk free,” Dr. Yao said.

Recognition that corticosteroids are associated with adverse events has been building for decades, according to the authors of an editorial that accompanies the new study.

“However, we commonly use short corticosteroid ‘bursts’ for minor ailments despite a lack of evidence for meaningful benefit. We are now learning that bursts as short as 3 days may increase risk for serious AEs, even in young and healthy people,” wrote editorialists Beth I. Wallace, MD, of the Center for Clinical Management Research at the VA Ann Arbor Healthcare System and the Institute for Healthcare Policy and Innovation at Michigan Medicine, Ann Arbor, and Dr. Waljee, who led the 2017 study.

Dr. Wallace and Dr. Waljee drew parallels between corticosteroid bursts and other short-term regimens, such as of antibiotics and opiates, in which prescriber preference and sometimes patient pressure play a role. “All of these treatments have well-defined indications but can cause net harm when used. We can thus conceive of a corticosteroid stewardship model of targeted interventions that aims to reduce inappropriate prescribing,” they wrote.

In an interview, Dr. Wallace, a rheumatologist who prescribes oral steroids fairly frequently, noted that the Taiwan study is the first to investigate steroid bursts. “Up till now, these very short courses have flown under the radar. Clinicians very commonly prescribe short courses to help relieve symptoms of self-limited conditions like bronchitis, and we assume that because the exposure duration is short, the risks are low, especially for patients who are otherwise healthy.”

She warned that the data in the current study indicate that these short bursts – even at the lower end of the 1- to 2-week courses American physicians prescribe most often – carry small but real increases in risk for serious AEs. “And these increases were seen in young, healthy people, not just in people with preexisting conditions,” she said. “So, we might need to start thinking harder about how we are prescribing even these very short courses of steroids and try to use steroids only when their meaningful benefits really outweigh the risk.”

She noted that a patient with a chronic inflammatory condition such as rheumatoid arthritis may benefit substantially from short-term steroids to treat a disease flare. In that specific case, the benefits of short-term steroids may outweigh the risks, Dr. Wallace said.

But not everyone thinks a new strategy is needed. For Whitney A. High, MD, associate professor of dermatology and pathology at the University of Colorado at Denver, Aurora, the overprescribing of short-term corticosteroids is not a problem, and dermatologists are already exercising caution.

Bruce Jancin/MDedge News

“I only prescribe these drugs short term to, at a guess, about 1 in 40 patients and only when a patient is miserable and quality of life is being seriously affected,” he said in an interview. “And that’s something that can’t be measured in a database study like the one from Taiwan but only in a risk-benefit analysis,” he said.

Furthermore, dermatologists have other drugs and technologies in their armamentarium, including topical steroids with occlusion or with wet wraps, phototherapy, phosphodiesterase inhibitors, calcipotriene, methotrexate and other immunosuppressive agents, and biologics. “In fact, many of these agents are specifically referred to as steroid-sparing,” Dr. High said.

Nor does he experience much pressure from patients to prescribe these drugs. “While occasionally I may encounter a patient who places pressure on me for oral steroids, it’s probably not nearly as frequently as providers in other fields are pressured to prescribe antibiotics or narcotics,” he said.

According to the Taiwanese researchers, the next step is to conduct more studies, including clinical trials, to determine optimal use of corticosteroids by monitoring adverse events. In the meantime, for practitioners such as Dr. Wallace and Dr. High, there is ample evidence from several recent studies of the harms of short-term corticosteroids, whereas the benefits for patients with self-limiting conditions remain uncertain. “This and other studies like it quite appropriately remind providers to avoid oral steroids when they’re not necessary and to seek alternatives where possible,” Dr. High said.

The study was supported by the National Health Research Institutes of Taiwan, the Ministry of Science and Technology of Taiwan, the Chang Gung Medical Foundation, and the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (NIH). Dr. Yao has disclosed no relevant financial relationships. Dr. Wu has received grants from GlaxoSmithKline outside the submitted work. The editorialists and Dr. High have disclosed no relevant financial relationships. Dr. Wallace received an NIH grant during the writing of the editorial.
 

A version of this article originally appeared on Medscape.com.

Extended use of corticosteroids for chronic inflammatory conditions puts patients at risk for serious adverse events (AEs), including cardiovascular disease, osteoporosis, cataracts, and diabetes. Now, a growing body of evidence suggests that even short bursts of these drugs are associated with serious risks.

Most recently, a population-based study of more than 2.6 million people found that taking corticosteroids for 14 days or less was associated with a substantially greater risk for gastrointestinal (GI) bleeding, sepsis, and heart failure, particularly within the first 30 days after therapy.

In the study, Tsung-Chieh Yao, MD, PhD, a professor in the division of allergy, asthma, and rheumatology in the department of pediatrics at Chang Gung Memorial Hospital in Taoyuan, Taiwan, and colleagues used a self-controlled case series to analyze data from Taiwan’s National Health Insurance Research Database of medical claims. They compared patients’ conditions in the period from 5 to 90 days before treatment to conditions from the periods from 5 to 30 days and from 31 to 90 days after therapy.

With a median duration of 3 days of treatment, the incidence rate ratios (IRRs) were 1.80 (95% confidence interval, 1.75-1.84) for GI bleeding, 1.99 (95% CI, 1.70-2.32) for sepsis, and 2.37 (95% CI, 2.13-2.63) for heart failure.

Given the findings, physicians should weigh the benefits against the risks of rare but potentially serious consequences of these anti-inflammatory drugs, according to the authors.

“After initiating patients on oral steroid bursts, physicians should be on the lookout for these severe adverse events, particularly within the first month after initiation of steroid therapy,” Dr. Yao said in an interview.

The findings were published online July 6 in Annals of Internal Medicine.

Of the 15,859,129 adult Asians in the Taiwanese database, the study included 2,623,327 adults aged 20-64 years who received single steroid bursts (14 days or less) between Jan. 1, 2013, and Dec. 31, 2015.

Almost 60% of the indications were for skin disorders, such as eczema and urticaria, and for respiratory tract infections, such as sinusitis and acute pharyngitis. Among specialties, dermatology, otolaryngology, family practice, internal medicine, and pediatrics accounted for 88% of prescriptions.

“Our findings are important for physicians and guideline developers because short-term use of oral corticosteroids is common and the real-world safety of this approach remains unclear,” the authors wrote. They acknowledged that the database did not provide information on such potential confounders as disease severity and lifestyle factors, nor did it include children and vulnerable individuals, which may limit the generalizability of the results.

The findings echo those of a 2017 cohort study conducted by researchers at the University of Michigan in Ann Arbor. That study, by Akbar K. Waljee, MD, assistant professor of gastroenterology, University of Michigan, Ann Arbor, and colleagues, included data on more than 1.5 million privately insured U.S. adults. The researchers included somewhat longer steroid bursts of up to 30 days’ duration and found that use of the drugs was associated with a greater than fivefold increased risk for sepsis, a more than threefold increased risk for venous thromboembolism, and a nearly twofold increased risk for fracture within 30 days of starting treatment.

Furthermore, the elevated risk persisted at prednisone-equivalent doses of less than 20 mg/d (IRR, 4.02 for sepsis, 3.61 for venous thromboembolism, and 1.83 for fracture; all P < .001).

The U.S. study also found that during the 3-year period from 2012 to 2014, more than 20% of patients were prescribed short-term oral corticosteroids.

“Both studies indicate that these short-term regimens are more common in the real world than was previously thought and are not risk free,” Dr. Yao said.

Recognition that corticosteroids are associated with adverse events has been building for decades, according to the authors of an editorial that accompanies the new study.

“However, we commonly use short corticosteroid ‘bursts’ for minor ailments despite a lack of evidence for meaningful benefit. We are now learning that bursts as short as 3 days may increase risk for serious AEs, even in young and healthy people,” wrote editorialists Beth I. Wallace, MD, of the Center for Clinical Management Research at the VA Ann Arbor Healthcare System and the Institute for Healthcare Policy and Innovation at Michigan Medicine, Ann Arbor, and Dr. Waljee, who led the 2017 study.

Dr. Wallace and Dr. Waljee drew parallels between corticosteroid bursts and other short-term regimens, such as of antibiotics and opiates, in which prescriber preference and sometimes patient pressure play a role. “All of these treatments have well-defined indications but can cause net harm when used. We can thus conceive of a corticosteroid stewardship model of targeted interventions that aims to reduce inappropriate prescribing,” they wrote.

In an interview, Dr. Wallace, a rheumatologist who prescribes oral steroids fairly frequently, noted that the Taiwan study is the first to investigate steroid bursts. “Up till now, these very short courses have flown under the radar. Clinicians very commonly prescribe short courses to help relieve symptoms of self-limited conditions like bronchitis, and we assume that because the exposure duration is short, the risks are low, especially for patients who are otherwise healthy.”

She warned that the data in the current study indicate that these short bursts – even at the lower end of the 1- to 2-week courses American physicians prescribe most often – carry small but real increases in risk for serious AEs. “And these increases were seen in young, healthy people, not just in people with preexisting conditions,” she said. “So, we might need to start thinking harder about how we are prescribing even these very short courses of steroids and try to use steroids only when their meaningful benefits really outweigh the risk.”

She noted that a patient with a chronic inflammatory condition such as rheumatoid arthritis may benefit substantially from short-term steroids to treat a disease flare. In that specific case, the benefits of short-term steroids may outweigh the risks, Dr. Wallace said.

But not everyone thinks a new strategy is needed. For Whitney A. High, MD, associate professor of dermatology and pathology at the University of Colorado at Denver, Aurora, the overprescribing of short-term corticosteroids is not a problem, and dermatologists are already exercising caution.

Bruce Jancin/MDedge News

“I only prescribe these drugs short term to, at a guess, about 1 in 40 patients and only when a patient is miserable and quality of life is being seriously affected,” he said in an interview. “And that’s something that can’t be measured in a database study like the one from Taiwan but only in a risk-benefit analysis,” he said.

Furthermore, dermatologists have other drugs and technologies in their armamentarium, including topical steroids with occlusion or with wet wraps, phototherapy, phosphodiesterase inhibitors, calcipotriene, methotrexate and other immunosuppressive agents, and biologics. “In fact, many of these agents are specifically referred to as steroid-sparing,” Dr. High said.

Nor does he experience much pressure from patients to prescribe these drugs. “While occasionally I may encounter a patient who places pressure on me for oral steroids, it’s probably not nearly as frequently as providers in other fields are pressured to prescribe antibiotics or narcotics,” he said.

According to the Taiwanese researchers, the next step is to conduct more studies, including clinical trials, to determine optimal use of corticosteroids by monitoring adverse events. In the meantime, for practitioners such as Dr. Wallace and Dr. High, there is ample evidence from several recent studies of the harms of short-term corticosteroids, whereas the benefits for patients with self-limiting conditions remain uncertain. “This and other studies like it quite appropriately remind providers to avoid oral steroids when they’re not necessary and to seek alternatives where possible,” Dr. High said.

The study was supported by the National Health Research Institutes of Taiwan, the Ministry of Science and Technology of Taiwan, the Chang Gung Medical Foundation, and the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (NIH). Dr. Yao has disclosed no relevant financial relationships. Dr. Wu has received grants from GlaxoSmithKline outside the submitted work. The editorialists and Dr. High have disclosed no relevant financial relationships. Dr. Wallace received an NIH grant during the writing of the editorial.
 

A version of this article originally appeared on Medscape.com.

Chronic pain is common in veterans, and early engagement in pain treatment is recommended to forestall consequences of untreated pain, including depression, disability, and substance use disorders. The Veterans Health Administration (VHA) employs a stepped care model of pain treatment, with the majority of pain care based in primary care (step 1), and an array of specialty/multimodal treatment options made available at each step in the model for patients with more complex problems, or those who do not respond to more conservative interventions.¹

Recognizing the need for comprehensive pain care, the US Congress passed the Comprehensive Addiction and Recovery Act, 21 USC §1521 (2016), which included provisions for VHA facilities to offer multimodal pain treatment and to report the availability of pain care options at each step in the stepped care model.^2, With the passage of the Veterans Access, Choice, and Accountability Act of 2014, 38 USC §101 (2014) and now the MISSION Act of 2018, 38 USC §703 (2018) veterans whose VHA facilities are too distant, who require care unavailable at that facility, or who have to wait too long to receive care are eligible for treatment at either VHA or non-VHA facilities.³ These laws allocate the same pool of funds to both VHA and community care and thus create an incentive to engage veterans in care within the VHA network so the funds are not spent out of network.⁴

An opportunity to connect veterans with VHA care arises at specialized VHA Compensation and Pension (C&P) clinics during examinations that determine whether a veteran’s health conditions were caused or exacerbated by their military service. Veterans file claims with the US Department of Veterans Affairs (VA) Veterans Benefits Administration (VBA), which sends the patient to either a VHA facility or private practitioners for these examinations. Although the number of examinations conducted each year is not available, there were 274,528 veterans newly awarded compensation in fiscal year 2018, and a substantial number of the total of 4,743,108 veterans with C&P awards had reevaluation examinations for at least 1 of their conditions during that year.⁵ Based largely on the compensation examination results, military service records, and medical records, veterans are granted a service-connected rating for conditions deemed related to military service. A service-connection rating between 0% and 100% is assigned by the VBA, with higher ratings indicating more impairment and, consequently, more financial compensation. Service-connection ratings also are used to decide which veterans are in the highest priority groups for receipt of VHA health care services and are exempt from copayments.

Although traditionally thought of as a forensic evaluation with no clinical purpose, the C&P examination process affords many opportunities to explain VHA care to veterans in distress who file claims.⁶ A randomized clinical trials (RCT) involving veterans with mental health claims and a second RCT including veterans with musculoskeletal claims each found that veterans use more VHA services if offered outreach at the time of the C&P examination.^7,8 In addition to clinical benefits, outreach around the time of C&P examinations also might mitigate the well documented adversarial aspects of the service-connection claims process.^6,9,10 Currently, such outreach is not part of routine VHA procedures. Ironically, it is the VBA and not VHA that contacts veterans who are awarded service-connection with information about their eligibility for VHA care based on their award.

Connecting veterans to pain treatment can involve clarifying eligibility for VHA care for veterans in whom eligibility is unknown, involving primary care providers (PCPs) who are the fulcrum of VHA pain care referrals, and motivating veterans to seek specific pain treatment modalities. Connecting veterans to treatment at the time of their compensation examinations also likely involves bidirectional cooperation between the specialized C&P clinics where veterans are examined and the clinics that provide treatment.

Relational coordination is a theoretical framework that can describe the horizontal relationships between different teams within the same medical facility. Relational coordination theorizes that communication between workgroups is related synergistically to the quality of relationships between workgroups. Relational coordination is better between workgroups that share goals and often have high levels of relational coordination, which is thought to be especially important when activities are ambiguous, require cooperation, and are conducted under time pressure.¹¹ High relational coordination also has been associated with high staff job satisfaction, high satisfaction with delivered services, and adherence to treatment guidelines.^12-14 An observational cohort study suggested that relational coordination can be improved by targeted interventions that bring workgroups together and facilitate intercommunication.¹⁵

To better understand referral and engagement for pain treatment at compensation examinations, VA staff from primary care, mental health, pain management, and C&P teams at the 8 VHA medical centers in New England were invited to complete a validated relational coordination survey.^11,16 A subset of invited staff participated in a semistructured interview about pain treatment referral practices within their medical centers.

Methods

Assessments were conducted as part of a mixed methods formative evaluation involving quantitative and qualitative methods for a clinical trial at the 8 VHA medical centers in New England. The trial is testing an intervention in which veterans presenting for service-connection examinations for musculoskeletal conditions receive brief counseling to engage them in nonopioid pain treatments. The VHA Central Institutional Review Board approved this formative evaluation and the clinical trial has begun (ClinicalTrials.gov NCT04062214).

Potential interviewees were involved in referrals to and provision of nonpharmacologic pain treatment and were identified by site investigators in the randomized trial. Identified interviewees were clinical and administrative staff belonging to VHA Primary Care, Pain Management, and Compensation and Pension clinics. A total of 83 staff were identified.

Semistructured Interviews

A subset of the 83 staff were invited to participate in a semistructured interview because their position impacted coordination of pain care at their facilities or they worked in C&P. Staff at a site were interviewed until no new themes emerged from additional interviews, and each of the 8 sites was represented. Interviews were conducted between June and August 2018. Standardized scripts describing the study and inviting participation in a semistructured interview were e-mailed to VA staff. At the time of the interview the study purpose was restated and consent for audiotaping was obtained. The interviews followed a guide designed to assess a relational coordination framework among various workgroups. The data in this manuscript were elicited by specific prompts concerning: (1) How veterans learn about pain care when they come through C&P; and (2) How staff in C&P communicate with treatment providers about veterans who have chronic pain. Each interview lasted about 30 minutes.

Relational Coordination Survey

All identified staff were invited to participate in a relational coordination survey. The survey was administered through VA REDCap. Survey invitations were e-mailed from REDCap to VA staff and included a description of the study and assurances of the confidentiality of data collected. Surveys took < 10 minutes to complete. To begin, respondents identified their primary workgroup (C&P, primary care, pain management, or administrative leadership or staff), secondary workgroup (if they were in > 1), and site. Respondents provided no other identifying information and were assured their responses would be confidential.

The survey consisted of 7 questions regarding beliefs about the quality of communication and interactions among workgroup members in obtaining a shared goal.¹¹ The shared goal in the survey used in this study was providing pain care services for veterans with musculoskeletal conditions. Using a 5-point Likert scale, the 7 questions concerned frequency, timeliness, and accuracy of communication; response to problems providing pain services; sharing goals; and knowledge and respect for respondent’s job function. Higher scores indicated better relational coordination among members of a workgroup. Using the survey’s 7 items, composite mean relational coordination scores were calculated for each of the 4 primary workgroups. To account for the possibility that a member rated their own workgroups, 2 scores were created for each workgroup; one included members of the workgroup and another excluded them.

Data Analysis

The audio-recorded semistructured interviews were transcribed and entered into Atlas.ti qualitative data analysis software. To identify cross-cutting themes, a semistructured telephone interview guide was developed by the qualitative study team that emphasized interrelationships between different clinical teams. The transcripts were then analyzed using the grounded theory approach, a systematic methodology to reduce themes from collected qualitative data. Two research staff read each transcript twice; first to familiarize themselves with the text and then, using open coding, to identify important concepts that emerged from the language and assign codes to segments of text. To ensure accuracy, researchers included suitable contextual information in the coding. Using the constant comparative method, research staff then met to examine the themes that emerged in the interviews, discuss and coalesce coding discrepancies, and compare perspectives.¹⁷

The composite score (mean of the 7 items and 95% CI) of the survey responses was analyzed to identify significant differences in coordination across the 4 workgroups. Analysis of variance (ANOVA) was used to examine each relational coordination score by respondents’ workgroup. Post hoc analyses examined relational coordination survey differences among the 4 respondent groups.

Results

Thirty-nine survey respondents participated in the semistructured interviews. C&P examiners expressed varying degrees of comfort with their role in extending access to pain care for veterans. Some of the examiners strongly believed that their role was purely forensic, and going beyond this forensic role to refer or recommend treatment to veterans would be a violation of their role to conduct a forensic examination. “We don’t have an ongoing therapeutic relationship with any of the patients,” a C&P examiner explained: “We see them once; they’re out the door. It’s forensic. We’re investigating the person as a claimant, we’re investigating it and using our tools to go and review information from 30, 40 years ago.”

Other examiners had a less strict approach for working with veterans in C&P, even though examiners are asked not to provide advice or therapy. One C&P examiner noted that because he “can’t watch people in pain,” during the examination this doctor recommends that patients go to the office that determines whether they are eligible for benefits and choose a PCP. Another C&P examiner concurred with this approach. “I certainly spend a little time with the veteran talking to them about their personal life, who they are, what they do, what they’ve done, what they’re going to do to kind of break the ice between us,” the second examiner explained. “At the end, I will make some suggestions to them. I’m comfortable doing that. I don’t know that everybody is.”

Many of the VHA providers we interviewed had little knowledge of the C&P process or whether C&P examiners had any role or responsibilities in referring veterans for pain care. Most VHA providers could not name any C&P examiners at their facility and were generally unfamiliar with the content of C&P examinations. One provider bluntly said, “I’ve never communicated with anyone in comp and pen [C&P].”

Another PCP also expressed concerns with referrals, suggesting that C&P and primary care “are totally separate and should remain separate,” the PCP explained. “My concern with getting referral from comp and pen is that is it then they’re seeking all sorts of treatment that they wouldn’t necessarily need or ask for otherwise.”

Conversely a different PCP had a positive outlook on how C&P examiners might help ease the transition into the VHA for veterans with pain, especially for newly discharged veterans. “Having comp and pen address these issues is really going to be helpful. I think it could be significant that the topic is introduced early on.”

Relational Coordination Survey

Relational coordination surveys were sent to 83 participants of whom 66 responded. Respondents were from C&P (n = 7), primary care (n = 16), pain medicine (n = 32), and administration (n = 11). Of the 66 respondents, 18 indicated a secondary workgroup. Respondents on 2 teams (primary/secondary) were primary care/administrative (n = 4), pain management/primary care (n = 4), primary care/pain management (n = 3), administrative/primary care (n = 3), and C&P/administrative (n = 1).

The relational coordination composite scores were lowest for C&P. This finding remained whether C&P staff surveys were included or removed from the C&P responses. As demonstrated by the 95% CI, when team members’ surveys were included, C&P scores (95% CI, 2.01-2.42) were significantly lower than the primary care (95% CI, 3.34-3.64) and pain management (95% CI, 3.61-3.96) groups. All the relational coordination composite scores were slightly lower when staff who described their own workgroup were removed (ie, respondents rated their own workgroups as having higher relational coordination than others did). Using the composite scores excluding same workgroup members, the composite scores of the C&P remained significantly lower than all 3 other workgroups (Table). Means values for each individual item in the C&P group were significantly less than all other group means for each item except for the question on responses to problems providing pain services (data not shown). On this item only, the mean C&P rating was > 3 (3.19), but this was still lower than the means of the primary care and pain management workgroups.

Further analyses were undertaken to understand the importance of stakeholders’ ratings of their own workgroup compared with ratings by others of that workgroup. A 1-way ANOVA of workgroup was conducted and displayed significant workgroup differences between member and nonmember relational coordination ratings on 3 of the 4 workgroup’s scores C&P (F = 5.75, 3, 62 df; P < .01) primary care (F = 4.30, 3, 62 df; P < .008) and pain management (F = 8.22, 3, 62 df; P < .001). Post hoc contrasts between the different workgroups doing the rating revealed: (1) significant differences in the assessment of the C&P workgroup between the C&P workgroup and both the primary care (P < .01) and pain management groups (P < .001) with C&P rating their own workgroup significantly higher; (2) a significant difference in the scoring of the primary care workgroup with the primary care group rating themselves significantly higher than the C&P group; and (3) significant differences in the scoring of the pain management workgroup with both pain management and primary care groups rating the pain management group significantly higher than the C&P group. The results were not substantially changed by removing the 18 respondents who identified themselves as being part of > 1 workgroup .

Discussion

Mixed methods revealed disparate viewpoints about the role of C&P in referring veterans to pain care services. Overall, C&P teams coordinated less with other workgroups than the other groups coordinated with each other, and the C&P clinics took only limited steps to engage veterans in VHA treatment. The relational coordination results appeared to be valid. The mean scores were near the middle of the relational coordination rating scale, with standard deviations indicating a range of responses. The lower relational coordination scores of the C&P group remained after removing stakeholders who were rating their own workgroup. Further support for the validity of the relational coordination survey results is that they were consistent with the reports of C&P clinic isolation in the semistructured interviews.

The interview data suggest that one reason the C&P teams had low relational coordination scores is that VA staff interpret the emphasis on evaluative rather than therapeutic examinations to preclude other attempts to engage veterans into VHA treatment, even though such treatment engagement is permitted within existing guidelines. VBA referrals for examinations say nothing, either way, about engaging veterans in VHA care. The relational coordination results suggest that an intervention that might increase treatment referrals from the C&P clinics would be to explain the (existing) policy allowing for outreach around the time of compensation examinations to VHA staff so this goal is clearly agreed-upon. Another approach to facilitating treatment engagement at the C&P examination is to use other interventions that have been associated with better relational coordination such as intergroup meetings, horizontal integration more generally, and an atmosphere is which people from different backgrounds feel empowered to speak frankly to each other.^15,18,19 An important linkage to forge is between C&P teams and the administrative workgroups responsible for verifying a veteran’s eligibility for VHA care and enrolling eligible veterans in VHA treatment. Having C&P clinicians who are familiar with the eligibility and treatment engagement processes would facilitate providing that information to veterans, without compromising the evaluative format of the compensation examination.

An interesting ancillary finding is that relational coordination ratings by members of 3 of the 4 workgroups were higher than ratings by other staff of that workgroup. A possible explanation for this finding is that workgroup members are more aware of the relational coordination efforts made by their own workgroup than those by other workgroups, and therefore rate their own workgroup higher. This also might be part of a broader self-aggrandizement heuristic that has been described in multiple domains.²⁰ Staff may apply this heuristic in reporting that their staff engage in more relational coordination, reflecting the social desirability of being cooperative.

There are simple facility-level interventions that would facilitate veterans access to care such as conducting C&P examinations for potentially treatment-eligible veterans at VHA facilities (vs conducted outside VHA) and having access to materials that explain the treatment options to veterans when they check in for their compensation examinations. The approach to C&P-based treatment engagement that was successfully employed in 2 clinical trials involved having counselors not connected with the C&P clinic contact veterans around the time of their compensation examination to explain VA treatment options and motivate veterans to pursue treatment.^8,9 This independent counselor approach is being evaluated in a larger study.

Limitations

These data are from a small number of VA staff evaluating veterans in a single region of the US. They do not show causation, and it is possible that relational coordination is not necessary for referrals from C&P clinics. Relational coordination might not be necessary when referral processes can be simply routinized with little need for communication.¹¹ However, other analyses in these clinics have found that pain treatment referrals in fact are not routinized, with substantial variability within and across institutions. Another possibility is that features that have been associated with less relational coordination, such as male gender and medical specialist guild, were disproportionately present in C&P clinics compared to the other clinics.²¹Finally, veterans may be eligible for priority VA care for reasons that do not involve service-connection claims (38 CFR § 17.37).

Conclusions

There have been public calls to improve the evaluation of service-connection claims such that this process includes approaches to engage veterans in treatment.²² Referring veterans to treatment when they come for C&P examinations will likely involve improving relational coordination between the C&P service and other parts of VHA. Nationwide, sites that integrate C&P more fully may have valuable lessons to impart about the benefits of such integration. An important step towards better relational coordination will be clarifying that engaging veterans in VHA care around the time of their C&P examinations is a facility-wide goal.

Acknowledgments

The authors thank Brian Linde and Efia James for their perspectives on C&P procedures. This work was supported by the Veterans Integrated Service Network 1 Mental Illness Research Education and Clinical Center (MIRECC) and National Institute of Health, National Center for Complementary and Integrative Health Project # 5UG3AT009758-02. (MIR, SM mPIs).

Chronic pain is common in veterans, and early engagement in pain treatment is recommended to forestall consequences of untreated pain, including depression, disability, and substance use disorders. The Veterans Health Administration (VHA) employs a stepped care model of pain treatment, with the majority of pain care based in primary care (step 1), and an array of specialty/multimodal treatment options made available at each step in the model for patients with more complex problems, or those who do not respond to more conservative interventions.¹

Recognizing the need for comprehensive pain care, the US Congress passed the Comprehensive Addiction and Recovery Act, 21 USC §1521 (2016), which included provisions for VHA facilities to offer multimodal pain treatment and to report the availability of pain care options at each step in the stepped care model.^2, With the passage of the Veterans Access, Choice, and Accountability Act of 2014, 38 USC §101 (2014) and now the MISSION Act of 2018, 38 USC §703 (2018) veterans whose VHA facilities are too distant, who require care unavailable at that facility, or who have to wait too long to receive care are eligible for treatment at either VHA or non-VHA facilities.³ These laws allocate the same pool of funds to both VHA and community care and thus create an incentive to engage veterans in care within the VHA network so the funds are not spent out of network.⁴

An opportunity to connect veterans with VHA care arises at specialized VHA Compensation and Pension (C&P) clinics during examinations that determine whether a veteran’s health conditions were caused or exacerbated by their military service. Veterans file claims with the US Department of Veterans Affairs (VA) Veterans Benefits Administration (VBA), which sends the patient to either a VHA facility or private practitioners for these examinations. Although the number of examinations conducted each year is not available, there were 274,528 veterans newly awarded compensation in fiscal year 2018, and a substantial number of the total of 4,743,108 veterans with C&P awards had reevaluation examinations for at least 1 of their conditions during that year.⁵ Based largely on the compensation examination results, military service records, and medical records, veterans are granted a service-connected rating for conditions deemed related to military service. A service-connection rating between 0% and 100% is assigned by the VBA, with higher ratings indicating more impairment and, consequently, more financial compensation. Service-connection ratings also are used to decide which veterans are in the highest priority groups for receipt of VHA health care services and are exempt from copayments.

Although traditionally thought of as a forensic evaluation with no clinical purpose, the C&P examination process affords many opportunities to explain VHA care to veterans in distress who file claims.⁶ A randomized clinical trials (RCT) involving veterans with mental health claims and a second RCT including veterans with musculoskeletal claims each found that veterans use more VHA services if offered outreach at the time of the C&P examination.^7,8 In addition to clinical benefits, outreach around the time of C&P examinations also might mitigate the well documented adversarial aspects of the service-connection claims process.^6,9,10 Currently, such outreach is not part of routine VHA procedures. Ironically, it is the VBA and not VHA that contacts veterans who are awarded service-connection with information about their eligibility for VHA care based on their award.

Connecting veterans to pain treatment can involve clarifying eligibility for VHA care for veterans in whom eligibility is unknown, involving primary care providers (PCPs) who are the fulcrum of VHA pain care referrals, and motivating veterans to seek specific pain treatment modalities. Connecting veterans to treatment at the time of their compensation examinations also likely involves bidirectional cooperation between the specialized C&P clinics where veterans are examined and the clinics that provide treatment.

Relational coordination is a theoretical framework that can describe the horizontal relationships between different teams within the same medical facility. Relational coordination theorizes that communication between workgroups is related synergistically to the quality of relationships between workgroups. Relational coordination is better between workgroups that share goals and often have high levels of relational coordination, which is thought to be especially important when activities are ambiguous, require cooperation, and are conducted under time pressure.¹¹ High relational coordination also has been associated with high staff job satisfaction, high satisfaction with delivered services, and adherence to treatment guidelines.^12-14 An observational cohort study suggested that relational coordination can be improved by targeted interventions that bring workgroups together and facilitate intercommunication.¹⁵

To better understand referral and engagement for pain treatment at compensation examinations, VA staff from primary care, mental health, pain management, and C&P teams at the 8 VHA medical centers in New England were invited to complete a validated relational coordination survey.^11,16 A subset of invited staff participated in a semistructured interview about pain treatment referral practices within their medical centers.

Methods

Assessments were conducted as part of a mixed methods formative evaluation involving quantitative and qualitative methods for a clinical trial at the 8 VHA medical centers in New England. The trial is testing an intervention in which veterans presenting for service-connection examinations for musculoskeletal conditions receive brief counseling to engage them in nonopioid pain treatments. The VHA Central Institutional Review Board approved this formative evaluation and the clinical trial has begun (ClinicalTrials.gov NCT04062214).

Potential interviewees were involved in referrals to and provision of nonpharmacologic pain treatment and were identified by site investigators in the randomized trial. Identified interviewees were clinical and administrative staff belonging to VHA Primary Care, Pain Management, and Compensation and Pension clinics. A total of 83 staff were identified.

Semistructured Interviews

A subset of the 83 staff were invited to participate in a semistructured interview because their position impacted coordination of pain care at their facilities or they worked in C&P. Staff at a site were interviewed until no new themes emerged from additional interviews, and each of the 8 sites was represented. Interviews were conducted between June and August 2018. Standardized scripts describing the study and inviting participation in a semistructured interview were e-mailed to VA staff. At the time of the interview the study purpose was restated and consent for audiotaping was obtained. The interviews followed a guide designed to assess a relational coordination framework among various workgroups. The data in this manuscript were elicited by specific prompts concerning: (1) How veterans learn about pain care when they come through C&P; and (2) How staff in C&P communicate with treatment providers about veterans who have chronic pain. Each interview lasted about 30 minutes.

Relational Coordination Survey

All identified staff were invited to participate in a relational coordination survey. The survey was administered through VA REDCap. Survey invitations were e-mailed from REDCap to VA staff and included a description of the study and assurances of the confidentiality of data collected. Surveys took < 10 minutes to complete. To begin, respondents identified their primary workgroup (C&P, primary care, pain management, or administrative leadership or staff), secondary workgroup (if they were in > 1), and site. Respondents provided no other identifying information and were assured their responses would be confidential.

The survey consisted of 7 questions regarding beliefs about the quality of communication and interactions among workgroup members in obtaining a shared goal.¹¹ The shared goal in the survey used in this study was providing pain care services for veterans with musculoskeletal conditions. Using a 5-point Likert scale, the 7 questions concerned frequency, timeliness, and accuracy of communication; response to problems providing pain services; sharing goals; and knowledge and respect for respondent’s job function. Higher scores indicated better relational coordination among members of a workgroup. Using the survey’s 7 items, composite mean relational coordination scores were calculated for each of the 4 primary workgroups. To account for the possibility that a member rated their own workgroups, 2 scores were created for each workgroup; one included members of the workgroup and another excluded them.

Data Analysis

The audio-recorded semistructured interviews were transcribed and entered into Atlas.ti qualitative data analysis software. To identify cross-cutting themes, a semistructured telephone interview guide was developed by the qualitative study team that emphasized interrelationships between different clinical teams. The transcripts were then analyzed using the grounded theory approach, a systematic methodology to reduce themes from collected qualitative data. Two research staff read each transcript twice; first to familiarize themselves with the text and then, using open coding, to identify important concepts that emerged from the language and assign codes to segments of text. To ensure accuracy, researchers included suitable contextual information in the coding. Using the constant comparative method, research staff then met to examine the themes that emerged in the interviews, discuss and coalesce coding discrepancies, and compare perspectives.¹⁷

The composite score (mean of the 7 items and 95% CI) of the survey responses was analyzed to identify significant differences in coordination across the 4 workgroups. Analysis of variance (ANOVA) was used to examine each relational coordination score by respondents’ workgroup. Post hoc analyses examined relational coordination survey differences among the 4 respondent groups.

Results

Thirty-nine survey respondents participated in the semistructured interviews. C&P examiners expressed varying degrees of comfort with their role in extending access to pain care for veterans. Some of the examiners strongly believed that their role was purely forensic, and going beyond this forensic role to refer or recommend treatment to veterans would be a violation of their role to conduct a forensic examination. “We don’t have an ongoing therapeutic relationship with any of the patients,” a C&P examiner explained: “We see them once; they’re out the door. It’s forensic. We’re investigating the person as a claimant, we’re investigating it and using our tools to go and review information from 30, 40 years ago.”

Other examiners had a less strict approach for working with veterans in C&P, even though examiners are asked not to provide advice or therapy. One C&P examiner noted that because he “can’t watch people in pain,” during the examination this doctor recommends that patients go to the office that determines whether they are eligible for benefits and choose a PCP. Another C&P examiner concurred with this approach. “I certainly spend a little time with the veteran talking to them about their personal life, who they are, what they do, what they’ve done, what they’re going to do to kind of break the ice between us,” the second examiner explained. “At the end, I will make some suggestions to them. I’m comfortable doing that. I don’t know that everybody is.”

Many of the VHA providers we interviewed had little knowledge of the C&P process or whether C&P examiners had any role or responsibilities in referring veterans for pain care. Most VHA providers could not name any C&P examiners at their facility and were generally unfamiliar with the content of C&P examinations. One provider bluntly said, “I’ve never communicated with anyone in comp and pen [C&P].”

Another PCP also expressed concerns with referrals, suggesting that C&P and primary care “are totally separate and should remain separate,” the PCP explained. “My concern with getting referral from comp and pen is that is it then they’re seeking all sorts of treatment that they wouldn’t necessarily need or ask for otherwise.”

Conversely a different PCP had a positive outlook on how C&P examiners might help ease the transition into the VHA for veterans with pain, especially for newly discharged veterans. “Having comp and pen address these issues is really going to be helpful. I think it could be significant that the topic is introduced early on.”

Relational Coordination Survey

Relational coordination surveys were sent to 83 participants of whom 66 responded. Respondents were from C&P (n = 7), primary care (n = 16), pain medicine (n = 32), and administration (n = 11). Of the 66 respondents, 18 indicated a secondary workgroup. Respondents on 2 teams (primary/secondary) were primary care/administrative (n = 4), pain management/primary care (n = 4), primary care/pain management (n = 3), administrative/primary care (n = 3), and C&P/administrative (n = 1).

The relational coordination composite scores were lowest for C&P. This finding remained whether C&P staff surveys were included or removed from the C&P responses. As demonstrated by the 95% CI, when team members’ surveys were included, C&P scores (95% CI, 2.01-2.42) were significantly lower than the primary care (95% CI, 3.34-3.64) and pain management (95% CI, 3.61-3.96) groups. All the relational coordination composite scores were slightly lower when staff who described their own workgroup were removed (ie, respondents rated their own workgroups as having higher relational coordination than others did). Using the composite scores excluding same workgroup members, the composite scores of the C&P remained significantly lower than all 3 other workgroups (Table). Means values for each individual item in the C&P group were significantly less than all other group means for each item except for the question on responses to problems providing pain services (data not shown). On this item only, the mean C&P rating was > 3 (3.19), but this was still lower than the means of the primary care and pain management workgroups.

Further analyses were undertaken to understand the importance of stakeholders’ ratings of their own workgroup compared with ratings by others of that workgroup. A 1-way ANOVA of workgroup was conducted and displayed significant workgroup differences between member and nonmember relational coordination ratings on 3 of the 4 workgroup’s scores C&P (F = 5.75, 3, 62 df; P < .01) primary care (F = 4.30, 3, 62 df; P < .008) and pain management (F = 8.22, 3, 62 df; P < .001). Post hoc contrasts between the different workgroups doing the rating revealed: (1) significant differences in the assessment of the C&P workgroup between the C&P workgroup and both the primary care (P < .01) and pain management groups (P < .001) with C&P rating their own workgroup significantly higher; (2) a significant difference in the scoring of the primary care workgroup with the primary care group rating themselves significantly higher than the C&P group; and (3) significant differences in the scoring of the pain management workgroup with both pain management and primary care groups rating the pain management group significantly higher than the C&P group. The results were not substantially changed by removing the 18 respondents who identified themselves as being part of > 1 workgroup .

Discussion

Mixed methods revealed disparate viewpoints about the role of C&P in referring veterans to pain care services. Overall, C&P teams coordinated less with other workgroups than the other groups coordinated with each other, and the C&P clinics took only limited steps to engage veterans in VHA treatment. The relational coordination results appeared to be valid. The mean scores were near the middle of the relational coordination rating scale, with standard deviations indicating a range of responses. The lower relational coordination scores of the C&P group remained after removing stakeholders who were rating their own workgroup. Further support for the validity of the relational coordination survey results is that they were consistent with the reports of C&P clinic isolation in the semistructured interviews.

The interview data suggest that one reason the C&P teams had low relational coordination scores is that VA staff interpret the emphasis on evaluative rather than therapeutic examinations to preclude other attempts to engage veterans into VHA treatment, even though such treatment engagement is permitted within existing guidelines. VBA referrals for examinations say nothing, either way, about engaging veterans in VHA care. The relational coordination results suggest that an intervention that might increase treatment referrals from the C&P clinics would be to explain the (existing) policy allowing for outreach around the time of compensation examinations to VHA staff so this goal is clearly agreed-upon. Another approach to facilitating treatment engagement at the C&P examination is to use other interventions that have been associated with better relational coordination such as intergroup meetings, horizontal integration more generally, and an atmosphere is which people from different backgrounds feel empowered to speak frankly to each other.^15,18,19 An important linkage to forge is between C&P teams and the administrative workgroups responsible for verifying a veteran’s eligibility for VHA care and enrolling eligible veterans in VHA treatment. Having C&P clinicians who are familiar with the eligibility and treatment engagement processes would facilitate providing that information to veterans, without compromising the evaluative format of the compensation examination.

An interesting ancillary finding is that relational coordination ratings by members of 3 of the 4 workgroups were higher than ratings by other staff of that workgroup. A possible explanation for this finding is that workgroup members are more aware of the relational coordination efforts made by their own workgroup than those by other workgroups, and therefore rate their own workgroup higher. This also might be part of a broader self-aggrandizement heuristic that has been described in multiple domains.²⁰ Staff may apply this heuristic in reporting that their staff engage in more relational coordination, reflecting the social desirability of being cooperative.

There are simple facility-level interventions that would facilitate veterans access to care such as conducting C&P examinations for potentially treatment-eligible veterans at VHA facilities (vs conducted outside VHA) and having access to materials that explain the treatment options to veterans when they check in for their compensation examinations. The approach to C&P-based treatment engagement that was successfully employed in 2 clinical trials involved having counselors not connected with the C&P clinic contact veterans around the time of their compensation examination to explain VA treatment options and motivate veterans to pursue treatment.^8,9 This independent counselor approach is being evaluated in a larger study.

Limitations

These data are from a small number of VA staff evaluating veterans in a single region of the US. They do not show causation, and it is possible that relational coordination is not necessary for referrals from C&P clinics. Relational coordination might not be necessary when referral processes can be simply routinized with little need for communication.¹¹ However, other analyses in these clinics have found that pain treatment referrals in fact are not routinized, with substantial variability within and across institutions. Another possibility is that features that have been associated with less relational coordination, such as male gender and medical specialist guild, were disproportionately present in C&P clinics compared to the other clinics.²¹Finally, veterans may be eligible for priority VA care for reasons that do not involve service-connection claims (38 CFR § 17.37).

Conclusions

There have been public calls to improve the evaluation of service-connection claims such that this process includes approaches to engage veterans in treatment.²² Referring veterans to treatment when they come for C&P examinations will likely involve improving relational coordination between the C&P service and other parts of VHA. Nationwide, sites that integrate C&P more fully may have valuable lessons to impart about the benefits of such integration. An important step towards better relational coordination will be clarifying that engaging veterans in VHA care around the time of their C&P examinations is a facility-wide goal.

Acknowledgments

The authors thank Brian Linde and Efia James for their perspectives on C&P procedures. This work was supported by the Veterans Integrated Service Network 1 Mental Illness Research Education and Clinical Center (MIRECC) and National Institute of Health, National Center for Complementary and Integrative Health Project # 5UG3AT009758-02. (MIR, SM mPIs).

Chronic pain is common in veterans, and early engagement in pain treatment is recommended to forestall consequences of untreated pain, including depression, disability, and substance use disorders. The Veterans Health Administration (VHA) employs a stepped care model of pain treatment, with the majority of pain care based in primary care (step 1), and an array of specialty/multimodal treatment options made available at each step in the model for patients with more complex problems, or those who do not respond to more conservative interventions.¹

Recognizing the need for comprehensive pain care, the US Congress passed the Comprehensive Addiction and Recovery Act, 21 USC §1521 (2016), which included provisions for VHA facilities to offer multimodal pain treatment and to report the availability of pain care options at each step in the stepped care model.^2, With the passage of the Veterans Access, Choice, and Accountability Act of 2014, 38 USC §101 (2014) and now the MISSION Act of 2018, 38 USC §703 (2018) veterans whose VHA facilities are too distant, who require care unavailable at that facility, or who have to wait too long to receive care are eligible for treatment at either VHA or non-VHA facilities.³ These laws allocate the same pool of funds to both VHA and community care and thus create an incentive to engage veterans in care within the VHA network so the funds are not spent out of network.⁴

An opportunity to connect veterans with VHA care arises at specialized VHA Compensation and Pension (C&P) clinics during examinations that determine whether a veteran’s health conditions were caused or exacerbated by their military service. Veterans file claims with the US Department of Veterans Affairs (VA) Veterans Benefits Administration (VBA), which sends the patient to either a VHA facility or private practitioners for these examinations. Although the number of examinations conducted each year is not available, there were 274,528 veterans newly awarded compensation in fiscal year 2018, and a substantial number of the total of 4,743,108 veterans with C&P awards had reevaluation examinations for at least 1 of their conditions during that year.⁵ Based largely on the compensation examination results, military service records, and medical records, veterans are granted a service-connected rating for conditions deemed related to military service. A service-connection rating between 0% and 100% is assigned by the VBA, with higher ratings indicating more impairment and, consequently, more financial compensation. Service-connection ratings also are used to decide which veterans are in the highest priority groups for receipt of VHA health care services and are exempt from copayments.

Although traditionally thought of as a forensic evaluation with no clinical purpose, the C&P examination process affords many opportunities to explain VHA care to veterans in distress who file claims.⁶ A randomized clinical trials (RCT) involving veterans with mental health claims and a second RCT including veterans with musculoskeletal claims each found that veterans use more VHA services if offered outreach at the time of the C&P examination.^7,8 In addition to clinical benefits, outreach around the time of C&P examinations also might mitigate the well documented adversarial aspects of the service-connection claims process.^6,9,10 Currently, such outreach is not part of routine VHA procedures. Ironically, it is the VBA and not VHA that contacts veterans who are awarded service-connection with information about their eligibility for VHA care based on their award.

Connecting veterans to pain treatment can involve clarifying eligibility for VHA care for veterans in whom eligibility is unknown, involving primary care providers (PCPs) who are the fulcrum of VHA pain care referrals, and motivating veterans to seek specific pain treatment modalities. Connecting veterans to treatment at the time of their compensation examinations also likely involves bidirectional cooperation between the specialized C&P clinics where veterans are examined and the clinics that provide treatment.

Relational coordination is a theoretical framework that can describe the horizontal relationships between different teams within the same medical facility. Relational coordination theorizes that communication between workgroups is related synergistically to the quality of relationships between workgroups. Relational coordination is better between workgroups that share goals and often have high levels of relational coordination, which is thought to be especially important when activities are ambiguous, require cooperation, and are conducted under time pressure.¹¹ High relational coordination also has been associated with high staff job satisfaction, high satisfaction with delivered services, and adherence to treatment guidelines.^12-14 An observational cohort study suggested that relational coordination can be improved by targeted interventions that bring workgroups together and facilitate intercommunication.¹⁵

To better understand referral and engagement for pain treatment at compensation examinations, VA staff from primary care, mental health, pain management, and C&P teams at the 8 VHA medical centers in New England were invited to complete a validated relational coordination survey.^11,16 A subset of invited staff participated in a semistructured interview about pain treatment referral practices within their medical centers.

Methods

Assessments were conducted as part of a mixed methods formative evaluation involving quantitative and qualitative methods for a clinical trial at the 8 VHA medical centers in New England. The trial is testing an intervention in which veterans presenting for service-connection examinations for musculoskeletal conditions receive brief counseling to engage them in nonopioid pain treatments. The VHA Central Institutional Review Board approved this formative evaluation and the clinical trial has begun (ClinicalTrials.gov NCT04062214).

Potential interviewees were involved in referrals to and provision of nonpharmacologic pain treatment and were identified by site investigators in the randomized trial. Identified interviewees were clinical and administrative staff belonging to VHA Primary Care, Pain Management, and Compensation and Pension clinics. A total of 83 staff were identified.

Semistructured Interviews

A subset of the 83 staff were invited to participate in a semistructured interview because their position impacted coordination of pain care at their facilities or they worked in C&P. Staff at a site were interviewed until no new themes emerged from additional interviews, and each of the 8 sites was represented. Interviews were conducted between June and August 2018. Standardized scripts describing the study and inviting participation in a semistructured interview were e-mailed to VA staff. At the time of the interview the study purpose was restated and consent for audiotaping was obtained. The interviews followed a guide designed to assess a relational coordination framework among various workgroups. The data in this manuscript were elicited by specific prompts concerning: (1) How veterans learn about pain care when they come through C&P; and (2) How staff in C&P communicate with treatment providers about veterans who have chronic pain. Each interview lasted about 30 minutes.

Relational Coordination Survey

All identified staff were invited to participate in a relational coordination survey. The survey was administered through VA REDCap. Survey invitations were e-mailed from REDCap to VA staff and included a description of the study and assurances of the confidentiality of data collected. Surveys took < 10 minutes to complete. To begin, respondents identified their primary workgroup (C&P, primary care, pain management, or administrative leadership or staff), secondary workgroup (if they were in > 1), and site. Respondents provided no other identifying information and were assured their responses would be confidential.

The survey consisted of 7 questions regarding beliefs about the quality of communication and interactions among workgroup members in obtaining a shared goal.¹¹ The shared goal in the survey used in this study was providing pain care services for veterans with musculoskeletal conditions. Using a 5-point Likert scale, the 7 questions concerned frequency, timeliness, and accuracy of communication; response to problems providing pain services; sharing goals; and knowledge and respect for respondent’s job function. Higher scores indicated better relational coordination among members of a workgroup. Using the survey’s 7 items, composite mean relational coordination scores were calculated for each of the 4 primary workgroups. To account for the possibility that a member rated their own workgroups, 2 scores were created for each workgroup; one included members of the workgroup and another excluded them.

Data Analysis

The audio-recorded semistructured interviews were transcribed and entered into Atlas.ti qualitative data analysis software. To identify cross-cutting themes, a semistructured telephone interview guide was developed by the qualitative study team that emphasized interrelationships between different clinical teams. The transcripts were then analyzed using the grounded theory approach, a systematic methodology to reduce themes from collected qualitative data. Two research staff read each transcript twice; first to familiarize themselves with the text and then, using open coding, to identify important concepts that emerged from the language and assign codes to segments of text. To ensure accuracy, researchers included suitable contextual information in the coding. Using the constant comparative method, research staff then met to examine the themes that emerged in the interviews, discuss and coalesce coding discrepancies, and compare perspectives.¹⁷

The composite score (mean of the 7 items and 95% CI) of the survey responses was analyzed to identify significant differences in coordination across the 4 workgroups. Analysis of variance (ANOVA) was used to examine each relational coordination score by respondents’ workgroup. Post hoc analyses examined relational coordination survey differences among the 4 respondent groups.

Results

Thirty-nine survey respondents participated in the semistructured interviews. C&P examiners expressed varying degrees of comfort with their role in extending access to pain care for veterans. Some of the examiners strongly believed that their role was purely forensic, and going beyond this forensic role to refer or recommend treatment to veterans would be a violation of their role to conduct a forensic examination. “We don’t have an ongoing therapeutic relationship with any of the patients,” a C&P examiner explained: “We see them once; they’re out the door. It’s forensic. We’re investigating the person as a claimant, we’re investigating it and using our tools to go and review information from 30, 40 years ago.”

Other examiners had a less strict approach for working with veterans in C&P, even though examiners are asked not to provide advice or therapy. One C&P examiner noted that because he “can’t watch people in pain,” during the examination this doctor recommends that patients go to the office that determines whether they are eligible for benefits and choose a PCP. Another C&P examiner concurred with this approach. “I certainly spend a little time with the veteran talking to them about their personal life, who they are, what they do, what they’ve done, what they’re going to do to kind of break the ice between us,” the second examiner explained. “At the end, I will make some suggestions to them. I’m comfortable doing that. I don’t know that everybody is.”

Many of the VHA providers we interviewed had little knowledge of the C&P process or whether C&P examiners had any role or responsibilities in referring veterans for pain care. Most VHA providers could not name any C&P examiners at their facility and were generally unfamiliar with the content of C&P examinations. One provider bluntly said, “I’ve never communicated with anyone in comp and pen [C&P].”

Another PCP also expressed concerns with referrals, suggesting that C&P and primary care “are totally separate and should remain separate,” the PCP explained. “My concern with getting referral from comp and pen is that is it then they’re seeking all sorts of treatment that they wouldn’t necessarily need or ask for otherwise.”

Conversely a different PCP had a positive outlook on how C&P examiners might help ease the transition into the VHA for veterans with pain, especially for newly discharged veterans. “Having comp and pen address these issues is really going to be helpful. I think it could be significant that the topic is introduced early on.”

Relational Coordination Survey

Relational coordination surveys were sent to 83 participants of whom 66 responded. Respondents were from C&P (n = 7), primary care (n = 16), pain medicine (n = 32), and administration (n = 11). Of the 66 respondents, 18 indicated a secondary workgroup. Respondents on 2 teams (primary/secondary) were primary care/administrative (n = 4), pain management/primary care (n = 4), primary care/pain management (n = 3), administrative/primary care (n = 3), and C&P/administrative (n = 1).

The relational coordination composite scores were lowest for C&P. This finding remained whether C&P staff surveys were included or removed from the C&P responses. As demonstrated by the 95% CI, when team members’ surveys were included, C&P scores (95% CI, 2.01-2.42) were significantly lower than the primary care (95% CI, 3.34-3.64) and pain management (95% CI, 3.61-3.96) groups. All the relational coordination composite scores were slightly lower when staff who described their own workgroup were removed (ie, respondents rated their own workgroups as having higher relational coordination than others did). Using the composite scores excluding same workgroup members, the composite scores of the C&P remained significantly lower than all 3 other workgroups (Table). Means values for each individual item in the C&P group were significantly less than all other group means for each item except for the question on responses to problems providing pain services (data not shown). On this item only, the mean C&P rating was > 3 (3.19), but this was still lower than the means of the primary care and pain management workgroups.

Further analyses were undertaken to understand the importance of stakeholders’ ratings of their own workgroup compared with ratings by others of that workgroup. A 1-way ANOVA of workgroup was conducted and displayed significant workgroup differences between member and nonmember relational coordination ratings on 3 of the 4 workgroup’s scores C&P (F = 5.75, 3, 62 df; P < .01) primary care (F = 4.30, 3, 62 df; P < .008) and pain management (F = 8.22, 3, 62 df; P < .001). Post hoc contrasts between the different workgroups doing the rating revealed: (1) significant differences in the assessment of the C&P workgroup between the C&P workgroup and both the primary care (P < .01) and pain management groups (P < .001) with C&P rating their own workgroup significantly higher; (2) a significant difference in the scoring of the primary care workgroup with the primary care group rating themselves significantly higher than the C&P group; and (3) significant differences in the scoring of the pain management workgroup with both pain management and primary care groups rating the pain management group significantly higher than the C&P group. The results were not substantially changed by removing the 18 respondents who identified themselves as being part of > 1 workgroup .

Discussion

Mixed methods revealed disparate viewpoints about the role of C&P in referring veterans to pain care services. Overall, C&P teams coordinated less with other workgroups than the other groups coordinated with each other, and the C&P clinics took only limited steps to engage veterans in VHA treatment. The relational coordination results appeared to be valid. The mean scores were near the middle of the relational coordination rating scale, with standard deviations indicating a range of responses. The lower relational coordination scores of the C&P group remained after removing stakeholders who were rating their own workgroup. Further support for the validity of the relational coordination survey results is that they were consistent with the reports of C&P clinic isolation in the semistructured interviews.

The interview data suggest that one reason the C&P teams had low relational coordination scores is that VA staff interpret the emphasis on evaluative rather than therapeutic examinations to preclude other attempts to engage veterans into VHA treatment, even though such treatment engagement is permitted within existing guidelines. VBA referrals for examinations say nothing, either way, about engaging veterans in VHA care. The relational coordination results suggest that an intervention that might increase treatment referrals from the C&P clinics would be to explain the (existing) policy allowing for outreach around the time of compensation examinations to VHA staff so this goal is clearly agreed-upon. Another approach to facilitating treatment engagement at the C&P examination is to use other interventions that have been associated with better relational coordination such as intergroup meetings, horizontal integration more generally, and an atmosphere is which people from different backgrounds feel empowered to speak frankly to each other.^15,18,19 An important linkage to forge is between C&P teams and the administrative workgroups responsible for verifying a veteran’s eligibility for VHA care and enrolling eligible veterans in VHA treatment. Having C&P clinicians who are familiar with the eligibility and treatment engagement processes would facilitate providing that information to veterans, without compromising the evaluative format of the compensation examination.

An interesting ancillary finding is that relational coordination ratings by members of 3 of the 4 workgroups were higher than ratings by other staff of that workgroup. A possible explanation for this finding is that workgroup members are more aware of the relational coordination efforts made by their own workgroup than those by other workgroups, and therefore rate their own workgroup higher. This also might be part of a broader self-aggrandizement heuristic that has been described in multiple domains.²⁰ Staff may apply this heuristic in reporting that their staff engage in more relational coordination, reflecting the social desirability of being cooperative.

There are simple facility-level interventions that would facilitate veterans access to care such as conducting C&P examinations for potentially treatment-eligible veterans at VHA facilities (vs conducted outside VHA) and having access to materials that explain the treatment options to veterans when they check in for their compensation examinations. The approach to C&P-based treatment engagement that was successfully employed in 2 clinical trials involved having counselors not connected with the C&P clinic contact veterans around the time of their compensation examination to explain VA treatment options and motivate veterans to pursue treatment.^8,9 This independent counselor approach is being evaluated in a larger study.

Limitations

These data are from a small number of VA staff evaluating veterans in a single region of the US. They do not show causation, and it is possible that relational coordination is not necessary for referrals from C&P clinics. Relational coordination might not be necessary when referral processes can be simply routinized with little need for communication.¹¹ However, other analyses in these clinics have found that pain treatment referrals in fact are not routinized, with substantial variability within and across institutions. Another possibility is that features that have been associated with less relational coordination, such as male gender and medical specialist guild, were disproportionately present in C&P clinics compared to the other clinics.²¹Finally, veterans may be eligible for priority VA care for reasons that do not involve service-connection claims (38 CFR § 17.37).

Conclusions

There have been public calls to improve the evaluation of service-connection claims such that this process includes approaches to engage veterans in treatment.²² Referring veterans to treatment when they come for C&P examinations will likely involve improving relational coordination between the C&P service and other parts of VHA. Nationwide, sites that integrate C&P more fully may have valuable lessons to impart about the benefits of such integration. An important step towards better relational coordination will be clarifying that engaging veterans in VHA care around the time of their C&P examinations is a facility-wide goal.

Acknowledgments

The authors thank Brian Linde and Efia James for their perspectives on C&P procedures. This work was supported by the Veterans Integrated Service Network 1 Mental Illness Research Education and Clinical Center (MIRECC) and National Institute of Health, National Center for Complementary and Integrative Health Project # 5UG3AT009758-02. (MIR, SM mPIs).

Rituximab is a genetically engineered chimeric immunoglobulin G1 monoclonal antibody. It functions by binding to the CD20 antigen on the surface of B-cell lymphocytes, leading to complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity.¹ The US Food and Drug Administration approved this therapy to treat patients with B-cell non-Hodgkin lymphoma and chronic lymphocytic leukemia, along with other nonmalignant indications, including pemphigus vulgaris and rheumatoid arthritis (RA). Historically, a significant amount of time and labor on behalf of medical personnel has been required to administer rituximab according to the original manufacturer’s labeling due to the boxed warning associated with infusion-related reactions (IRRs).²

Originally, the elongated infusion times that were recommended for rituximab were largely due to the perceived risk of serious infusion-related adverse drug reactions. Slower infusion times should reduce the risk of a reaction and are considered to be a good option for those patients who are at a high risk of having a severe IRRs to rituximab. Examples of high-risk patients from previous studies include those with significant cardiovascular disease, a circulating lymphocyte count ≤ 5,000/µL at the start of infusion, and those who have previously had a reaction to rituximab.^3-5 In appropriate patients, research has shown a decreasing incidence of all-grade IRRs for patients who are prescribed rituximab as they receive more doses of the drug.^2,6 The ability to identify suitable patients for 90-minute infusions of rituximab and the prospect of better health system resource utilization has led investigators to study the effects of shortened infusion times.

The RATE trial addressed this subject with a phase 3 safety study on the effects of a 90-minute rituximab infusion for patients with previously untreated diffuse large B-cell and follicular lymphoma.³ The patients in this study received their first dose of rituximab using the traditional infusion approach. If it was well-tolerated, they received subsequent rituximab infusions using a 90-minute protocol. Only 1.1% of patients who had previously received a rituximab infusion developed a grade 3 or 4 IRR when receiving a faster infusion of the drug for the first time.³ This result led to the addition of instructions for a 90-minute infusion to the package insert.²

In contrast to the RATE trial, the RATE-RA trial evaluated the incidence of IRRs in patients who received rituximab for nonmalignant indications. This study assessed patients with RA receiving rituximab for > 120 minutes. The authors reported 0.6% of the patients in the study developed a grade 3 or 4 IRR associated with the first 120-minute infusion of the medication.⁵ The researchers concluded that rituximab can be administered at a faster rate during second and subsequent infusions in patients who have been shown to tolerate traditional infusions without increasing the risk or severity of IRRs.⁵

The US Department of Veterans Affairs (VA) Richard L. Roudebush VA Medical Center (RLRVAMC) in Indianapolis, Indiana, uses traditional directions for the infusion of rituximab due to perceived tolerability and safety concerns specifically in a veteran population—even while other VA medical centers have implemented shortened infusion protocols. This also is despite the fact that available research shows rapid infusions of the drug are well tolerated in a variety of community settings.^7,8 Anticipated benefits of implementing a protocol include savings in chair time at the institution’s infusion clinic along with increased nursing and patient satisfaction. This project was conducted to prepare, implement, and assess the safety of a 90-minute rituximab protocol at the RLRVAMC.

Methods

Proactive measures were required before and during the implementation of the 90-minute protocol to ensure patient safety and staff satisfaction. Updates to the RLRVAMC policy for the management of medical emergencies within the infusion center were reviewed and approved by the acute care committee and nursing leadership. A protocol was developed to identify eligible patients, outline the hypersensitivity protocol, instruct pharmacy personnel on admixture preparation, and provide a titration schedule based on dose. Order sets also were created to assist health care providers (HCPs) with the prescribing of rituximab for nonantineoplastic indications. Educational materials were crafted to assist with order verification, product preparation, labeling, and programming of infusion pumps. Live education was provided for physicians, pharmacists, and nurses to ensure smooth implementation of the protocol and appropriate management of medical emergencies based on the updated policy.

Study Design

Nursing staff in the infusion clinic were surveyed once before a live education session and again after the conclusion of the study. The purpose of the survey was to assess the prior experience and current comfort level of the nursing staff with administering rituximab over 90 minutes. Nurses were asked the following questions: (1) Do you have prior experience administering rituximab via 90-minute infusion; and (2) do you feel comfortable administering rituximab via 90-minute infusion?

A weekly report of patients who received rituximab between November 1, 2018 through April 1, 2019 at the RLRVAMC was generated. HCPs were alerted to eligible patients based on protocol requirements. The HCPs then made the final determination and entered orders accordingly.

This study was a retrospective chart review of all who patients received a rapid infusion of rituximab. Patients who were included if they were aged ≥ 18 years, received rituximab infusions in the RLRVAMC infusion clinic, had an absolute lymphocyte count ≤ 5,000/mm³ at the time of their rapid infusions, had no significant baseline cardiovascular disease or respiratory compromise, and had no prior grade 3 or 4 rituximab IRRs as defined by Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0.⁹ This study was a quality improvement initiative and considered exempt by the institutional review board. All data were deidentified and secured to ensure patient privacy.

The primary endpoint for this study was the incidence of grade 3 or 4 IRRs associated with the rapid infusion of rituximab. Secondary endpoints included the proportion of patients who experienced a grade 3 or 4 infusion reaction, who received proper treatment according to the institution’s hypersensitivity protocol, savings in infusion clinic chair time, and nursing satisfaction with education and implementation of the rapid infusion rituximab protocol.

The following data were collected for all included patients: demographics, lactic acid dehydrogenase level, white blood cell count, and absolute lymphocyte count prior to rituximab infusion, indication for treatment, dose of rituximab for 90-minute infusion, date of infusion, starting time, ending time, number of previous rituximab infusions within the past 3 months, symptoms of infusion reactions during rituximab infusion, and grade of any infusion reactions that occurred.

Estimated savings in infusion clinic chair time was calculated by taking the difference in time between each completed rapid infusion and the estimated amount of time it would have taken for each patient to receive a traditional infusion. The estimated amount of time for traditional infusion was determined by following the institution’s protocol for administering rituximab to patients who previously tolerated their first dose of the drug (eg, 100 mg/h starting rate and increasing by 100 mg/h every 30 minutes to a maximum infusion rate of 400 mg/h). All endpoints were analyzed using descriptive statistics.

Results

Between November 1, 2018 and April 1, 2019, 11 patients received a total of 24 rapid infusions of rituximab. The majority of patients included in the study were older males, and the most common indication for rapid infusion was follicular lymphoma (Table 1).

Primary Endpoint

All patients who received a rapid infusion of rituximab were reviewed in the analysis of the primary and secondary endpoints. Among the 24 rapid infusions of rituximab, 1 infusion was stopped due to the patient experiencing a grade 3 IRR according to criteria from CTCAE Version 5.0. The patient was found to have dysphagia at baseline and experienced severe symptoms in the days following the first infusion that put the patient at high risk for subsequent infusion related concerns. Eligibility criteria for the 90-minute protocol were updated based on these findings. No patient experienced a grade 4 or 5 IRR. The remaining 23 infusions were well tolerated by the patients with no clinically significant events.

Secondary Endpoints

The patient who experienced a grade 3 IRR to rituximab received proper treatment by infusion clinic nurses according to the RLRVAMC hypersensitivity protocol. Patients who received rapid infusions of rituximab had a mean length of infusion of 95.0 minutes. This was in contrast to the mean time of each patient’s previous nonrapid infusion of 134.3 minutes. The difference between the 2 values equated to a savings in infusion clinic chair mean time of 39.3 minutes per patient.

Nurses were asked whether they had prior experience administering rituximab via 90-minute infusion and whether they felt comfortable administering a 90-minute rituximab infusion. Before the live education session, none of the nurses surveyed had prior experience or felt comfortable administering rituximab over 90 minutes. When the nurses were surveyed poststudy, all reported that they were experienced administering rituximab and felt comfortable with the process (Table 2).

Discussion

The infusion of rituximab has been associated with significant challenges related to the time and labor required. Although a vast number of institutions across the country now infuse the medication over an abbreviated time, HCP concerns for patient safety and appropriate use of hypersensitivity protocol in a veteran population delayed implementation at RLRVAMC. The results from this quality improvement initiative highlight the positive impact of the proactive measures that were used to implement the rapid infusion protocol for rituximab on improving HCP prescribing rates, nursing satisfaction, and appropriate management of IRRs.

Rapid infusion saved on average 39.3 minutes per patient in infusion clinic chair time. Each successful rapid infusion of rituximab potentially opened additional time in clinic for ≥ 1 patients to receive an infusion therapy. The RLRVAMC usually operated at maximum capacity, so the ability to accommodate more patients helped decrease hospital admittances for time-sensitive infusions.

The initial criteria used to screen patients to determine whether a rapid infusion of rituximab would be appropriate was based on inclusion and exclusion criteria for past studies on the same subject.^3-5 The incidence of hypersensitivity reactions associated with study participants who received rapid rituximab infusions also resembles past research done on the subject, which is important to note due to prior misconceptions of staff at the institution of a higher risk of reaction in this specific veteran population. One patient with RA experienced a grade 3 IRR in this study. Although this patient met the original inclusion criteria, the patient had baseline dysphagia, and following the first infusion, reported to the emergency department (ED) with symptoms of delayed anaphylaxis. In this case, the order for rapid infusion was placed in advance and the prescriber was unaware of the ED visit. Based on this event, eligibility criteria for 90-minute rituximab infusions were updated to include additional information specifying that candidates for a rapid infusion also may have no baseline airway compromise. This hypersensitivity reaction also highlighted the need for decision support technology to assist HCPs in patient selection as well as empowering nursing and pharmacy staff to identify concerns once they place orders.

Over the course of the study, investigators assisted the HCPs with preparation of orders for the rapid infusion of rituximab for antineoplastic indications. Due to feasibility issues with this practice moving forward, order sets containing rituximab were updated to include a 90-minute option. This created a more standardized process that allowed HCPs to screen potential patients on their own. The expectation is that HCPs will be more likely to order 90-minute infusions for eligible patients in the future with this efficient and safer process.

Limitations

The small sample size in this study was a limitation. Retrospective data related to the management of infusion reactions and length of infusions were collected from nursing notes. The prospective use of a standardized evaluation tool for adverse drug reactions as well as bar code medication administration technology would improve the data available for this study. Additional studies also would be useful to validate the results.

Conclusions

The proactive measures that were used to implement the rapid infusion rituximab protocol improved HCP prescribing rates, nursing satisfaction, and the management of IRRs. Potential time savings with each infusion was significant. This study confirmed appropriateness of rapid administration of rituximab in this veteran population and has increased interest in implementing other rapid infusion protocols. Protocols, education, and order sets are being developed for daratumumab and infliximab.

Rituximab is a genetically engineered chimeric immunoglobulin G1 monoclonal antibody. It functions by binding to the CD20 antigen on the surface of B-cell lymphocytes, leading to complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity.¹ The US Food and Drug Administration approved this therapy to treat patients with B-cell non-Hodgkin lymphoma and chronic lymphocytic leukemia, along with other nonmalignant indications, including pemphigus vulgaris and rheumatoid arthritis (RA). Historically, a significant amount of time and labor on behalf of medical personnel has been required to administer rituximab according to the original manufacturer’s labeling due to the boxed warning associated with infusion-related reactions (IRRs).²

Originally, the elongated infusion times that were recommended for rituximab were largely due to the perceived risk of serious infusion-related adverse drug reactions. Slower infusion times should reduce the risk of a reaction and are considered to be a good option for those patients who are at a high risk of having a severe IRRs to rituximab. Examples of high-risk patients from previous studies include those with significant cardiovascular disease, a circulating lymphocyte count ≤ 5,000/µL at the start of infusion, and those who have previously had a reaction to rituximab.^3-5 In appropriate patients, research has shown a decreasing incidence of all-grade IRRs for patients who are prescribed rituximab as they receive more doses of the drug.^2,6 The ability to identify suitable patients for 90-minute infusions of rituximab and the prospect of better health system resource utilization has led investigators to study the effects of shortened infusion times.

The RATE trial addressed this subject with a phase 3 safety study on the effects of a 90-minute rituximab infusion for patients with previously untreated diffuse large B-cell and follicular lymphoma.³ The patients in this study received their first dose of rituximab using the traditional infusion approach. If it was well-tolerated, they received subsequent rituximab infusions using a 90-minute protocol. Only 1.1% of patients who had previously received a rituximab infusion developed a grade 3 or 4 IRR when receiving a faster infusion of the drug for the first time.³ This result led to the addition of instructions for a 90-minute infusion to the package insert.²

In contrast to the RATE trial, the RATE-RA trial evaluated the incidence of IRRs in patients who received rituximab for nonmalignant indications. This study assessed patients with RA receiving rituximab for > 120 minutes. The authors reported 0.6% of the patients in the study developed a grade 3 or 4 IRR associated with the first 120-minute infusion of the medication.⁵ The researchers concluded that rituximab can be administered at a faster rate during second and subsequent infusions in patients who have been shown to tolerate traditional infusions without increasing the risk or severity of IRRs.⁵

The US Department of Veterans Affairs (VA) Richard L. Roudebush VA Medical Center (RLRVAMC) in Indianapolis, Indiana, uses traditional directions for the infusion of rituximab due to perceived tolerability and safety concerns specifically in a veteran population—even while other VA medical centers have implemented shortened infusion protocols. This also is despite the fact that available research shows rapid infusions of the drug are well tolerated in a variety of community settings.^7,8 Anticipated benefits of implementing a protocol include savings in chair time at the institution’s infusion clinic along with increased nursing and patient satisfaction. This project was conducted to prepare, implement, and assess the safety of a 90-minute rituximab protocol at the RLRVAMC.

Methods

Proactive measures were required before and during the implementation of the 90-minute protocol to ensure patient safety and staff satisfaction. Updates to the RLRVAMC policy for the management of medical emergencies within the infusion center were reviewed and approved by the acute care committee and nursing leadership. A protocol was developed to identify eligible patients, outline the hypersensitivity protocol, instruct pharmacy personnel on admixture preparation, and provide a titration schedule based on dose. Order sets also were created to assist health care providers (HCPs) with the prescribing of rituximab for nonantineoplastic indications. Educational materials were crafted to assist with order verification, product preparation, labeling, and programming of infusion pumps. Live education was provided for physicians, pharmacists, and nurses to ensure smooth implementation of the protocol and appropriate management of medical emergencies based on the updated policy.

Study Design

Nursing staff in the infusion clinic were surveyed once before a live education session and again after the conclusion of the study. The purpose of the survey was to assess the prior experience and current comfort level of the nursing staff with administering rituximab over 90 minutes. Nurses were asked the following questions: (1) Do you have prior experience administering rituximab via 90-minute infusion; and (2) do you feel comfortable administering rituximab via 90-minute infusion?

A weekly report of patients who received rituximab between November 1, 2018 through April 1, 2019 at the RLRVAMC was generated. HCPs were alerted to eligible patients based on protocol requirements. The HCPs then made the final determination and entered orders accordingly.

This study was a retrospective chart review of all who patients received a rapid infusion of rituximab. Patients who were included if they were aged ≥ 18 years, received rituximab infusions in the RLRVAMC infusion clinic, had an absolute lymphocyte count ≤ 5,000/mm³ at the time of their rapid infusions, had no significant baseline cardiovascular disease or respiratory compromise, and had no prior grade 3 or 4 rituximab IRRs as defined by Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0.⁹ This study was a quality improvement initiative and considered exempt by the institutional review board. All data were deidentified and secured to ensure patient privacy.

The primary endpoint for this study was the incidence of grade 3 or 4 IRRs associated with the rapid infusion of rituximab. Secondary endpoints included the proportion of patients who experienced a grade 3 or 4 infusion reaction, who received proper treatment according to the institution’s hypersensitivity protocol, savings in infusion clinic chair time, and nursing satisfaction with education and implementation of the rapid infusion rituximab protocol.

The following data were collected for all included patients: demographics, lactic acid dehydrogenase level, white blood cell count, and absolute lymphocyte count prior to rituximab infusion, indication for treatment, dose of rituximab for 90-minute infusion, date of infusion, starting time, ending time, number of previous rituximab infusions within the past 3 months, symptoms of infusion reactions during rituximab infusion, and grade of any infusion reactions that occurred.

Estimated savings in infusion clinic chair time was calculated by taking the difference in time between each completed rapid infusion and the estimated amount of time it would have taken for each patient to receive a traditional infusion. The estimated amount of time for traditional infusion was determined by following the institution’s protocol for administering rituximab to patients who previously tolerated their first dose of the drug (eg, 100 mg/h starting rate and increasing by 100 mg/h every 30 minutes to a maximum infusion rate of 400 mg/h). All endpoints were analyzed using descriptive statistics.

Results

Between November 1, 2018 and April 1, 2019, 11 patients received a total of 24 rapid infusions of rituximab. The majority of patients included in the study were older males, and the most common indication for rapid infusion was follicular lymphoma (Table 1).

Primary Endpoint

All patients who received a rapid infusion of rituximab were reviewed in the analysis of the primary and secondary endpoints. Among the 24 rapid infusions of rituximab, 1 infusion was stopped due to the patient experiencing a grade 3 IRR according to criteria from CTCAE Version 5.0. The patient was found to have dysphagia at baseline and experienced severe symptoms in the days following the first infusion that put the patient at high risk for subsequent infusion related concerns. Eligibility criteria for the 90-minute protocol were updated based on these findings. No patient experienced a grade 4 or 5 IRR. The remaining 23 infusions were well tolerated by the patients with no clinically significant events.

Secondary Endpoints

The patient who experienced a grade 3 IRR to rituximab received proper treatment by infusion clinic nurses according to the RLRVAMC hypersensitivity protocol. Patients who received rapid infusions of rituximab had a mean length of infusion of 95.0 minutes. This was in contrast to the mean time of each patient’s previous nonrapid infusion of 134.3 minutes. The difference between the 2 values equated to a savings in infusion clinic chair mean time of 39.3 minutes per patient.

Nurses were asked whether they had prior experience administering rituximab via 90-minute infusion and whether they felt comfortable administering a 90-minute rituximab infusion. Before the live education session, none of the nurses surveyed had prior experience or felt comfortable administering rituximab over 90 minutes. When the nurses were surveyed poststudy, all reported that they were experienced administering rituximab and felt comfortable with the process (Table 2).

Discussion

The infusion of rituximab has been associated with significant challenges related to the time and labor required. Although a vast number of institutions across the country now infuse the medication over an abbreviated time, HCP concerns for patient safety and appropriate use of hypersensitivity protocol in a veteran population delayed implementation at RLRVAMC. The results from this quality improvement initiative highlight the positive impact of the proactive measures that were used to implement the rapid infusion protocol for rituximab on improving HCP prescribing rates, nursing satisfaction, and appropriate management of IRRs.

Rapid infusion saved on average 39.3 minutes per patient in infusion clinic chair time. Each successful rapid infusion of rituximab potentially opened additional time in clinic for ≥ 1 patients to receive an infusion therapy. The RLRVAMC usually operated at maximum capacity, so the ability to accommodate more patients helped decrease hospital admittances for time-sensitive infusions.

The initial criteria used to screen patients to determine whether a rapid infusion of rituximab would be appropriate was based on inclusion and exclusion criteria for past studies on the same subject.^3-5 The incidence of hypersensitivity reactions associated with study participants who received rapid rituximab infusions also resembles past research done on the subject, which is important to note due to prior misconceptions of staff at the institution of a higher risk of reaction in this specific veteran population. One patient with RA experienced a grade 3 IRR in this study. Although this patient met the original inclusion criteria, the patient had baseline dysphagia, and following the first infusion, reported to the emergency department (ED) with symptoms of delayed anaphylaxis. In this case, the order for rapid infusion was placed in advance and the prescriber was unaware of the ED visit. Based on this event, eligibility criteria for 90-minute rituximab infusions were updated to include additional information specifying that candidates for a rapid infusion also may have no baseline airway compromise. This hypersensitivity reaction also highlighted the need for decision support technology to assist HCPs in patient selection as well as empowering nursing and pharmacy staff to identify concerns once they place orders.

Over the course of the study, investigators assisted the HCPs with preparation of orders for the rapid infusion of rituximab for antineoplastic indications. Due to feasibility issues with this practice moving forward, order sets containing rituximab were updated to include a 90-minute option. This created a more standardized process that allowed HCPs to screen potential patients on their own. The expectation is that HCPs will be more likely to order 90-minute infusions for eligible patients in the future with this efficient and safer process.

Limitations

The small sample size in this study was a limitation. Retrospective data related to the management of infusion reactions and length of infusions were collected from nursing notes. The prospective use of a standardized evaluation tool for adverse drug reactions as well as bar code medication administration technology would improve the data available for this study. Additional studies also would be useful to validate the results.

Conclusions

The proactive measures that were used to implement the rapid infusion rituximab protocol improved HCP prescribing rates, nursing satisfaction, and the management of IRRs. Potential time savings with each infusion was significant. This study confirmed appropriateness of rapid administration of rituximab in this veteran population and has increased interest in implementing other rapid infusion protocols. Protocols, education, and order sets are being developed for daratumumab and infliximab.

Rituximab is a genetically engineered chimeric immunoglobulin G1 monoclonal antibody. It functions by binding to the CD20 antigen on the surface of B-cell lymphocytes, leading to complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity.¹ The US Food and Drug Administration approved this therapy to treat patients with B-cell non-Hodgkin lymphoma and chronic lymphocytic leukemia, along with other nonmalignant indications, including pemphigus vulgaris and rheumatoid arthritis (RA). Historically, a significant amount of time and labor on behalf of medical personnel has been required to administer rituximab according to the original manufacturer’s labeling due to the boxed warning associated with infusion-related reactions (IRRs).²

Originally, the elongated infusion times that were recommended for rituximab were largely due to the perceived risk of serious infusion-related adverse drug reactions. Slower infusion times should reduce the risk of a reaction and are considered to be a good option for those patients who are at a high risk of having a severe IRRs to rituximab. Examples of high-risk patients from previous studies include those with significant cardiovascular disease, a circulating lymphocyte count ≤ 5,000/µL at the start of infusion, and those who have previously had a reaction to rituximab.^3-5 In appropriate patients, research has shown a decreasing incidence of all-grade IRRs for patients who are prescribed rituximab as they receive more doses of the drug.^2,6 The ability to identify suitable patients for 90-minute infusions of rituximab and the prospect of better health system resource utilization has led investigators to study the effects of shortened infusion times.

The RATE trial addressed this subject with a phase 3 safety study on the effects of a 90-minute rituximab infusion for patients with previously untreated diffuse large B-cell and follicular lymphoma.³ The patients in this study received their first dose of rituximab using the traditional infusion approach. If it was well-tolerated, they received subsequent rituximab infusions using a 90-minute protocol. Only 1.1% of patients who had previously received a rituximab infusion developed a grade 3 or 4 IRR when receiving a faster infusion of the drug for the first time.³ This result led to the addition of instructions for a 90-minute infusion to the package insert.²

In contrast to the RATE trial, the RATE-RA trial evaluated the incidence of IRRs in patients who received rituximab for nonmalignant indications. This study assessed patients with RA receiving rituximab for > 120 minutes. The authors reported 0.6% of the patients in the study developed a grade 3 or 4 IRR associated with the first 120-minute infusion of the medication.⁵ The researchers concluded that rituximab can be administered at a faster rate during second and subsequent infusions in patients who have been shown to tolerate traditional infusions without increasing the risk or severity of IRRs.⁵

The US Department of Veterans Affairs (VA) Richard L. Roudebush VA Medical Center (RLRVAMC) in Indianapolis, Indiana, uses traditional directions for the infusion of rituximab due to perceived tolerability and safety concerns specifically in a veteran population—even while other VA medical centers have implemented shortened infusion protocols. This also is despite the fact that available research shows rapid infusions of the drug are well tolerated in a variety of community settings.^7,8 Anticipated benefits of implementing a protocol include savings in chair time at the institution’s infusion clinic along with increased nursing and patient satisfaction. This project was conducted to prepare, implement, and assess the safety of a 90-minute rituximab protocol at the RLRVAMC.

Methods

Proactive measures were required before and during the implementation of the 90-minute protocol to ensure patient safety and staff satisfaction. Updates to the RLRVAMC policy for the management of medical emergencies within the infusion center were reviewed and approved by the acute care committee and nursing leadership. A protocol was developed to identify eligible patients, outline the hypersensitivity protocol, instruct pharmacy personnel on admixture preparation, and provide a titration schedule based on dose. Order sets also were created to assist health care providers (HCPs) with the prescribing of rituximab for nonantineoplastic indications. Educational materials were crafted to assist with order verification, product preparation, labeling, and programming of infusion pumps. Live education was provided for physicians, pharmacists, and nurses to ensure smooth implementation of the protocol and appropriate management of medical emergencies based on the updated policy.

Study Design

Nursing staff in the infusion clinic were surveyed once before a live education session and again after the conclusion of the study. The purpose of the survey was to assess the prior experience and current comfort level of the nursing staff with administering rituximab over 90 minutes. Nurses were asked the following questions: (1) Do you have prior experience administering rituximab via 90-minute infusion; and (2) do you feel comfortable administering rituximab via 90-minute infusion?

A weekly report of patients who received rituximab between November 1, 2018 through April 1, 2019 at the RLRVAMC was generated. HCPs were alerted to eligible patients based on protocol requirements. The HCPs then made the final determination and entered orders accordingly.

This study was a retrospective chart review of all who patients received a rapid infusion of rituximab. Patients who were included if they were aged ≥ 18 years, received rituximab infusions in the RLRVAMC infusion clinic, had an absolute lymphocyte count ≤ 5,000/mm³ at the time of their rapid infusions, had no significant baseline cardiovascular disease or respiratory compromise, and had no prior grade 3 or 4 rituximab IRRs as defined by Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0.⁹ This study was a quality improvement initiative and considered exempt by the institutional review board. All data were deidentified and secured to ensure patient privacy.

The primary endpoint for this study was the incidence of grade 3 or 4 IRRs associated with the rapid infusion of rituximab. Secondary endpoints included the proportion of patients who experienced a grade 3 or 4 infusion reaction, who received proper treatment according to the institution’s hypersensitivity protocol, savings in infusion clinic chair time, and nursing satisfaction with education and implementation of the rapid infusion rituximab protocol.

The following data were collected for all included patients: demographics, lactic acid dehydrogenase level, white blood cell count, and absolute lymphocyte count prior to rituximab infusion, indication for treatment, dose of rituximab for 90-minute infusion, date of infusion, starting time, ending time, number of previous rituximab infusions within the past 3 months, symptoms of infusion reactions during rituximab infusion, and grade of any infusion reactions that occurred.

Estimated savings in infusion clinic chair time was calculated by taking the difference in time between each completed rapid infusion and the estimated amount of time it would have taken for each patient to receive a traditional infusion. The estimated amount of time for traditional infusion was determined by following the institution’s protocol for administering rituximab to patients who previously tolerated their first dose of the drug (eg, 100 mg/h starting rate and increasing by 100 mg/h every 30 minutes to a maximum infusion rate of 400 mg/h). All endpoints were analyzed using descriptive statistics.

Results

Between November 1, 2018 and April 1, 2019, 11 patients received a total of 24 rapid infusions of rituximab. The majority of patients included in the study were older males, and the most common indication for rapid infusion was follicular lymphoma (Table 1).

Primary Endpoint

All patients who received a rapid infusion of rituximab were reviewed in the analysis of the primary and secondary endpoints. Among the 24 rapid infusions of rituximab, 1 infusion was stopped due to the patient experiencing a grade 3 IRR according to criteria from CTCAE Version 5.0. The patient was found to have dysphagia at baseline and experienced severe symptoms in the days following the first infusion that put the patient at high risk for subsequent infusion related concerns. Eligibility criteria for the 90-minute protocol were updated based on these findings. No patient experienced a grade 4 or 5 IRR. The remaining 23 infusions were well tolerated by the patients with no clinically significant events.

Secondary Endpoints

The patient who experienced a grade 3 IRR to rituximab received proper treatment by infusion clinic nurses according to the RLRVAMC hypersensitivity protocol. Patients who received rapid infusions of rituximab had a mean length of infusion of 95.0 minutes. This was in contrast to the mean time of each patient’s previous nonrapid infusion of 134.3 minutes. The difference between the 2 values equated to a savings in infusion clinic chair mean time of 39.3 minutes per patient.

Nurses were asked whether they had prior experience administering rituximab via 90-minute infusion and whether they felt comfortable administering a 90-minute rituximab infusion. Before the live education session, none of the nurses surveyed had prior experience or felt comfortable administering rituximab over 90 minutes. When the nurses were surveyed poststudy, all reported that they were experienced administering rituximab and felt comfortable with the process (Table 2).

Discussion

The infusion of rituximab has been associated with significant challenges related to the time and labor required. Although a vast number of institutions across the country now infuse the medication over an abbreviated time, HCP concerns for patient safety and appropriate use of hypersensitivity protocol in a veteran population delayed implementation at RLRVAMC. The results from this quality improvement initiative highlight the positive impact of the proactive measures that were used to implement the rapid infusion protocol for rituximab on improving HCP prescribing rates, nursing satisfaction, and appropriate management of IRRs.

Rapid infusion saved on average 39.3 minutes per patient in infusion clinic chair time. Each successful rapid infusion of rituximab potentially opened additional time in clinic for ≥ 1 patients to receive an infusion therapy. The RLRVAMC usually operated at maximum capacity, so the ability to accommodate more patients helped decrease hospital admittances for time-sensitive infusions.

The initial criteria used to screen patients to determine whether a rapid infusion of rituximab would be appropriate was based on inclusion and exclusion criteria for past studies on the same subject.^3-5 The incidence of hypersensitivity reactions associated with study participants who received rapid rituximab infusions also resembles past research done on the subject, which is important to note due to prior misconceptions of staff at the institution of a higher risk of reaction in this specific veteran population. One patient with RA experienced a grade 3 IRR in this study. Although this patient met the original inclusion criteria, the patient had baseline dysphagia, and following the first infusion, reported to the emergency department (ED) with symptoms of delayed anaphylaxis. In this case, the order for rapid infusion was placed in advance and the prescriber was unaware of the ED visit. Based on this event, eligibility criteria for 90-minute rituximab infusions were updated to include additional information specifying that candidates for a rapid infusion also may have no baseline airway compromise. This hypersensitivity reaction also highlighted the need for decision support technology to assist HCPs in patient selection as well as empowering nursing and pharmacy staff to identify concerns once they place orders.

Over the course of the study, investigators assisted the HCPs with preparation of orders for the rapid infusion of rituximab for antineoplastic indications. Due to feasibility issues with this practice moving forward, order sets containing rituximab were updated to include a 90-minute option. This created a more standardized process that allowed HCPs to screen potential patients on their own. The expectation is that HCPs will be more likely to order 90-minute infusions for eligible patients in the future with this efficient and safer process.

Limitations

The small sample size in this study was a limitation. Retrospective data related to the management of infusion reactions and length of infusions were collected from nursing notes. The prospective use of a standardized evaluation tool for adverse drug reactions as well as bar code medication administration technology would improve the data available for this study. Additional studies also would be useful to validate the results.

Conclusions

The proactive measures that were used to implement the rapid infusion rituximab protocol improved HCP prescribing rates, nursing satisfaction, and the management of IRRs. Potential time savings with each infusion was significant. This study confirmed appropriateness of rapid administration of rituximab in this veteran population and has increased interest in implementing other rapid infusion protocols. Protocols, education, and order sets are being developed for daratumumab and infliximab.

Consolidated mail outpatient pharmacy (CMOP) is an automated prescription order processing and delivery system developed by the US Department of Veterans Affairs (VA) in 1994 to provide medications to VA patients.¹ In fiscal year (FY) 2016, CMOP filled about 80% of VA outpatient prescriptions.²

Formalized by the 2010 Memorandum of Understanding between Indian Health Service (IHS) and VA, CMOP is a partnership undertaken to improve the delivery of care to patients by both agencies.³ The number of prescriptions filled by CMOP for IHS patients increased from 1,972 in FY 2010 to 840,109 in FY 2018.⁴ In the fourth quarter of FY 2018, there were 94 CMOP-enrolled IHS federal and tribal sites.⁵ It is only appropriate that a growing number of IHS sites are adopting CMOP considering the evidence for mail-order pharmacy on better patient adherence, improved health outcomes, and potential cost savings.^6-9 Furthermore, using a centralized pharmacy operation, such as CMOP, can lead to better quality services.¹⁰

Crownpoint Health Care Facility (CHCF) serves > 30,000 American Indians and is in Crownpoint, New Mexico, a small community of about 3,000 people.¹¹ Most of the patients served by the facility live in distant places. Many of these underserved patients do not have a stable means of transportation.¹² Therefore, these patients may have difficulty traveling to the facility for their health care needs, including medication pickups. More than 2.5 million American Indians and Alaska Natives IHS beneficiaries face similar challenges due to the rurality of their communities.¹³ CMOP can be a method to increase access to care for this vulnerable population. However, the utilization of CMOP varies significantly among IHS facilities. While some IHS facilities process large numbers of prescriptions through CMOP, other facilities process few, if any. There also are IHS facilities, such as CHCF, which are at the initial stage of implementing CMOP or trying to increase the volume of prescriptions processed through CMOP. Although the utilization of CMOP has grown exponentially among IHS facilities, there is currently no available resource that summarizes the relative advantages and disadvantages, the challenges and opportunities, and the strengths and weaknesses of implementing CMOP for IHS facilities

Methods

A questionnaire encompassing various aspects of CMOP prescription processing was developed and distributed to the primary CMOP contacts for IHS facilities. The questionnaire was first distributed by e-mail on December 19, 2018. It was e-mailed for a second time on January 16, 2019, and the questionnaire was open for responses until the end of January 2019 (Table).

Results

Forty-four of 94 CMOP-enrolled IHS sites responded to the questionnaire. Most sites train the majority of their pharmacists in CMOP prescription processing. Overall, 310 of 347 pharmacists (89%) in these 44 IHS sites can process prescriptions through CMOP. Thirty-one sites have all their pharmacists trained in CMOP prescription processing. Only 1 facility had less than half (2 of 17 pharmacists) of its pharmacists trained in CMOP prescription processing. More than half the total number of pharmacists, 185 out of 347 (53%), check electronic messages via Resource and Patient Management System (RPMS) MailMan to get information about prescriptions rejected by CMOP. Twenty sites have all their pharmacists check messages about CMOP rejections. However, 2 facilities reported that they do not check the rejection messages at all. Twenty-six of the 44 responding sites (59%) transmit prescriptions to CMOP manually in the electronic system. The rest (18 of 44) rely on the auto-transmission (AT) setup to transmit the CMOP-suspended prescriptions at specified times of the day.

Half the sites (8 of 16) that rely on patients asking for prescriptions to be mailed at the time of refill request do not use any method to designate a CMOP patient. Twenty-four sites use the narrative field on the patient’s profile in RPMS, the health information system used by most IHS facilities, to designate CMOP patients. Eighteen sites use pop-up messages on ScriptPro, a pharmacy automation system, as a designation method. Most of the sites (12 of 15) that use both RPMS and ScriptPro designation methods do not require patients to ask for prescriptions to be mailed at the time of refill request; prescriptions for these patients are routed through CMOP unless patients request otherwise. Only 3 of 44 sites use both methods and rely on patients asking for prescriptions to be mailed at the time of refill request. Some other reported designation methods were using the electronic health record (EHR) posting box, keeping a manual list of CMOP patients, and solely utilizing the Prescription Mail Delivery field in RPMS. Three sites also noted that they keep manual lists to auto-refill prescriptions through CMOP.

Thirty sites (68%) reported that they process every prescription through CMOP even if the patient had prescriptions with specified CMOP quantities. Only 8 sites (18%) said that they used the local mail-out program to keep the same days’ supply for all medication orders. For patients with CMOP-ineligible prescriptions, 34 of the 44 sites (77%) process the eligible prescriptions through CMOP and refill the rest of the prescriptions locally. Six sites (14%) process all medication orders locally for patients with any CMOP-ineligible prescriptions.

Only 12 of 44 sites (27%) involve pharmacy technicians in CMOP prescription processing. Five sites have technicians process prescription refills through CMOP. Two of these sites mentioned the strategy of technicians suspending the prescriptions to be sent to CMOP on the refill due date. Other technician roles included tracking CMOP packages, checking electronic messages for CMOP rejections, and signing up patients for CMOP.

Only 3 of the 44 sites (7%) have measured patient satisfaction with the CMOP program. One of these 3 sites reported that the overall satisfaction was high with CMOP. This site administered the survey to patients who came to the clinic for appointments. The second facility called patients and asked for their feedback. The third site conducted the survey by using student pharmacists. Two sites reported that they use the survey results from the CMOP-conducted patient satisfaction surveys, although they have not measured patient satisfaction at their specific facilities.

Most sites have not assessed CMOP’s impact on their insurance (point of sale) collections. However, 13 sites (30%) reported that they believe they are losing on collections by utilizing CMOP. The use of repackaged products by CMOP, which are usually nonreimbursable, is an issue that was mentioned multiple times. In contrast, 2 sites mentioned that CMOP has led to increased insurance collections for their facilities.

Discussion

The utility of CMOP among the responding IHS sites varies quite significantly. Some sites appreciate the convenience of CMOP while acknowledging its limitations, such as the possible decrease in insurance collections, lengthy prescription processing time, or medication backorders. However, some sites have reserved CMOP for special circumstances (eg, mailing refrigerated items to the patient’s street address) due to various complexities that may come with CMOP. One site reported that it compares IHS contract drug prices with VA contract drug prices quarterly to determine which prescriptions should be sent through CMOP.

Most of the IHS pharmacists (89%) are trained in CMOP prescription processing. If an IHS site wants to increase its volume of CMOP prescriptions, it is sensible to train as many pharmacists as possible so that the responsibility does not fall on a few pharmacists. Newly hired pharmacists can receive guidance from trained pharmacists. Designation methods for CMOP patients can be beneficial for these pharmacists to identify CMOP-enrolled patients, especially if the site does not require patients to ask for prescriptions to be mailed at the time of refill request. Only 3 sites (7%) use multiple designation methods in addition to relying on patients to ask for prescriptions to be mailed. Proper implementation of designation methods can remove this extra burden on patients. Conversely, requiring patients to ask for prescriptions to be sent through CMOP can prevent spontaneous mail-outs if a CMOP-designated patient wants to pick up prescriptions locally. Overall, 16 sites (36%) rely on patients asking for prescriptions to be mailed.

One of the main benefits of CMOP is the ability to mail refrigerated items. Local pharmacy mail-out programs may not have this ability. Patients at rural locations often use post office (PO) boxes because they are unable to receive postal services at their physical addresses; however, they may receive packages through United Parcel Service (UPS) at their physical addresses. CMOP uses UPS to send refrigerated items, but UPS does not deliver to PO boxes. Therefore, remotely located sites like CHCF have difficulty in fully optimizing this benefit. One solution is documenting both the physical and mailing addresses on the patient’s EHR, which enables CMOP to send refrigerated items to the patient’s home address via UPS and mail the rest of the prescriptions to the patient’s PO box address with the US Postal Service. The physical address must be listed above the PO box address to ensure that refrigerated items are not rejected by CMOP. Furthermore, both the physical address and the PO box address must be in the same city for this method to work. Two sites noted mailing refrigerated items as one of the major challenges in CMOP prescription processing.

CMOP-enrolled patients must be educated about requesting medications 7 to 10 days before they run out. There is no standard time line for prescriptions filled by CMOP. However, 1 site reported that it may take up to “10 days from time requested to mailbox.” This delay leads to pharmacies facing a dilemma as processing prescriptions too early can lead to insurance rejections, but processing them too late can lead to the patient not receiving the medication by the time they run out of their current supply. However, CMOP provides the ability to track prescriptions sent through CMOP. Pharmacists and technicians need to have access to BestWay Parcel Services Client Portal (genco-mms.bestwayparcel.com) to track CMOP packages. Tracking CMOP prescriptions is a way pharmacy technicians can be involved in CMOP prescription processing. Technicians seem to be underutilized, as only 27% of the responding sites utilize them to some degree in the CMOP process. One site delegated the responsibility of checking CMOP rejection messages to pharmacy technicians. Since 2 of the responding sites do not check CMOP rejection messages at all, this is an excellent opportunity to get pharmacy technicians involved.

A CMOP auto-refill program can potentially be utilized to avoid missed or late medications. In an auto-refill program, a pharmacist can refill prescriptions through CMOP on the due date without a patient request. They may get rejected by insurance the first time they are processed through CMOP for refilling too early if the processing time is taken into account. However, the subsequent refills do not have to consider the CMOP processing time as they would already be synchronized based on the last refill date. Though, if CMOP is out of stock on a medication and it is expected to be available soon, CMOP may take a few extra days to either fill the prescription or reject it if the drug stays unavailable. One of the sites reported “the amount of time [CMOP] holds medications if they are out of stock” as “the hardest thing to work around.” A couple of sites also mentioned the longer than usual delay in processing prescriptions by CMOP during the holidays as one of the major challenges.

CMOP use of repackaged products also may lead insurance companies to deny reimbursement. Repackaged products are usually cheaper to buy.¹⁴ However, most insurances do not reimburse for prescriptions filled with these products.¹⁵ The local drug file on RPMS may have a national drug code (NDC) that is reimbursable by insurance, but CMOP will change it to the repackaged NDC if they are filling the prescription with a repackaged product. One potential solution to this problem would be filling these prescriptions locally. Furthermore, insurance claims are processed when the prescriptions are filled by CMOP. Sites cannot return/cancel the prescription anymore at that point. Therefore, the inability to see real-time rejections as the medication orders are processed on-site makes it challenging to prevent avoidable insurance rejections, such as a refill too soon. One site calculated that it lost $26,386.45 by utilizing CMOP from January 9, 2018 to December 12, 2018. However, it is unclear whether this loss was representative of other sites. It is also worth noting that IHS sites can save a substantial amount of money on certain products by utilizing CMOP because VA buys these products at a reduced price.¹⁶

CMOP-transmitted prescriptions can be rejected for various reasons, such as CMOP manufacturer’s backorder, a different quantity from CMOP stock size, etc. Information about these rejected prescriptions is accessed through electronic messages on RPMS. CMOP does not dispense less than a full, unopened package for most over-the-counter (OTC) medications. The quantity on these prescriptions must be equal to or multiples of the package size for them to be filled by CMOP. This can lead to a patient having prescriptions with different days’ supplies, which results in various refill due dates. If a site has a local mail-out program available, it can potentially keep the same days’ supply for all prescriptions by mailing these OTC medications locally rather than utilizing CMOP. However, this can partially negate CMOP’s benefit of reduced workload.

CMOP also has specified quantities on some prescription medications. One survey respondent viewed “the quantity and day supply required by CMOP” as a negative influence on the site’s insurance collection. It is possible that CMOP does not carry all the medications that a CMOP-enrolled patient is prescribed. Most sites (77%) still send eligible prescriptions through CMOP for the patients who also have CMOP-ineligible prescriptions. There are a small number of sites (14%) that utilize local mail-out program for the patients with any CMOP-ineligible prescriptions, possibly to simplify the process. Schedule II controlled substances cannot be processed through CMOP either; however, facilities may have local policies that prohibit mailing any controlled substances.

Prescriptions can be manually transmitted to CMOP, or they can be automatically transmitted based on the run time and frequency of the auto-transmission setup. The prescriptions that are waiting to be transmitted to CMOP must be in the “suspended” status. The apparent advantage of relying on auto-transmission is that you do not have to complete the steps manually to transmit suspended CMOP prescriptions, thereby making the process more convenient. However, the manual transmission can be utilized as a checkpoint to verify that prescriptions were properly suspended for CMOP, as the prescription status changes from “S” (suspended) to “AT” once the transmission is completed. If a prescription is not properly suspended for CMOP, the status will remain as S even after manual transmission. More than half (59%) of the responding sites must find the manual transmission feature useful as they use it either over or in addition to the auto-transmission setup.

Despite the challenges, many IHS sites process thousands of monthly prescriptions through CMOP. Of the 94 CMOP-enrolled IHS sites, 17 processed > 1,000 prescriptions from March 27, 2019 to April 25, 2019.¹⁷ Five sites processed > 5,000 prescriptions.¹⁷ At the rate of > 5,000 prescriptions per month, the yearly CMOP prescription count will be > 60,000. That is more than one-third of the prescriptions processed by CHCF in 2018. By handling these prescriptions through CMOP, it can decrease pharmacy filling and dispensing workload, thereby freeing pharmacists to participate in other services.¹⁸ Furthermore, implementing CMOP does not incur any cost for the IHS site. There is a nondrug cost for each prescription that is filled through CMOP. This cost was $2.67 during FY 2016.¹⁹ The fee covers prescription vial, label, packaging for mail, postage, personnel, building overhead, and equipment capitalization.¹⁹ The nondrug cost of filling a prescription locally at the site can potentially exceed the cost charged by CMOP.¹⁹

A lack of objective data exists to assess the net impact of CMOP on patients. Different theoretical assumptions can be made, such as CMOP resulting in better patient adherence. However, there is no objective information about how much CMOP improves patient adherence if it does at all. Though J.D. Power US Pharmacy Study ranks CMOP as “among the best” mail-order pharmacies in customer satisfaction, only 3 of the 44 responding sites have measured patient satisfaction locally.²⁰ Only 1 site had objective data about CMOP’s impact on the point of sale. Therefore, it is currently difficult to perform a cost-benefit analysis of the CMOP program. There are opportunities for further studies on these topics.

Limitations

One limitation of this study is that < 50% of the CMOP-enrolled sites (44 of 94) responded to the questionnaire. It is possible that the facilities that had a significantly positive or negative experience with CMOP were more inclined to share their views. Therefore, it is difficult to conclude whether the responding sites are an accurate representative sample. Another limitation of the study was the questionnaire design and the reliance on free-text responses as opposed to structured data. The free-text responses had to be analyzed manually to determine whether they fall in the same category, thereby increasing the risk of interpretation error.

Conclusion

CMOP has its unique challenges but provides many benefits that local pharmacy mail-out programs may not possess, such as the abilities to mail refrigerated items and track packages. One must be familiar with CMOP’s various idiosyncrasies to make the best use of the program. Extensive staff education and orientation for new staff members must be done to familiarize them with the program. Nevertheless, the successful implementation of CMOP can lead to reduced pharmacy workload while increasing access to care for patients with transportation issues.

Acknowledgments
The authors thank LCDR Karsten Smith, PharmD, BCGP, the IHS CMOP Coordinator for providing the list of primary CMOP contacts and CDR Kendall Van Tyle, PharmD, BCPS, for proofreading the article.

Consolidated mail outpatient pharmacy (CMOP) is an automated prescription order processing and delivery system developed by the US Department of Veterans Affairs (VA) in 1994 to provide medications to VA patients.¹ In fiscal year (FY) 2016, CMOP filled about 80% of VA outpatient prescriptions.²

Formalized by the 2010 Memorandum of Understanding between Indian Health Service (IHS) and VA, CMOP is a partnership undertaken to improve the delivery of care to patients by both agencies.³ The number of prescriptions filled by CMOP for IHS patients increased from 1,972 in FY 2010 to 840,109 in FY 2018.⁴ In the fourth quarter of FY 2018, there were 94 CMOP-enrolled IHS federal and tribal sites.⁵ It is only appropriate that a growing number of IHS sites are adopting CMOP considering the evidence for mail-order pharmacy on better patient adherence, improved health outcomes, and potential cost savings.^6-9 Furthermore, using a centralized pharmacy operation, such as CMOP, can lead to better quality services.¹⁰

Crownpoint Health Care Facility (CHCF) serves > 30,000 American Indians and is in Crownpoint, New Mexico, a small community of about 3,000 people.¹¹ Most of the patients served by the facility live in distant places. Many of these underserved patients do not have a stable means of transportation.¹² Therefore, these patients may have difficulty traveling to the facility for their health care needs, including medication pickups. More than 2.5 million American Indians and Alaska Natives IHS beneficiaries face similar challenges due to the rurality of their communities.¹³ CMOP can be a method to increase access to care for this vulnerable population. However, the utilization of CMOP varies significantly among IHS facilities. While some IHS facilities process large numbers of prescriptions through CMOP, other facilities process few, if any. There also are IHS facilities, such as CHCF, which are at the initial stage of implementing CMOP or trying to increase the volume of prescriptions processed through CMOP. Although the utilization of CMOP has grown exponentially among IHS facilities, there is currently no available resource that summarizes the relative advantages and disadvantages, the challenges and opportunities, and the strengths and weaknesses of implementing CMOP for IHS facilities

Methods

A questionnaire encompassing various aspects of CMOP prescription processing was developed and distributed to the primary CMOP contacts for IHS facilities. The questionnaire was first distributed by e-mail on December 19, 2018. It was e-mailed for a second time on January 16, 2019, and the questionnaire was open for responses until the end of January 2019 (Table).

Results

Forty-four of 94 CMOP-enrolled IHS sites responded to the questionnaire. Most sites train the majority of their pharmacists in CMOP prescription processing. Overall, 310 of 347 pharmacists (89%) in these 44 IHS sites can process prescriptions through CMOP. Thirty-one sites have all their pharmacists trained in CMOP prescription processing. Only 1 facility had less than half (2 of 17 pharmacists) of its pharmacists trained in CMOP prescription processing. More than half the total number of pharmacists, 185 out of 347 (53%), check electronic messages via Resource and Patient Management System (RPMS) MailMan to get information about prescriptions rejected by CMOP. Twenty sites have all their pharmacists check messages about CMOP rejections. However, 2 facilities reported that they do not check the rejection messages at all. Twenty-six of the 44 responding sites (59%) transmit prescriptions to CMOP manually in the electronic system. The rest (18 of 44) rely on the auto-transmission (AT) setup to transmit the CMOP-suspended prescriptions at specified times of the day.

Half the sites (8 of 16) that rely on patients asking for prescriptions to be mailed at the time of refill request do not use any method to designate a CMOP patient. Twenty-four sites use the narrative field on the patient’s profile in RPMS, the health information system used by most IHS facilities, to designate CMOP patients. Eighteen sites use pop-up messages on ScriptPro, a pharmacy automation system, as a designation method. Most of the sites (12 of 15) that use both RPMS and ScriptPro designation methods do not require patients to ask for prescriptions to be mailed at the time of refill request; prescriptions for these patients are routed through CMOP unless patients request otherwise. Only 3 of 44 sites use both methods and rely on patients asking for prescriptions to be mailed at the time of refill request. Some other reported designation methods were using the electronic health record (EHR) posting box, keeping a manual list of CMOP patients, and solely utilizing the Prescription Mail Delivery field in RPMS. Three sites also noted that they keep manual lists to auto-refill prescriptions through CMOP.

Thirty sites (68%) reported that they process every prescription through CMOP even if the patient had prescriptions with specified CMOP quantities. Only 8 sites (18%) said that they used the local mail-out program to keep the same days’ supply for all medication orders. For patients with CMOP-ineligible prescriptions, 34 of the 44 sites (77%) process the eligible prescriptions through CMOP and refill the rest of the prescriptions locally. Six sites (14%) process all medication orders locally for patients with any CMOP-ineligible prescriptions.

Only 12 of 44 sites (27%) involve pharmacy technicians in CMOP prescription processing. Five sites have technicians process prescription refills through CMOP. Two of these sites mentioned the strategy of technicians suspending the prescriptions to be sent to CMOP on the refill due date. Other technician roles included tracking CMOP packages, checking electronic messages for CMOP rejections, and signing up patients for CMOP.

Only 3 of the 44 sites (7%) have measured patient satisfaction with the CMOP program. One of these 3 sites reported that the overall satisfaction was high with CMOP. This site administered the survey to patients who came to the clinic for appointments. The second facility called patients and asked for their feedback. The third site conducted the survey by using student pharmacists. Two sites reported that they use the survey results from the CMOP-conducted patient satisfaction surveys, although they have not measured patient satisfaction at their specific facilities.

Most sites have not assessed CMOP’s impact on their insurance (point of sale) collections. However, 13 sites (30%) reported that they believe they are losing on collections by utilizing CMOP. The use of repackaged products by CMOP, which are usually nonreimbursable, is an issue that was mentioned multiple times. In contrast, 2 sites mentioned that CMOP has led to increased insurance collections for their facilities.

Discussion

The utility of CMOP among the responding IHS sites varies quite significantly. Some sites appreciate the convenience of CMOP while acknowledging its limitations, such as the possible decrease in insurance collections, lengthy prescription processing time, or medication backorders. However, some sites have reserved CMOP for special circumstances (eg, mailing refrigerated items to the patient’s street address) due to various complexities that may come with CMOP. One site reported that it compares IHS contract drug prices with VA contract drug prices quarterly to determine which prescriptions should be sent through CMOP.

Most of the IHS pharmacists (89%) are trained in CMOP prescription processing. If an IHS site wants to increase its volume of CMOP prescriptions, it is sensible to train as many pharmacists as possible so that the responsibility does not fall on a few pharmacists. Newly hired pharmacists can receive guidance from trained pharmacists. Designation methods for CMOP patients can be beneficial for these pharmacists to identify CMOP-enrolled patients, especially if the site does not require patients to ask for prescriptions to be mailed at the time of refill request. Only 3 sites (7%) use multiple designation methods in addition to relying on patients to ask for prescriptions to be mailed. Proper implementation of designation methods can remove this extra burden on patients. Conversely, requiring patients to ask for prescriptions to be sent through CMOP can prevent spontaneous mail-outs if a CMOP-designated patient wants to pick up prescriptions locally. Overall, 16 sites (36%) rely on patients asking for prescriptions to be mailed.

One of the main benefits of CMOP is the ability to mail refrigerated items. Local pharmacy mail-out programs may not have this ability. Patients at rural locations often use post office (PO) boxes because they are unable to receive postal services at their physical addresses; however, they may receive packages through United Parcel Service (UPS) at their physical addresses. CMOP uses UPS to send refrigerated items, but UPS does not deliver to PO boxes. Therefore, remotely located sites like CHCF have difficulty in fully optimizing this benefit. One solution is documenting both the physical and mailing addresses on the patient’s EHR, which enables CMOP to send refrigerated items to the patient’s home address via UPS and mail the rest of the prescriptions to the patient’s PO box address with the US Postal Service. The physical address must be listed above the PO box address to ensure that refrigerated items are not rejected by CMOP. Furthermore, both the physical address and the PO box address must be in the same city for this method to work. Two sites noted mailing refrigerated items as one of the major challenges in CMOP prescription processing.

CMOP-enrolled patients must be educated about requesting medications 7 to 10 days before they run out. There is no standard time line for prescriptions filled by CMOP. However, 1 site reported that it may take up to “10 days from time requested to mailbox.” This delay leads to pharmacies facing a dilemma as processing prescriptions too early can lead to insurance rejections, but processing them too late can lead to the patient not receiving the medication by the time they run out of their current supply. However, CMOP provides the ability to track prescriptions sent through CMOP. Pharmacists and technicians need to have access to BestWay Parcel Services Client Portal (genco-mms.bestwayparcel.com) to track CMOP packages. Tracking CMOP prescriptions is a way pharmacy technicians can be involved in CMOP prescription processing. Technicians seem to be underutilized, as only 27% of the responding sites utilize them to some degree in the CMOP process. One site delegated the responsibility of checking CMOP rejection messages to pharmacy technicians. Since 2 of the responding sites do not check CMOP rejection messages at all, this is an excellent opportunity to get pharmacy technicians involved.

A CMOP auto-refill program can potentially be utilized to avoid missed or late medications. In an auto-refill program, a pharmacist can refill prescriptions through CMOP on the due date without a patient request. They may get rejected by insurance the first time they are processed through CMOP for refilling too early if the processing time is taken into account. However, the subsequent refills do not have to consider the CMOP processing time as they would already be synchronized based on the last refill date. Though, if CMOP is out of stock on a medication and it is expected to be available soon, CMOP may take a few extra days to either fill the prescription or reject it if the drug stays unavailable. One of the sites reported “the amount of time [CMOP] holds medications if they are out of stock” as “the hardest thing to work around.” A couple of sites also mentioned the longer than usual delay in processing prescriptions by CMOP during the holidays as one of the major challenges.

CMOP use of repackaged products also may lead insurance companies to deny reimbursement. Repackaged products are usually cheaper to buy.¹⁴ However, most insurances do not reimburse for prescriptions filled with these products.¹⁵ The local drug file on RPMS may have a national drug code (NDC) that is reimbursable by insurance, but CMOP will change it to the repackaged NDC if they are filling the prescription with a repackaged product. One potential solution to this problem would be filling these prescriptions locally. Furthermore, insurance claims are processed when the prescriptions are filled by CMOP. Sites cannot return/cancel the prescription anymore at that point. Therefore, the inability to see real-time rejections as the medication orders are processed on-site makes it challenging to prevent avoidable insurance rejections, such as a refill too soon. One site calculated that it lost $26,386.45 by utilizing CMOP from January 9, 2018 to December 12, 2018. However, it is unclear whether this loss was representative of other sites. It is also worth noting that IHS sites can save a substantial amount of money on certain products by utilizing CMOP because VA buys these products at a reduced price.¹⁶

CMOP-transmitted prescriptions can be rejected for various reasons, such as CMOP manufacturer’s backorder, a different quantity from CMOP stock size, etc. Information about these rejected prescriptions is accessed through electronic messages on RPMS. CMOP does not dispense less than a full, unopened package for most over-the-counter (OTC) medications. The quantity on these prescriptions must be equal to or multiples of the package size for them to be filled by CMOP. This can lead to a patient having prescriptions with different days’ supplies, which results in various refill due dates. If a site has a local mail-out program available, it can potentially keep the same days’ supply for all prescriptions by mailing these OTC medications locally rather than utilizing CMOP. However, this can partially negate CMOP’s benefit of reduced workload.

CMOP also has specified quantities on some prescription medications. One survey respondent viewed “the quantity and day supply required by CMOP” as a negative influence on the site’s insurance collection. It is possible that CMOP does not carry all the medications that a CMOP-enrolled patient is prescribed. Most sites (77%) still send eligible prescriptions through CMOP for the patients who also have CMOP-ineligible prescriptions. There are a small number of sites (14%) that utilize local mail-out program for the patients with any CMOP-ineligible prescriptions, possibly to simplify the process. Schedule II controlled substances cannot be processed through CMOP either; however, facilities may have local policies that prohibit mailing any controlled substances.

Prescriptions can be manually transmitted to CMOP, or they can be automatically transmitted based on the run time and frequency of the auto-transmission setup. The prescriptions that are waiting to be transmitted to CMOP must be in the “suspended” status. The apparent advantage of relying on auto-transmission is that you do not have to complete the steps manually to transmit suspended CMOP prescriptions, thereby making the process more convenient. However, the manual transmission can be utilized as a checkpoint to verify that prescriptions were properly suspended for CMOP, as the prescription status changes from “S” (suspended) to “AT” once the transmission is completed. If a prescription is not properly suspended for CMOP, the status will remain as S even after manual transmission. More than half (59%) of the responding sites must find the manual transmission feature useful as they use it either over or in addition to the auto-transmission setup.

Despite the challenges, many IHS sites process thousands of monthly prescriptions through CMOP. Of the 94 CMOP-enrolled IHS sites, 17 processed > 1,000 prescriptions from March 27, 2019 to April 25, 2019.¹⁷ Five sites processed > 5,000 prescriptions.¹⁷ At the rate of > 5,000 prescriptions per month, the yearly CMOP prescription count will be > 60,000. That is more than one-third of the prescriptions processed by CHCF in 2018. By handling these prescriptions through CMOP, it can decrease pharmacy filling and dispensing workload, thereby freeing pharmacists to participate in other services.¹⁸ Furthermore, implementing CMOP does not incur any cost for the IHS site. There is a nondrug cost for each prescription that is filled through CMOP. This cost was $2.67 during FY 2016.¹⁹ The fee covers prescription vial, label, packaging for mail, postage, personnel, building overhead, and equipment capitalization.¹⁹ The nondrug cost of filling a prescription locally at the site can potentially exceed the cost charged by CMOP.¹⁹

A lack of objective data exists to assess the net impact of CMOP on patients. Different theoretical assumptions can be made, such as CMOP resulting in better patient adherence. However, there is no objective information about how much CMOP improves patient adherence if it does at all. Though J.D. Power US Pharmacy Study ranks CMOP as “among the best” mail-order pharmacies in customer satisfaction, only 3 of the 44 responding sites have measured patient satisfaction locally.²⁰ Only 1 site had objective data about CMOP’s impact on the point of sale. Therefore, it is currently difficult to perform a cost-benefit analysis of the CMOP program. There are opportunities for further studies on these topics.

Limitations

One limitation of this study is that < 50% of the CMOP-enrolled sites (44 of 94) responded to the questionnaire. It is possible that the facilities that had a significantly positive or negative experience with CMOP were more inclined to share their views. Therefore, it is difficult to conclude whether the responding sites are an accurate representative sample. Another limitation of the study was the questionnaire design and the reliance on free-text responses as opposed to structured data. The free-text responses had to be analyzed manually to determine whether they fall in the same category, thereby increasing the risk of interpretation error.

Conclusion

CMOP has its unique challenges but provides many benefits that local pharmacy mail-out programs may not possess, such as the abilities to mail refrigerated items and track packages. One must be familiar with CMOP’s various idiosyncrasies to make the best use of the program. Extensive staff education and orientation for new staff members must be done to familiarize them with the program. Nevertheless, the successful implementation of CMOP can lead to reduced pharmacy workload while increasing access to care for patients with transportation issues.

Acknowledgments
The authors thank LCDR Karsten Smith, PharmD, BCGP, the IHS CMOP Coordinator for providing the list of primary CMOP contacts and CDR Kendall Van Tyle, PharmD, BCPS, for proofreading the article.

Consolidated mail outpatient pharmacy (CMOP) is an automated prescription order processing and delivery system developed by the US Department of Veterans Affairs (VA) in 1994 to provide medications to VA patients.¹ In fiscal year (FY) 2016, CMOP filled about 80% of VA outpatient prescriptions.²

Formalized by the 2010 Memorandum of Understanding between Indian Health Service (IHS) and VA, CMOP is a partnership undertaken to improve the delivery of care to patients by both agencies.³ The number of prescriptions filled by CMOP for IHS patients increased from 1,972 in FY 2010 to 840,109 in FY 2018.⁴ In the fourth quarter of FY 2018, there were 94 CMOP-enrolled IHS federal and tribal sites.⁵ It is only appropriate that a growing number of IHS sites are adopting CMOP considering the evidence for mail-order pharmacy on better patient adherence, improved health outcomes, and potential cost savings.^6-9 Furthermore, using a centralized pharmacy operation, such as CMOP, can lead to better quality services.¹⁰

Crownpoint Health Care Facility (CHCF) serves > 30,000 American Indians and is in Crownpoint, New Mexico, a small community of about 3,000 people.¹¹ Most of the patients served by the facility live in distant places. Many of these underserved patients do not have a stable means of transportation.¹² Therefore, these patients may have difficulty traveling to the facility for their health care needs, including medication pickups. More than 2.5 million American Indians and Alaska Natives IHS beneficiaries face similar challenges due to the rurality of their communities.¹³ CMOP can be a method to increase access to care for this vulnerable population. However, the utilization of CMOP varies significantly among IHS facilities. While some IHS facilities process large numbers of prescriptions through CMOP, other facilities process few, if any. There also are IHS facilities, such as CHCF, which are at the initial stage of implementing CMOP or trying to increase the volume of prescriptions processed through CMOP. Although the utilization of CMOP has grown exponentially among IHS facilities, there is currently no available resource that summarizes the relative advantages and disadvantages, the challenges and opportunities, and the strengths and weaknesses of implementing CMOP for IHS facilities

Methods

A questionnaire encompassing various aspects of CMOP prescription processing was developed and distributed to the primary CMOP contacts for IHS facilities. The questionnaire was first distributed by e-mail on December 19, 2018. It was e-mailed for a second time on January 16, 2019, and the questionnaire was open for responses until the end of January 2019 (Table).

Results

Forty-four of 94 CMOP-enrolled IHS sites responded to the questionnaire. Most sites train the majority of their pharmacists in CMOP prescription processing. Overall, 310 of 347 pharmacists (89%) in these 44 IHS sites can process prescriptions through CMOP. Thirty-one sites have all their pharmacists trained in CMOP prescription processing. Only 1 facility had less than half (2 of 17 pharmacists) of its pharmacists trained in CMOP prescription processing. More than half the total number of pharmacists, 185 out of 347 (53%), check electronic messages via Resource and Patient Management System (RPMS) MailMan to get information about prescriptions rejected by CMOP. Twenty sites have all their pharmacists check messages about CMOP rejections. However, 2 facilities reported that they do not check the rejection messages at all. Twenty-six of the 44 responding sites (59%) transmit prescriptions to CMOP manually in the electronic system. The rest (18 of 44) rely on the auto-transmission (AT) setup to transmit the CMOP-suspended prescriptions at specified times of the day.

Half the sites (8 of 16) that rely on patients asking for prescriptions to be mailed at the time of refill request do not use any method to designate a CMOP patient. Twenty-four sites use the narrative field on the patient’s profile in RPMS, the health information system used by most IHS facilities, to designate CMOP patients. Eighteen sites use pop-up messages on ScriptPro, a pharmacy automation system, as a designation method. Most of the sites (12 of 15) that use both RPMS and ScriptPro designation methods do not require patients to ask for prescriptions to be mailed at the time of refill request; prescriptions for these patients are routed through CMOP unless patients request otherwise. Only 3 of 44 sites use both methods and rely on patients asking for prescriptions to be mailed at the time of refill request. Some other reported designation methods were using the electronic health record (EHR) posting box, keeping a manual list of CMOP patients, and solely utilizing the Prescription Mail Delivery field in RPMS. Three sites also noted that they keep manual lists to auto-refill prescriptions through CMOP.

Thirty sites (68%) reported that they process every prescription through CMOP even if the patient had prescriptions with specified CMOP quantities. Only 8 sites (18%) said that they used the local mail-out program to keep the same days’ supply for all medication orders. For patients with CMOP-ineligible prescriptions, 34 of the 44 sites (77%) process the eligible prescriptions through CMOP and refill the rest of the prescriptions locally. Six sites (14%) process all medication orders locally for patients with any CMOP-ineligible prescriptions.

Only 12 of 44 sites (27%) involve pharmacy technicians in CMOP prescription processing. Five sites have technicians process prescription refills through CMOP. Two of these sites mentioned the strategy of technicians suspending the prescriptions to be sent to CMOP on the refill due date. Other technician roles included tracking CMOP packages, checking electronic messages for CMOP rejections, and signing up patients for CMOP.

Only 3 of the 44 sites (7%) have measured patient satisfaction with the CMOP program. One of these 3 sites reported that the overall satisfaction was high with CMOP. This site administered the survey to patients who came to the clinic for appointments. The second facility called patients and asked for their feedback. The third site conducted the survey by using student pharmacists. Two sites reported that they use the survey results from the CMOP-conducted patient satisfaction surveys, although they have not measured patient satisfaction at their specific facilities.

Most sites have not assessed CMOP’s impact on their insurance (point of sale) collections. However, 13 sites (30%) reported that they believe they are losing on collections by utilizing CMOP. The use of repackaged products by CMOP, which are usually nonreimbursable, is an issue that was mentioned multiple times. In contrast, 2 sites mentioned that CMOP has led to increased insurance collections for their facilities.

Discussion

The utility of CMOP among the responding IHS sites varies quite significantly. Some sites appreciate the convenience of CMOP while acknowledging its limitations, such as the possible decrease in insurance collections, lengthy prescription processing time, or medication backorders. However, some sites have reserved CMOP for special circumstances (eg, mailing refrigerated items to the patient’s street address) due to various complexities that may come with CMOP. One site reported that it compares IHS contract drug prices with VA contract drug prices quarterly to determine which prescriptions should be sent through CMOP.

Most of the IHS pharmacists (89%) are trained in CMOP prescription processing. If an IHS site wants to increase its volume of CMOP prescriptions, it is sensible to train as many pharmacists as possible so that the responsibility does not fall on a few pharmacists. Newly hired pharmacists can receive guidance from trained pharmacists. Designation methods for CMOP patients can be beneficial for these pharmacists to identify CMOP-enrolled patients, especially if the site does not require patients to ask for prescriptions to be mailed at the time of refill request. Only 3 sites (7%) use multiple designation methods in addition to relying on patients to ask for prescriptions to be mailed. Proper implementation of designation methods can remove this extra burden on patients. Conversely, requiring patients to ask for prescriptions to be sent through CMOP can prevent spontaneous mail-outs if a CMOP-designated patient wants to pick up prescriptions locally. Overall, 16 sites (36%) rely on patients asking for prescriptions to be mailed.

One of the main benefits of CMOP is the ability to mail refrigerated items. Local pharmacy mail-out programs may not have this ability. Patients at rural locations often use post office (PO) boxes because they are unable to receive postal services at their physical addresses; however, they may receive packages through United Parcel Service (UPS) at their physical addresses. CMOP uses UPS to send refrigerated items, but UPS does not deliver to PO boxes. Therefore, remotely located sites like CHCF have difficulty in fully optimizing this benefit. One solution is documenting both the physical and mailing addresses on the patient’s EHR, which enables CMOP to send refrigerated items to the patient’s home address via UPS and mail the rest of the prescriptions to the patient’s PO box address with the US Postal Service. The physical address must be listed above the PO box address to ensure that refrigerated items are not rejected by CMOP. Furthermore, both the physical address and the PO box address must be in the same city for this method to work. Two sites noted mailing refrigerated items as one of the major challenges in CMOP prescription processing.

CMOP-enrolled patients must be educated about requesting medications 7 to 10 days before they run out. There is no standard time line for prescriptions filled by CMOP. However, 1 site reported that it may take up to “10 days from time requested to mailbox.” This delay leads to pharmacies facing a dilemma as processing prescriptions too early can lead to insurance rejections, but processing them too late can lead to the patient not receiving the medication by the time they run out of their current supply. However, CMOP provides the ability to track prescriptions sent through CMOP. Pharmacists and technicians need to have access to BestWay Parcel Services Client Portal (genco-mms.bestwayparcel.com) to track CMOP packages. Tracking CMOP prescriptions is a way pharmacy technicians can be involved in CMOP prescription processing. Technicians seem to be underutilized, as only 27% of the responding sites utilize them to some degree in the CMOP process. One site delegated the responsibility of checking CMOP rejection messages to pharmacy technicians. Since 2 of the responding sites do not check CMOP rejection messages at all, this is an excellent opportunity to get pharmacy technicians involved.

A CMOP auto-refill program can potentially be utilized to avoid missed or late medications. In an auto-refill program, a pharmacist can refill prescriptions through CMOP on the due date without a patient request. They may get rejected by insurance the first time they are processed through CMOP for refilling too early if the processing time is taken into account. However, the subsequent refills do not have to consider the CMOP processing time as they would already be synchronized based on the last refill date. Though, if CMOP is out of stock on a medication and it is expected to be available soon, CMOP may take a few extra days to either fill the prescription or reject it if the drug stays unavailable. One of the sites reported “the amount of time [CMOP] holds medications if they are out of stock” as “the hardest thing to work around.” A couple of sites also mentioned the longer than usual delay in processing prescriptions by CMOP during the holidays as one of the major challenges.

CMOP use of repackaged products also may lead insurance companies to deny reimbursement. Repackaged products are usually cheaper to buy.¹⁴ However, most insurances do not reimburse for prescriptions filled with these products.¹⁵ The local drug file on RPMS may have a national drug code (NDC) that is reimbursable by insurance, but CMOP will change it to the repackaged NDC if they are filling the prescription with a repackaged product. One potential solution to this problem would be filling these prescriptions locally. Furthermore, insurance claims are processed when the prescriptions are filled by CMOP. Sites cannot return/cancel the prescription anymore at that point. Therefore, the inability to see real-time rejections as the medication orders are processed on-site makes it challenging to prevent avoidable insurance rejections, such as a refill too soon. One site calculated that it lost $26,386.45 by utilizing CMOP from January 9, 2018 to December 12, 2018. However, it is unclear whether this loss was representative of other sites. It is also worth noting that IHS sites can save a substantial amount of money on certain products by utilizing CMOP because VA buys these products at a reduced price.¹⁶

CMOP-transmitted prescriptions can be rejected for various reasons, such as CMOP manufacturer’s backorder, a different quantity from CMOP stock size, etc. Information about these rejected prescriptions is accessed through electronic messages on RPMS. CMOP does not dispense less than a full, unopened package for most over-the-counter (OTC) medications. The quantity on these prescriptions must be equal to or multiples of the package size for them to be filled by CMOP. This can lead to a patient having prescriptions with different days’ supplies, which results in various refill due dates. If a site has a local mail-out program available, it can potentially keep the same days’ supply for all prescriptions by mailing these OTC medications locally rather than utilizing CMOP. However, this can partially negate CMOP’s benefit of reduced workload.

CMOP also has specified quantities on some prescription medications. One survey respondent viewed “the quantity and day supply required by CMOP” as a negative influence on the site’s insurance collection. It is possible that CMOP does not carry all the medications that a CMOP-enrolled patient is prescribed. Most sites (77%) still send eligible prescriptions through CMOP for the patients who also have CMOP-ineligible prescriptions. There are a small number of sites (14%) that utilize local mail-out program for the patients with any CMOP-ineligible prescriptions, possibly to simplify the process. Schedule II controlled substances cannot be processed through CMOP either; however, facilities may have local policies that prohibit mailing any controlled substances.

Prescriptions can be manually transmitted to CMOP, or they can be automatically transmitted based on the run time and frequency of the auto-transmission setup. The prescriptions that are waiting to be transmitted to CMOP must be in the “suspended” status. The apparent advantage of relying on auto-transmission is that you do not have to complete the steps manually to transmit suspended CMOP prescriptions, thereby making the process more convenient. However, the manual transmission can be utilized as a checkpoint to verify that prescriptions were properly suspended for CMOP, as the prescription status changes from “S” (suspended) to “AT” once the transmission is completed. If a prescription is not properly suspended for CMOP, the status will remain as S even after manual transmission. More than half (59%) of the responding sites must find the manual transmission feature useful as they use it either over or in addition to the auto-transmission setup.

Despite the challenges, many IHS sites process thousands of monthly prescriptions through CMOP. Of the 94 CMOP-enrolled IHS sites, 17 processed > 1,000 prescriptions from March 27, 2019 to April 25, 2019.¹⁷ Five sites processed > 5,000 prescriptions.¹⁷ At the rate of > 5,000 prescriptions per month, the yearly CMOP prescription count will be > 60,000. That is more than one-third of the prescriptions processed by CHCF in 2018. By handling these prescriptions through CMOP, it can decrease pharmacy filling and dispensing workload, thereby freeing pharmacists to participate in other services.¹⁸ Furthermore, implementing CMOP does not incur any cost for the IHS site. There is a nondrug cost for each prescription that is filled through CMOP. This cost was $2.67 during FY 2016.¹⁹ The fee covers prescription vial, label, packaging for mail, postage, personnel, building overhead, and equipment capitalization.¹⁹ The nondrug cost of filling a prescription locally at the site can potentially exceed the cost charged by CMOP.¹⁹

A lack of objective data exists to assess the net impact of CMOP on patients. Different theoretical assumptions can be made, such as CMOP resulting in better patient adherence. However, there is no objective information about how much CMOP improves patient adherence if it does at all. Though J.D. Power US Pharmacy Study ranks CMOP as “among the best” mail-order pharmacies in customer satisfaction, only 3 of the 44 responding sites have measured patient satisfaction locally.²⁰ Only 1 site had objective data about CMOP’s impact on the point of sale. Therefore, it is currently difficult to perform a cost-benefit analysis of the CMOP program. There are opportunities for further studies on these topics.

Limitations

One limitation of this study is that < 50% of the CMOP-enrolled sites (44 of 94) responded to the questionnaire. It is possible that the facilities that had a significantly positive or negative experience with CMOP were more inclined to share their views. Therefore, it is difficult to conclude whether the responding sites are an accurate representative sample. Another limitation of the study was the questionnaire design and the reliance on free-text responses as opposed to structured data. The free-text responses had to be analyzed manually to determine whether they fall in the same category, thereby increasing the risk of interpretation error.

Conclusion

CMOP has its unique challenges but provides many benefits that local pharmacy mail-out programs may not possess, such as the abilities to mail refrigerated items and track packages. One must be familiar with CMOP’s various idiosyncrasies to make the best use of the program. Extensive staff education and orientation for new staff members must be done to familiarize them with the program. Nevertheless, the successful implementation of CMOP can lead to reduced pharmacy workload while increasing access to care for patients with transportation issues.

Acknowledgments
The authors thank LCDR Karsten Smith, PharmD, BCGP, the IHS CMOP Coordinator for providing the list of primary CMOP contacts and CDR Kendall Van Tyle, PharmD, BCPS, for proofreading the article.

An estimated 3.5 million people in the US have chronic hepatitis C virus (HCV) infection, and between 10% and 20% of those developed cirrhosis over 20 to 30 years.¹ There are at least 6 genotypes (GTs) of HCV, with GT1 being the most common in the US and previously one of the most difficult to treat.^2,3 The goal of treatment is to achieve viral cure, called sustained virologic response (SVR) when HCV viral load remains undetectable several weeks after therapy completion. In the 2000s, pegylated interferon (pegIFN) and ribavirin (RBV) were the standard of care.² For patients with GT1 infections, an SVR of 40 to 50% was commonly seen after 48 weeks of pegIFN/RBV regimens compared with 70 to 80% SVR for GT2 or GT3 after 24 weeks of pegIFN/RBV therapy.² However, treatment has evolved rapidly (Table 1).^2-17

In 2011, the US Food and Drug Administration (FDA) approved the protease inhibitors (PIs) boceprevir and telaprevir, which added a new class of agents with increased SVR for patients with GT1 infection; however, pegIFN and RBV were still needed for treatment.⁴ In addition, both PIs required multiple doses per day and strict adherence to an 8-hour schedule.⁴ Boceprevir required treatment with RBV and pegIFN for 48 weeks unless futility rule was met at 24 weeks of treatment (ie, viral load still detectable).⁴ The SVR in patients with GT1 infection improved to > 65% for patients in clinical trials.² FDA approval of the direct-acting antivirals (DAAs) sofosbuvir and simeprevir in late 2013 decreased the usual duration of therapy to only 12 weeks with improved SVR rates 12 weeks posttherapy (SVR12) to 90% or higher.^2,6,10

FDA approval of ledipasvir (LDV)/sofosbuvir (SOF) in October 2014 resulted in the first interferon-free all-oral regimen indicated for HCV GT1 infection.¹¹ In December 2014, FDA approved a combination of paritaprevir, ritonavir, ombitasvir, and dasabuvir (PrOD).¹² In 2015 GT-specific approvals were issued for daclastavir to be used with SOV for GT1 and GT3 and a combination similar to PrOD without dasabuvir (PrO) for GT4.¹³ In 2016, a combination of elbasvir (ERB) and grazoprevir (GZP) was approved for GT1 and GT4.¹⁴

In 2016, a pangenotypic DAA of SOF and velpatasvir (VEL) was approved.¹⁵ Most recently, combinations of SOF, VEL, and voxilaprevir (VOX), and glecaprevir (GLE) and pibrentasvir (PIB) were approved for patients with previous DAA treatment failures.^{7, 8,16,17} These oral regimens avoided the significant adverse events (AEs) associated with pegIFN and RBV (eg, thrombocytopenia, depression), were expected to improve treatment adherence and shorten duration of therapy.

The West Palm Beach Veterans Affairs has had a nurse practitioner (NP)-based HCV treatment clinic since the late 1990s. When PIs became available, a CPS started reviewing patient electronic health records (EHRs) and monitored response to therapy along with the NP to ensure discontinuation of therapy if futility criteria were met.⁷ Our unpublished experience showed SVR > 60% with both boceprevir and SOF regimens and > 90% with oral DAA regimens.

This review will provide the SVR rates for patients that needed retreatment for HCV infection since 2015 until December 2019. We treated all willing patients, beginning with the patients who had experienced failures with previous regimens. Patients first received education on HCV infection and treatment options in a group class then they were seen by the NP individually for specific education on treatment. The CPS reviewed the patient’s medical record to assess for appropriate therapy, possible drug-drug interactions and contraindications to therapy. In addition, patient outcomes (eg, viral load, AEs) were documented by the CPS in collaboration with the NP throughout treatment until viral load for SVR evaluation was obtained.

Methods

A retrospective EHR review of patients retreated from January 2015 to December 2019 was conducted. Data collected included age, sex, HCV GT, previous therapy, new medications prescribed, creatinine clearance, and achievement of SVR12. This retrospective review was approved by the facility’s scientific advisory committee as part of performance improvement efforts. Descriptive statistics are provided.

Results

Boceprevir

We treated 31 patients with boceprevir of which 3 met futility rule and 28 completed therapy. Eighteen of 28 responded (64%) to the treatment. The 10 patients who failed treatment were retreated with LDV/SOF, and all achieved SVR.

Sofosbuvir

A total of 53 patients were treated with SOF, RBV, and pegIFN for 12 weeks. Forty-one achieved SVR (77%). Of the 12 who failed therapy, all have been retreated and achieved SVR (Table 2).

Interferon-Free DAA Oral Regimens

More than 900 patients have been treated with interferon-free regimens since 2015 and outcomes were documented for > 800 patients. The SVR rates by GT were as follows: GT1 639 of 676 (95%); GT2 76 of 79 (96%); GT3 40 of 48 (83%); and GT4 6 of 6 (100%). Eighty-four percent of patients had GT1 infection. The median age of patient was 62 years, 72% were treatment naïve, and 35% having cirrhosis (based on liver biopsy or FIB4 score).¹⁸

Of 48 treatment failures, 30 patients were retreated; the rest of the patients were lost to follow-up (n = 9) or unable to receive retreatment (n = 9) mainly due to decompensated cirrhosis or liver cancer and short life expectancy. The median age of patient in this retreatment group was 62 years, 62% had cirrhosis, and most were infected with GT1. The average creatinine clearance was 73 mL/min. Twenty-two patients who failed therapy with ledipasvir/SOF were retreated (Table 3). A total of 13 patients out of the 19 tested eventually achieved SVR (68%). Four of the patients who had treatment failure again had GT1 infection and the other 2 GT3. All had cirrhosis.

Thirty-five patients were treated with PrOD, and 32 achieved SVR (91%). All 3 patients were retreated. One patient each achieved SVR with ERB/GZP, SOF/VEL and SOF/VEL/VOX. Fifty patients were treated with ERB/GZP and 45 achieved SVR (90%). All 5 treatment failures were retreated. Four achieved SVR and 1 was lost to follow-up (Table 4). Overall, of 30 patients who were retreated after failure with an all-oral DAA regimen, 27 patients had SVR values available and 21 achieved it (78%).

Discussion

Overall SVR was very high for patients who received oral treatment for HCV infection. A low number of patients failed therapy and were retreated. Patients who failed therapy again were similar in age but were more likely to have cirrhosis when compared with the overall interferon-free treated group. Thus, prompt treatment after HCV detection and before disease progression may improve treatment outcomes. Achieving SVR has been shown to improve fibrosis, portal hypertension, splenomegaly and cirrhosis, and reduce the risk of hepatocellular carcinoma by 70% and liver-related mortality by 90%.^19-21/

Patients who failed therapy primarily had GT1—the most prevalent GT treated. A higher prevalence of GT1 is expected since it is the most common GT in the US.⁶ However, disease progression occurs more rapidly in those with GT3 and is more difficult to treat.²² The overall response rate was lower with this GT (83%) in this report, with only 1 of 3 patients retreated achieving an SVR.

Similar results are documented in retreatment trials.²³ In the POLARIS-1 trial, treatment with SOF/VEL/VOX resulted in an overall response rate of 96% but only 91% for patients with GT3, compared with 95 to 100% for GTs 1, 2, or 4.²³ In the current report, only 1 patient (GT1) failed retreatment with SOF/VEL/VOX. At this time, there are no clear treatment options for this patient. However, patients who fail GLE/PIB (none so far in the current report) may be able to receive SOF/VEL/VOX.²⁴ In a small study, 29 of 31 patients achieved SVR with SOF/VEL/VOX after GLE/PIB failure (12 of 13 GT1 and 17 of 18 GT3).²⁴

Limitations

This review was an observational, nonrandomized design, and only 1 medical center was involved. These results may not be applicable to other patient populations without a clinic set up with routine follow-ups to encourage adherence and completion of therapy.

Conclusions

Treatment of HCV infection has improved significantly over the past 10 years. Use of DAAs results in SVR for > 90% of patients, especially if the disease had not progressed to cirrhosis. Failure after retreatment for HCV infection was rare as well. Given that cirrhosis seems to increase the chance of treatment failure, it is imperative to identify candidates for treatment before the infection has progressed to cirrhosis. Patients infected with GT3 in particular should be more aggressively identified and treated.

Acknowledgments
The authors thank Nick P. Becky, PharmD, for his contributions to the identification of patients needing treatment for their HCV infection and review of initial manuscript information.

An estimated 3.5 million people in the US have chronic hepatitis C virus (HCV) infection, and between 10% and 20% of those developed cirrhosis over 20 to 30 years.¹ There are at least 6 genotypes (GTs) of HCV, with GT1 being the most common in the US and previously one of the most difficult to treat.^2,3 The goal of treatment is to achieve viral cure, called sustained virologic response (SVR) when HCV viral load remains undetectable several weeks after therapy completion. In the 2000s, pegylated interferon (pegIFN) and ribavirin (RBV) were the standard of care.² For patients with GT1 infections, an SVR of 40 to 50% was commonly seen after 48 weeks of pegIFN/RBV regimens compared with 70 to 80% SVR for GT2 or GT3 after 24 weeks of pegIFN/RBV therapy.² However, treatment has evolved rapidly (Table 1).^2-17

In 2011, the US Food and Drug Administration (FDA) approved the protease inhibitors (PIs) boceprevir and telaprevir, which added a new class of agents with increased SVR for patients with GT1 infection; however, pegIFN and RBV were still needed for treatment.⁴ In addition, both PIs required multiple doses per day and strict adherence to an 8-hour schedule.⁴ Boceprevir required treatment with RBV and pegIFN for 48 weeks unless futility rule was met at 24 weeks of treatment (ie, viral load still detectable).⁴ The SVR in patients with GT1 infection improved to > 65% for patients in clinical trials.² FDA approval of the direct-acting antivirals (DAAs) sofosbuvir and simeprevir in late 2013 decreased the usual duration of therapy to only 12 weeks with improved SVR rates 12 weeks posttherapy (SVR12) to 90% or higher.^2,6,10

FDA approval of ledipasvir (LDV)/sofosbuvir (SOF) in October 2014 resulted in the first interferon-free all-oral regimen indicated for HCV GT1 infection.¹¹ In December 2014, FDA approved a combination of paritaprevir, ritonavir, ombitasvir, and dasabuvir (PrOD).¹² In 2015 GT-specific approvals were issued for daclastavir to be used with SOV for GT1 and GT3 and a combination similar to PrOD without dasabuvir (PrO) for GT4.¹³ In 2016, a combination of elbasvir (ERB) and grazoprevir (GZP) was approved for GT1 and GT4.¹⁴

In 2016, a pangenotypic DAA of SOF and velpatasvir (VEL) was approved.¹⁵ Most recently, combinations of SOF, VEL, and voxilaprevir (VOX), and glecaprevir (GLE) and pibrentasvir (PIB) were approved for patients with previous DAA treatment failures.^{7, 8,16,17} These oral regimens avoided the significant adverse events (AEs) associated with pegIFN and RBV (eg, thrombocytopenia, depression), were expected to improve treatment adherence and shorten duration of therapy.

The West Palm Beach Veterans Affairs has had a nurse practitioner (NP)-based HCV treatment clinic since the late 1990s. When PIs became available, a CPS started reviewing patient electronic health records (EHRs) and monitored response to therapy along with the NP to ensure discontinuation of therapy if futility criteria were met.⁷ Our unpublished experience showed SVR > 60% with both boceprevir and SOF regimens and > 90% with oral DAA regimens.

This review will provide the SVR rates for patients that needed retreatment for HCV infection since 2015 until December 2019. We treated all willing patients, beginning with the patients who had experienced failures with previous regimens. Patients first received education on HCV infection and treatment options in a group class then they were seen by the NP individually for specific education on treatment. The CPS reviewed the patient’s medical record to assess for appropriate therapy, possible drug-drug interactions and contraindications to therapy. In addition, patient outcomes (eg, viral load, AEs) were documented by the CPS in collaboration with the NP throughout treatment until viral load for SVR evaluation was obtained.

Methods

A retrospective EHR review of patients retreated from January 2015 to December 2019 was conducted. Data collected included age, sex, HCV GT, previous therapy, new medications prescribed, creatinine clearance, and achievement of SVR12. This retrospective review was approved by the facility’s scientific advisory committee as part of performance improvement efforts. Descriptive statistics are provided.

Results

Boceprevir

We treated 31 patients with boceprevir of which 3 met futility rule and 28 completed therapy. Eighteen of 28 responded (64%) to the treatment. The 10 patients who failed treatment were retreated with LDV/SOF, and all achieved SVR.

Sofosbuvir

A total of 53 patients were treated with SOF, RBV, and pegIFN for 12 weeks. Forty-one achieved SVR (77%). Of the 12 who failed therapy, all have been retreated and achieved SVR (Table 2).

Interferon-Free DAA Oral Regimens

More than 900 patients have been treated with interferon-free regimens since 2015 and outcomes were documented for > 800 patients. The SVR rates by GT were as follows: GT1 639 of 676 (95%); GT2 76 of 79 (96%); GT3 40 of 48 (83%); and GT4 6 of 6 (100%). Eighty-four percent of patients had GT1 infection. The median age of patient was 62 years, 72% were treatment naïve, and 35% having cirrhosis (based on liver biopsy or FIB4 score).¹⁸

Of 48 treatment failures, 30 patients were retreated; the rest of the patients were lost to follow-up (n = 9) or unable to receive retreatment (n = 9) mainly due to decompensated cirrhosis or liver cancer and short life expectancy. The median age of patient in this retreatment group was 62 years, 62% had cirrhosis, and most were infected with GT1. The average creatinine clearance was 73 mL/min. Twenty-two patients who failed therapy with ledipasvir/SOF were retreated (Table 3). A total of 13 patients out of the 19 tested eventually achieved SVR (68%). Four of the patients who had treatment failure again had GT1 infection and the other 2 GT3. All had cirrhosis.

Thirty-five patients were treated with PrOD, and 32 achieved SVR (91%). All 3 patients were retreated. One patient each achieved SVR with ERB/GZP, SOF/VEL and SOF/VEL/VOX. Fifty patients were treated with ERB/GZP and 45 achieved SVR (90%). All 5 treatment failures were retreated. Four achieved SVR and 1 was lost to follow-up (Table 4). Overall, of 30 patients who were retreated after failure with an all-oral DAA regimen, 27 patients had SVR values available and 21 achieved it (78%).

Discussion

Overall SVR was very high for patients who received oral treatment for HCV infection. A low number of patients failed therapy and were retreated. Patients who failed therapy again were similar in age but were more likely to have cirrhosis when compared with the overall interferon-free treated group. Thus, prompt treatment after HCV detection and before disease progression may improve treatment outcomes. Achieving SVR has been shown to improve fibrosis, portal hypertension, splenomegaly and cirrhosis, and reduce the risk of hepatocellular carcinoma by 70% and liver-related mortality by 90%.^19-21/

Patients who failed therapy primarily had GT1—the most prevalent GT treated. A higher prevalence of GT1 is expected since it is the most common GT in the US.⁶ However, disease progression occurs more rapidly in those with GT3 and is more difficult to treat.²² The overall response rate was lower with this GT (83%) in this report, with only 1 of 3 patients retreated achieving an SVR.

Similar results are documented in retreatment trials.²³ In the POLARIS-1 trial, treatment with SOF/VEL/VOX resulted in an overall response rate of 96% but only 91% for patients with GT3, compared with 95 to 100% for GTs 1, 2, or 4.²³ In the current report, only 1 patient (GT1) failed retreatment with SOF/VEL/VOX. At this time, there are no clear treatment options for this patient. However, patients who fail GLE/PIB (none so far in the current report) may be able to receive SOF/VEL/VOX.²⁴ In a small study, 29 of 31 patients achieved SVR with SOF/VEL/VOX after GLE/PIB failure (12 of 13 GT1 and 17 of 18 GT3).²⁴

Limitations

This review was an observational, nonrandomized design, and only 1 medical center was involved. These results may not be applicable to other patient populations without a clinic set up with routine follow-ups to encourage adherence and completion of therapy.

Conclusions

Treatment of HCV infection has improved significantly over the past 10 years. Use of DAAs results in SVR for > 90% of patients, especially if the disease had not progressed to cirrhosis. Failure after retreatment for HCV infection was rare as well. Given that cirrhosis seems to increase the chance of treatment failure, it is imperative to identify candidates for treatment before the infection has progressed to cirrhosis. Patients infected with GT3 in particular should be more aggressively identified and treated.

Acknowledgments
The authors thank Nick P. Becky, PharmD, for his contributions to the identification of patients needing treatment for their HCV infection and review of initial manuscript information.

An estimated 3.5 million people in the US have chronic hepatitis C virus (HCV) infection, and between 10% and 20% of those developed cirrhosis over 20 to 30 years.¹ There are at least 6 genotypes (GTs) of HCV, with GT1 being the most common in the US and previously one of the most difficult to treat.^2,3 The goal of treatment is to achieve viral cure, called sustained virologic response (SVR) when HCV viral load remains undetectable several weeks after therapy completion. In the 2000s, pegylated interferon (pegIFN) and ribavirin (RBV) were the standard of care.² For patients with GT1 infections, an SVR of 40 to 50% was commonly seen after 48 weeks of pegIFN/RBV regimens compared with 70 to 80% SVR for GT2 or GT3 after 24 weeks of pegIFN/RBV therapy.² However, treatment has evolved rapidly (Table 1).^2-17

In 2011, the US Food and Drug Administration (FDA) approved the protease inhibitors (PIs) boceprevir and telaprevir, which added a new class of agents with increased SVR for patients with GT1 infection; however, pegIFN and RBV were still needed for treatment.⁴ In addition, both PIs required multiple doses per day and strict adherence to an 8-hour schedule.⁴ Boceprevir required treatment with RBV and pegIFN for 48 weeks unless futility rule was met at 24 weeks of treatment (ie, viral load still detectable).⁴ The SVR in patients with GT1 infection improved to > 65% for patients in clinical trials.² FDA approval of the direct-acting antivirals (DAAs) sofosbuvir and simeprevir in late 2013 decreased the usual duration of therapy to only 12 weeks with improved SVR rates 12 weeks posttherapy (SVR12) to 90% or higher.^2,6,10

FDA approval of ledipasvir (LDV)/sofosbuvir (SOF) in October 2014 resulted in the first interferon-free all-oral regimen indicated for HCV GT1 infection.¹¹ In December 2014, FDA approved a combination of paritaprevir, ritonavir, ombitasvir, and dasabuvir (PrOD).¹² In 2015 GT-specific approvals were issued for daclastavir to be used with SOV for GT1 and GT3 and a combination similar to PrOD without dasabuvir (PrO) for GT4.¹³ In 2016, a combination of elbasvir (ERB) and grazoprevir (GZP) was approved for GT1 and GT4.¹⁴

In 2016, a pangenotypic DAA of SOF and velpatasvir (VEL) was approved.¹⁵ Most recently, combinations of SOF, VEL, and voxilaprevir (VOX), and glecaprevir (GLE) and pibrentasvir (PIB) were approved for patients with previous DAA treatment failures.^{7, 8,16,17} These oral regimens avoided the significant adverse events (AEs) associated with pegIFN and RBV (eg, thrombocytopenia, depression), were expected to improve treatment adherence and shorten duration of therapy.

The West Palm Beach Veterans Affairs has had a nurse practitioner (NP)-based HCV treatment clinic since the late 1990s. When PIs became available, a CPS started reviewing patient electronic health records (EHRs) and monitored response to therapy along with the NP to ensure discontinuation of therapy if futility criteria were met.⁷ Our unpublished experience showed SVR > 60% with both boceprevir and SOF regimens and > 90% with oral DAA regimens.

This review will provide the SVR rates for patients that needed retreatment for HCV infection since 2015 until December 2019. We treated all willing patients, beginning with the patients who had experienced failures with previous regimens. Patients first received education on HCV infection and treatment options in a group class then they were seen by the NP individually for specific education on treatment. The CPS reviewed the patient’s medical record to assess for appropriate therapy, possible drug-drug interactions and contraindications to therapy. In addition, patient outcomes (eg, viral load, AEs) were documented by the CPS in collaboration with the NP throughout treatment until viral load for SVR evaluation was obtained.

Methods

A retrospective EHR review of patients retreated from January 2015 to December 2019 was conducted. Data collected included age, sex, HCV GT, previous therapy, new medications prescribed, creatinine clearance, and achievement of SVR12. This retrospective review was approved by the facility’s scientific advisory committee as part of performance improvement efforts. Descriptive statistics are provided.

Results

Boceprevir

We treated 31 patients with boceprevir of which 3 met futility rule and 28 completed therapy. Eighteen of 28 responded (64%) to the treatment. The 10 patients who failed treatment were retreated with LDV/SOF, and all achieved SVR.

Sofosbuvir

A total of 53 patients were treated with SOF, RBV, and pegIFN for 12 weeks. Forty-one achieved SVR (77%). Of the 12 who failed therapy, all have been retreated and achieved SVR (Table 2).

Interferon-Free DAA Oral Regimens

More than 900 patients have been treated with interferon-free regimens since 2015 and outcomes were documented for > 800 patients. The SVR rates by GT were as follows: GT1 639 of 676 (95%); GT2 76 of 79 (96%); GT3 40 of 48 (83%); and GT4 6 of 6 (100%). Eighty-four percent of patients had GT1 infection. The median age of patient was 62 years, 72% were treatment naïve, and 35% having cirrhosis (based on liver biopsy or FIB4 score).¹⁸

Of 48 treatment failures, 30 patients were retreated; the rest of the patients were lost to follow-up (n = 9) or unable to receive retreatment (n = 9) mainly due to decompensated cirrhosis or liver cancer and short life expectancy. The median age of patient in this retreatment group was 62 years, 62% had cirrhosis, and most were infected with GT1. The average creatinine clearance was 73 mL/min. Twenty-two patients who failed therapy with ledipasvir/SOF were retreated (Table 3). A total of 13 patients out of the 19 tested eventually achieved SVR (68%). Four of the patients who had treatment failure again had GT1 infection and the other 2 GT3. All had cirrhosis.

Thirty-five patients were treated with PrOD, and 32 achieved SVR (91%). All 3 patients were retreated. One patient each achieved SVR with ERB/GZP, SOF/VEL and SOF/VEL/VOX. Fifty patients were treated with ERB/GZP and 45 achieved SVR (90%). All 5 treatment failures were retreated. Four achieved SVR and 1 was lost to follow-up (Table 4). Overall, of 30 patients who were retreated after failure with an all-oral DAA regimen, 27 patients had SVR values available and 21 achieved it (78%).

Discussion

Overall SVR was very high for patients who received oral treatment for HCV infection. A low number of patients failed therapy and were retreated. Patients who failed therapy again were similar in age but were more likely to have cirrhosis when compared with the overall interferon-free treated group. Thus, prompt treatment after HCV detection and before disease progression may improve treatment outcomes. Achieving SVR has been shown to improve fibrosis, portal hypertension, splenomegaly and cirrhosis, and reduce the risk of hepatocellular carcinoma by 70% and liver-related mortality by 90%.^19-21/

Patients who failed therapy primarily had GT1—the most prevalent GT treated. A higher prevalence of GT1 is expected since it is the most common GT in the US.⁶ However, disease progression occurs more rapidly in those with GT3 and is more difficult to treat.²² The overall response rate was lower with this GT (83%) in this report, with only 1 of 3 patients retreated achieving an SVR.

Similar results are documented in retreatment trials.²³ In the POLARIS-1 trial, treatment with SOF/VEL/VOX resulted in an overall response rate of 96% but only 91% for patients with GT3, compared with 95 to 100% for GTs 1, 2, or 4.²³ In the current report, only 1 patient (GT1) failed retreatment with SOF/VEL/VOX. At this time, there are no clear treatment options for this patient. However, patients who fail GLE/PIB (none so far in the current report) may be able to receive SOF/VEL/VOX.²⁴ In a small study, 29 of 31 patients achieved SVR with SOF/VEL/VOX after GLE/PIB failure (12 of 13 GT1 and 17 of 18 GT3).²⁴

Limitations

This review was an observational, nonrandomized design, and only 1 medical center was involved. These results may not be applicable to other patient populations without a clinic set up with routine follow-ups to encourage adherence and completion of therapy.

Conclusions

Treatment of HCV infection has improved significantly over the past 10 years. Use of DAAs results in SVR for > 90% of patients, especially if the disease had not progressed to cirrhosis. Failure after retreatment for HCV infection was rare as well. Given that cirrhosis seems to increase the chance of treatment failure, it is imperative to identify candidates for treatment before the infection has progressed to cirrhosis. Patients infected with GT3 in particular should be more aggressively identified and treated.

Acknowledgments
The authors thank Nick P. Becky, PharmD, for his contributions to the identification of patients needing treatment for their HCV infection and review of initial manuscript information.

With a rapidly growing older adult population, increased attention has been given to cognitive changes that occur with age, with a focus on optimizing the cognitive health of aging individuals.¹ Given the absence of pharmaceutical treatments to prevent cognitive decline, there is an increased need for health care systems to offer alternative or behavioral interventions that can mitigate the effects of cognitive decline in aging.

Notably, many individuals are able to maintain or even improve cognitive functioning throughout their lifespan, with some research implicating health behaviors as an important factor for promoting brain health with age. Specifically, sleep, exercise, eating habits, social engagement, and cognitive stimulation have been linked to improved cognitive functioning.^2-8 In addition to the potential benefits for brain health, there is evidence that greater investment in attaining health goals is associated with subjective reports of higher well-being, fewer mental health symptoms, lower physical health stresses, decreased caregiver burden, and increased functional independence linked with longer independent living.⁹ The latter has a substantial financial impact, such that the positive consequence of increased independence is likely staving off the need for admission to assisted living and adult family homes, which can be costly.

Despite the role of health behaviors in brain aging and overall health and functioning, research indicates that only a small number of older adults (12.8%) follow recommended guidelines for healthy lifestyle factors.¹⁰ Education has been identified as one factor associated with the likelihood of engaging in positive health behaviors, prompting the delivery of health-education interventions. Most psychoeducational interventions have traditionally focused on one aspect of behavior change at a time (eg, sleep); however, Gross and colleaguesconducted a meta-analysis of cognitive interventions and in addition to the overall positive benefits (effect size 0.38), they also found suggestive evidence that interventions that combined multiple training strategies were associated with larger training gains (P = .04) after adjusting for multiple comparisons.¹¹ For example, Miller and colleagues found a significant improvement on both subjective and objective measures of memory following a multicomponent approach that combined training in memory skills, stress reduction, nutrition, and physical activity.¹²

In addition to the potential positive impacts of health behaviors on brain health, findings suggest that targeted emphasis on health behavior change may have the potential to stave off mild cognitiveimpairment (MCI) or dementia even if for a short time. Given the increasing prevalence rates of MCI with age (6.7% in adults aged 60-64 years, reaching 25.2% in adults aged 80-84 years¹³) and dementia (prevalence of MCI converting to dementia is 18-40%¹⁴), as well as the corresponding emotional, financial, and family-oriented consequences (eg, impact on the well-being of family caregivers), the need for behavioral interventions that seek to optimize brain health is becoming increasingly apparent.

More than 9 million veterans are now aged ≥ 65 years.¹⁵ In addition to representing nearly half of all veterans and a sizable portion of aging adults in the US, older veterans are at increased risk of frailty, mortality, and high rates of chronic medical/mental health conditions that can lead to accelerated cognitive aging.^6-17 Together, these conditions highlight the importance of developing comprehensive psychoeducational and behavioral interventions in this population. To address this need, we developed a novel psychoeducation and behavior change group called the Healthy Aging Project-Brain (HAP-B, pronounced “happy”). The HAP-B intervention was designed to promote healthy brain aging by using empirically supported health behavior change strategies, including education, personalized goal setting, and community support. The primary aim of this project was to develop and implement an intervention that was feasible and acceptable (eg, could be implemented in our setting, was appropriate for a veteran population) and to determine any positive outcomes/preliminary effects on overall health and well-being.

Methods

We recruited veterans aged ≥ 50 years through primary care clinics and self-referrals via flyers in the US Department of Veterans Affairs (VA) Puget Sound Health Care System (VAPSHCS), Seattle Division hospital. We targeted the “worried well” and welcomed veterans with MCI and mental health diagnoses. Notably, if there were significant mental health and/or substance use concerns, we encouraged veterans to seek focused care and stabilization prior to or concurrent with group participation. Exclusion criteria included presence of suicidality/homicidality, untreated or unstable substance use disorder, or a diagnosis of dementia. Exclusion criteria were assessed by the referring health care providers (HCPs), when appropriate, and through a health record review. Group facilitators used their clinical judgment to monitor participants if they began experiencing more severe cognitive impairment or acute mental health concerns. Although we did not encounter any of these instances, facilitators were prepared to discuss any concerns with the veteran and their referring HCP. Participants sampled were from 1 of 5 groups offered between January 2018 and March 2019. A waiver from the institutional review board was obtained after meeting criteria for quality improvement/quality assurance (QI/QA) for this study.

Procedures

At the initial stages of development, our team conducted a needs assessment to identify health-related areas where HCPs felt veterans would benefit from additional education and support. The needs assessment was conducted across primary care, geriatric extended care, and the Geriatric Research, Education, and Clinical Center (GRECC) at VAPSHCS. Combining the needs assessment results with the available research base, we identified sleep, physical activity, social engagement, and cognitive stimulation as areas for focus. Notably, although nutrition has been identified as an important factor in cognitive aging, a diet and nutrition class was already available to older veterans at the Seattle VA; hence, we chose to limit overlap by not covering this topic in our group.

The group was offered on a quarterly basis as six 90-minute psychoeducational classes to allow time for didactics, discussion, and practice without overloading participants with information. Each group consisted of 4 to 9 veterans led by 2 cofacilitators. Group structure allowed for feedback and ideas from group members as well as accountability for engaging in behavior change. Cognitive functioning was not formally evaluated. Attendees were asked but not required to complete questionnaires before the classes began and again at completion. In addition at the completion of each group, feedback was collected from veterans and used to modify group content (Figure).

Two pilot groups were implemented in early and mid-2018 with iterative changes after each group. Then we revised the assessment battery and implemented the current version (v1.0), which was first offered in the fall of 2018 and was used with the final 3 groups. Noteworthy changes included weekly check-ins to assess use of health behavior logs and progress toward individual goals, additional pre-and postgroup measures, and in vivo skills practice relevant to the topic being discussed that day.

Each session began with a check-in, which included a review of daily logs and SMART (specific, measurable, attainable, relevant/realistic, and timebound) goals from the previous week.¹⁸ This allowed for praise/reinforcement of health behaviors as well as discussion of potential barriers. Second, an overview of research focusing on the relationship between aging, brain health, and the topic of the day was presented. As an example, in the discussion of social engagement, research was presented about the link between social isolation and cognitive decline; the indirect benefits of social support (eg, social support is linked to improved physical and mental health, which, in turn, is associated with less cognitive decline); and the direct benefits of social support (eg, high levels of emotional support are associated with better cognitive function) (Table 1).⁶

Next, facilitators reviewed skills and strategies to improve functioning in the topic of discussion. During the social engagement group, for example, facilitators discussed tips to improve social skills (eg, asking open-ended questions) and how to build social support into a daily routine (eg, scheduling weekly phone calls with family and friends). Following this discussion of skills, an activity was practiced, reinforcing learned material. During the social engagement group, veterans were invited to use small talk strategies with fellow group members. Finally, group sessions ended with each participant identifying a SMART goal for the coming week and troubleshooting potential barriers to success. SMART goals were kept broad, so veterans could choose a goal related to the topic discussed at the group that day (eg, scheduling a phone call with a friend twice in the coming week during the social engagement-focused group) or choose any other goal to focus on (eg, a sleep-related goal). Similarly, goals could change week to week, or could remain the same throughout the 6-week classes.

Measures

The questionnaires used for QI/QA analyses included the Satisfaction with Life Scale (SWLS); Geriatric Depression Scale-Short Form (GDS-S); Social Support Survey Instrument (SSSI); Pittsburg Sleep Quality Index (PSQI); Medical Outcomes Survey-Short Form (MOS-36 SF); and a self-efficacy scale (adapted from Huckans and colleagues for traumatic brain injury).^19-24 Written feedback was collected at the end of the last group to assess perception of progress, self-perceived behavior change, what was helpful or unhelpful, and how likely the participants were to recommend the group to other veterans (0 to 3, very unlikely to very likely).

To promote consistency with other health and behavior change interventions at the VA, HAP-B used resources from the Whole Health model SMART goals. Research supports the use of self-monitoring techniques like SMART goals for behavior change.²⁵

To facilitate skills practice and self-monitoring between classes, veterans were asked to complete 2 homework assignments. First, at the end of each group, each veteran identified a specific SMART goal to focus on and track in the coming week. Goals were unique to each veteran and allowed to change from week to week. Group discussion around SMART goals involved plans for how to address potential barriers; progress toward goals was discussed at the beginning of the following group. Second, veterans were asked to complete a worksheet used to track progress toward the weekly SMART goal and the specific health behaviors related to the 4 domains targeted by HAP-B. For example, when tracking sleep behaviors, veterans noted bedtime, waketime, number of times they woke up during the night, and length of daytime naps if applicable. Tracking logs were provided at the end of each class for personal purposes only. We asked veterans to rate themselves each week on whether they used the tracking sheet to monitor health behaviors; and how successful they were at accomplishing their previously identified SMART goal. We recorded responses on a 0 to 2 scale (0, not good; 1, fair; 2, good). This rating system was developed and implemented in later groups to promote self-monitoring, accountability, and discussion of potential barriers. However, due to the small sample that completed these ratings and the absence of objective corroborating data, these ratings were not included in the current analyses.

Every participant received a manual in binder format, which provided the didactic information for each group session, skills and strategies discussed in each session, and relevant resources in both the VA and community. For example, social engagement resources included information about volunteer opportunities, VA groups that focus on developing interpersonal skills, and recommendations from past group members on social events (eg, dance lessons at a senior center). We also developed a facilitator version of the manual in which we added comments and guidance on topics for discussion. Materials were developed with the goal of optimizing the ease of dissemination to other sites.

Results

Across the 5 groups, 31 veterans enrolled as participants and completed the initial intake measures, with an average of 6 participants per group (range 4-9). The majority (80%) attended at least 5 of the 6 classes. The mean age was 70.7 years, and 90% of participants were men. Seventy percent of participants self-identified as white, 32% African American, and 3% Native American, which is consistent with VAPSHCS demographics. Of the 31 participants, 16 had a mental health diagnosis, and 6 had a cognitive diagnosis.

At the start of the class, the mean (SD) reports of participants were mild depressive symptoms 5.96 (3.8) on the GDS scale, moderate levels of self-efficacy 3.69 (0.5) on the self-efficacy scale, and moderate levels of satisfaction with life 18.08 (6.8) on the SWLS scale (Table 2). Data from 25 of 31 veterans who completed both pregroup and postgroup surveys were analyzed and paired samples t tests without corrections indicated a reduction in depressive symptoms (P = .01), improved self-efficacy (P = .08), and improved satisfaction with life (P = .03). There were no significant differences in self-reported sleep quality or perceived social support from pregroup to postgroup evaluations. Because the sample size was smaller for the MOS-36, which was not used until group 3, and the subscales are composed of few items each, we conducted exploratory analyses of the 8 MOS-36 subscales and found that well-being, physical functioning, role limitations due to physical and emotional functioning, and energy/fatigue significantly improved over time (Ps < .04).

Discussion

The ability and need to promote brain health with age is an emerging priority as our aging population grows. A growing body of evidence supports the role of health behaviors in healthy brain aging. Education and skills training in a group setting provides a supportive, cost-effective approach for increasing overall health in aging adults. Yet older adults are statistically less likely to engage in these behaviors on a regular basis. The current investigation provides preliminary support for a model of care that uses a comprehensive, experiential psychoeducational approach to facilitate behavior change in older adults. Our aim was to develop and implement an intervention that was feasible and acceptable to our older veterans and to determine any positive outcomes/preliminary effects on overall health and well-being.

Participants indicated that they enjoyed the group, learned new skills (per participant feedback and facilitator observation), and experienced improvements in mood, self-efficacy, and life satisfaction. Given the participants’ positive response to the group and its content, as well as continued referrals by HCPs to this group and low difficulty with ongoing recruitment, this program was deemed both feasible and acceptable in our veteran health care setting. Questions remain about the extent to which participants modified their health behaviors given that we did not collect objective measurements of behaviors (eg, time spent exercising), the duration of behavior change (ie, how long during and after the group were behaviors maintained), and the role of premorbid or concurrent characteristics that may moderate the effect of the intervention on health-related outcomes (eg, sleep quality, perceived social support, overall functioning, concurrent interventions, medications).

Strengths and Limitations

This study had a limited sample size and no control group. However, evidence of significant improvements in depressive symptoms, self-efficacy, and life satisfaction in the development groups without a control group is encouraging. This is particularly noteworthy given that older veterans as a group have higher rates of frailty and mortality than do other similarly aged counterparts.¹⁷An additional weakness is the absence of a brief cognitive assessment or other formal assessment as part of the inclusion/exclusion criteria. However, this program development project provides data from a realistic condition (recruited broadly and with few exclusions, offered in similar format as other VA classes), thus adding strength to the interpretation and possibly the generalizability of these findings.

Conclusions

Future directions include disseminating HAP-B materials and procedures across a variety of sites, both VA and non-VA. In line with this goal, we hope to increase sample size and sample diversity while optimizing protocol integrity during the exportation phase. With a greater sample size and power, we aim to examine the role of self-efficacy and other premorbid factors (eg, cognitive functioning at baseline) as mediators for observed changes in pre-/postmeasures and outcomes. We also hope to incorporate objective measures of behavior change, such as fitness trackers, heart rate/pulse monitors, and actigraphy for monitoring sleep. Finally, we are interested in conducting follow-up with past and future participants to detect changes that may occur with learning new skills following the completion of the group (eg, changes in sleep behavior that take time to take effect) and the extent to which participants continue to use the health behavior skills and strategies to maintain or enhance progress in behavioral goals. Finally, although this intervention was initially designed for use with older veterans receiving health care through the VA, we believe the concepts and work products described here can be used with older adults across a wide range of health care settings. Providers interested in trialing HAP-B at their local site are encouraged to contact the authors.

With a rapidly growing older adult population, increased attention has been given to cognitive changes that occur with age, with a focus on optimizing the cognitive health of aging individuals.¹ Given the absence of pharmaceutical treatments to prevent cognitive decline, there is an increased need for health care systems to offer alternative or behavioral interventions that can mitigate the effects of cognitive decline in aging.

Notably, many individuals are able to maintain or even improve cognitive functioning throughout their lifespan, with some research implicating health behaviors as an important factor for promoting brain health with age. Specifically, sleep, exercise, eating habits, social engagement, and cognitive stimulation have been linked to improved cognitive functioning.^2-8 In addition to the potential benefits for brain health, there is evidence that greater investment in attaining health goals is associated with subjective reports of higher well-being, fewer mental health symptoms, lower physical health stresses, decreased caregiver burden, and increased functional independence linked with longer independent living.⁹ The latter has a substantial financial impact, such that the positive consequence of increased independence is likely staving off the need for admission to assisted living and adult family homes, which can be costly.

Despite the role of health behaviors in brain aging and overall health and functioning, research indicates that only a small number of older adults (12.8%) follow recommended guidelines for healthy lifestyle factors.¹⁰ Education has been identified as one factor associated with the likelihood of engaging in positive health behaviors, prompting the delivery of health-education interventions. Most psychoeducational interventions have traditionally focused on one aspect of behavior change at a time (eg, sleep); however, Gross and colleaguesconducted a meta-analysis of cognitive interventions and in addition to the overall positive benefits (effect size 0.38), they also found suggestive evidence that interventions that combined multiple training strategies were associated with larger training gains (P = .04) after adjusting for multiple comparisons.¹¹ For example, Miller and colleagues found a significant improvement on both subjective and objective measures of memory following a multicomponent approach that combined training in memory skills, stress reduction, nutrition, and physical activity.¹²

In addition to the potential positive impacts of health behaviors on brain health, findings suggest that targeted emphasis on health behavior change may have the potential to stave off mild cognitiveimpairment (MCI) or dementia even if for a short time. Given the increasing prevalence rates of MCI with age (6.7% in adults aged 60-64 years, reaching 25.2% in adults aged 80-84 years¹³) and dementia (prevalence of MCI converting to dementia is 18-40%¹⁴), as well as the corresponding emotional, financial, and family-oriented consequences (eg, impact on the well-being of family caregivers), the need for behavioral interventions that seek to optimize brain health is becoming increasingly apparent.

More than 9 million veterans are now aged ≥ 65 years.¹⁵ In addition to representing nearly half of all veterans and a sizable portion of aging adults in the US, older veterans are at increased risk of frailty, mortality, and high rates of chronic medical/mental health conditions that can lead to accelerated cognitive aging.^6-17 Together, these conditions highlight the importance of developing comprehensive psychoeducational and behavioral interventions in this population. To address this need, we developed a novel psychoeducation and behavior change group called the Healthy Aging Project-Brain (HAP-B, pronounced “happy”). The HAP-B intervention was designed to promote healthy brain aging by using empirically supported health behavior change strategies, including education, personalized goal setting, and community support. The primary aim of this project was to develop and implement an intervention that was feasible and acceptable (eg, could be implemented in our setting, was appropriate for a veteran population) and to determine any positive outcomes/preliminary effects on overall health and well-being.

Methods

We recruited veterans aged ≥ 50 years through primary care clinics and self-referrals via flyers in the US Department of Veterans Affairs (VA) Puget Sound Health Care System (VAPSHCS), Seattle Division hospital. We targeted the “worried well” and welcomed veterans with MCI and mental health diagnoses. Notably, if there were significant mental health and/or substance use concerns, we encouraged veterans to seek focused care and stabilization prior to or concurrent with group participation. Exclusion criteria included presence of suicidality/homicidality, untreated or unstable substance use disorder, or a diagnosis of dementia. Exclusion criteria were assessed by the referring health care providers (HCPs), when appropriate, and through a health record review. Group facilitators used their clinical judgment to monitor participants if they began experiencing more severe cognitive impairment or acute mental health concerns. Although we did not encounter any of these instances, facilitators were prepared to discuss any concerns with the veteran and their referring HCP. Participants sampled were from 1 of 5 groups offered between January 2018 and March 2019. A waiver from the institutional review board was obtained after meeting criteria for quality improvement/quality assurance (QI/QA) for this study.

Procedures

At the initial stages of development, our team conducted a needs assessment to identify health-related areas where HCPs felt veterans would benefit from additional education and support. The needs assessment was conducted across primary care, geriatric extended care, and the Geriatric Research, Education, and Clinical Center (GRECC) at VAPSHCS. Combining the needs assessment results with the available research base, we identified sleep, physical activity, social engagement, and cognitive stimulation as areas for focus. Notably, although nutrition has been identified as an important factor in cognitive aging, a diet and nutrition class was already available to older veterans at the Seattle VA; hence, we chose to limit overlap by not covering this topic in our group.

The group was offered on a quarterly basis as six 90-minute psychoeducational classes to allow time for didactics, discussion, and practice without overloading participants with information. Each group consisted of 4 to 9 veterans led by 2 cofacilitators. Group structure allowed for feedback and ideas from group members as well as accountability for engaging in behavior change. Cognitive functioning was not formally evaluated. Attendees were asked but not required to complete questionnaires before the classes began and again at completion. In addition at the completion of each group, feedback was collected from veterans and used to modify group content (Figure).

Two pilot groups were implemented in early and mid-2018 with iterative changes after each group. Then we revised the assessment battery and implemented the current version (v1.0), which was first offered in the fall of 2018 and was used with the final 3 groups. Noteworthy changes included weekly check-ins to assess use of health behavior logs and progress toward individual goals, additional pre-and postgroup measures, and in vivo skills practice relevant to the topic being discussed that day.

Each session began with a check-in, which included a review of daily logs and SMART (specific, measurable, attainable, relevant/realistic, and timebound) goals from the previous week.¹⁸ This allowed for praise/reinforcement of health behaviors as well as discussion of potential barriers. Second, an overview of research focusing on the relationship between aging, brain health, and the topic of the day was presented. As an example, in the discussion of social engagement, research was presented about the link between social isolation and cognitive decline; the indirect benefits of social support (eg, social support is linked to improved physical and mental health, which, in turn, is associated with less cognitive decline); and the direct benefits of social support (eg, high levels of emotional support are associated with better cognitive function) (Table 1).⁶

Next, facilitators reviewed skills and strategies to improve functioning in the topic of discussion. During the social engagement group, for example, facilitators discussed tips to improve social skills (eg, asking open-ended questions) and how to build social support into a daily routine (eg, scheduling weekly phone calls with family and friends). Following this discussion of skills, an activity was practiced, reinforcing learned material. During the social engagement group, veterans were invited to use small talk strategies with fellow group members. Finally, group sessions ended with each participant identifying a SMART goal for the coming week and troubleshooting potential barriers to success. SMART goals were kept broad, so veterans could choose a goal related to the topic discussed at the group that day (eg, scheduling a phone call with a friend twice in the coming week during the social engagement-focused group) or choose any other goal to focus on (eg, a sleep-related goal). Similarly, goals could change week to week, or could remain the same throughout the 6-week classes.

Measures

The questionnaires used for QI/QA analyses included the Satisfaction with Life Scale (SWLS); Geriatric Depression Scale-Short Form (GDS-S); Social Support Survey Instrument (SSSI); Pittsburg Sleep Quality Index (PSQI); Medical Outcomes Survey-Short Form (MOS-36 SF); and a self-efficacy scale (adapted from Huckans and colleagues for traumatic brain injury).^19-24 Written feedback was collected at the end of the last group to assess perception of progress, self-perceived behavior change, what was helpful or unhelpful, and how likely the participants were to recommend the group to other veterans (0 to 3, very unlikely to very likely).

To promote consistency with other health and behavior change interventions at the VA, HAP-B used resources from the Whole Health model SMART goals. Research supports the use of self-monitoring techniques like SMART goals for behavior change.²⁵

To facilitate skills practice and self-monitoring between classes, veterans were asked to complete 2 homework assignments. First, at the end of each group, each veteran identified a specific SMART goal to focus on and track in the coming week. Goals were unique to each veteran and allowed to change from week to week. Group discussion around SMART goals involved plans for how to address potential barriers; progress toward goals was discussed at the beginning of the following group. Second, veterans were asked to complete a worksheet used to track progress toward the weekly SMART goal and the specific health behaviors related to the 4 domains targeted by HAP-B. For example, when tracking sleep behaviors, veterans noted bedtime, waketime, number of times they woke up during the night, and length of daytime naps if applicable. Tracking logs were provided at the end of each class for personal purposes only. We asked veterans to rate themselves each week on whether they used the tracking sheet to monitor health behaviors; and how successful they were at accomplishing their previously identified SMART goal. We recorded responses on a 0 to 2 scale (0, not good; 1, fair; 2, good). This rating system was developed and implemented in later groups to promote self-monitoring, accountability, and discussion of potential barriers. However, due to the small sample that completed these ratings and the absence of objective corroborating data, these ratings were not included in the current analyses.

Every participant received a manual in binder format, which provided the didactic information for each group session, skills and strategies discussed in each session, and relevant resources in both the VA and community. For example, social engagement resources included information about volunteer opportunities, VA groups that focus on developing interpersonal skills, and recommendations from past group members on social events (eg, dance lessons at a senior center). We also developed a facilitator version of the manual in which we added comments and guidance on topics for discussion. Materials were developed with the goal of optimizing the ease of dissemination to other sites.

Results

Across the 5 groups, 31 veterans enrolled as participants and completed the initial intake measures, with an average of 6 participants per group (range 4-9). The majority (80%) attended at least 5 of the 6 classes. The mean age was 70.7 years, and 90% of participants were men. Seventy percent of participants self-identified as white, 32% African American, and 3% Native American, which is consistent with VAPSHCS demographics. Of the 31 participants, 16 had a mental health diagnosis, and 6 had a cognitive diagnosis.

At the start of the class, the mean (SD) reports of participants were mild depressive symptoms 5.96 (3.8) on the GDS scale, moderate levels of self-efficacy 3.69 (0.5) on the self-efficacy scale, and moderate levels of satisfaction with life 18.08 (6.8) on the SWLS scale (Table 2). Data from 25 of 31 veterans who completed both pregroup and postgroup surveys were analyzed and paired samples t tests without corrections indicated a reduction in depressive symptoms (P = .01), improved self-efficacy (P = .08), and improved satisfaction with life (P = .03). There were no significant differences in self-reported sleep quality or perceived social support from pregroup to postgroup evaluations. Because the sample size was smaller for the MOS-36, which was not used until group 3, and the subscales are composed of few items each, we conducted exploratory analyses of the 8 MOS-36 subscales and found that well-being, physical functioning, role limitations due to physical and emotional functioning, and energy/fatigue significantly improved over time (Ps < .04).

Discussion

The ability and need to promote brain health with age is an emerging priority as our aging population grows. A growing body of evidence supports the role of health behaviors in healthy brain aging. Education and skills training in a group setting provides a supportive, cost-effective approach for increasing overall health in aging adults. Yet older adults are statistically less likely to engage in these behaviors on a regular basis. The current investigation provides preliminary support for a model of care that uses a comprehensive, experiential psychoeducational approach to facilitate behavior change in older adults. Our aim was to develop and implement an intervention that was feasible and acceptable to our older veterans and to determine any positive outcomes/preliminary effects on overall health and well-being.

Participants indicated that they enjoyed the group, learned new skills (per participant feedback and facilitator observation), and experienced improvements in mood, self-efficacy, and life satisfaction. Given the participants’ positive response to the group and its content, as well as continued referrals by HCPs to this group and low difficulty with ongoing recruitment, this program was deemed both feasible and acceptable in our veteran health care setting. Questions remain about the extent to which participants modified their health behaviors given that we did not collect objective measurements of behaviors (eg, time spent exercising), the duration of behavior change (ie, how long during and after the group were behaviors maintained), and the role of premorbid or concurrent characteristics that may moderate the effect of the intervention on health-related outcomes (eg, sleep quality, perceived social support, overall functioning, concurrent interventions, medications).

Strengths and Limitations

This study had a limited sample size and no control group. However, evidence of significant improvements in depressive symptoms, self-efficacy, and life satisfaction in the development groups without a control group is encouraging. This is particularly noteworthy given that older veterans as a group have higher rates of frailty and mortality than do other similarly aged counterparts.¹⁷An additional weakness is the absence of a brief cognitive assessment or other formal assessment as part of the inclusion/exclusion criteria. However, this program development project provides data from a realistic condition (recruited broadly and with few exclusions, offered in similar format as other VA classes), thus adding strength to the interpretation and possibly the generalizability of these findings.

Conclusions

Future directions include disseminating HAP-B materials and procedures across a variety of sites, both VA and non-VA. In line with this goal, we hope to increase sample size and sample diversity while optimizing protocol integrity during the exportation phase. With a greater sample size and power, we aim to examine the role of self-efficacy and other premorbid factors (eg, cognitive functioning at baseline) as mediators for observed changes in pre-/postmeasures and outcomes. We also hope to incorporate objective measures of behavior change, such as fitness trackers, heart rate/pulse monitors, and actigraphy for monitoring sleep. Finally, we are interested in conducting follow-up with past and future participants to detect changes that may occur with learning new skills following the completion of the group (eg, changes in sleep behavior that take time to take effect) and the extent to which participants continue to use the health behavior skills and strategies to maintain or enhance progress in behavioral goals. Finally, although this intervention was initially designed for use with older veterans receiving health care through the VA, we believe the concepts and work products described here can be used with older adults across a wide range of health care settings. Providers interested in trialing HAP-B at their local site are encouraged to contact the authors.

With a rapidly growing older adult population, increased attention has been given to cognitive changes that occur with age, with a focus on optimizing the cognitive health of aging individuals.¹ Given the absence of pharmaceutical treatments to prevent cognitive decline, there is an increased need for health care systems to offer alternative or behavioral interventions that can mitigate the effects of cognitive decline in aging.

Notably, many individuals are able to maintain or even improve cognitive functioning throughout their lifespan, with some research implicating health behaviors as an important factor for promoting brain health with age. Specifically, sleep, exercise, eating habits, social engagement, and cognitive stimulation have been linked to improved cognitive functioning.^2-8 In addition to the potential benefits for brain health, there is evidence that greater investment in attaining health goals is associated with subjective reports of higher well-being, fewer mental health symptoms, lower physical health stresses, decreased caregiver burden, and increased functional independence linked with longer independent living.⁹ The latter has a substantial financial impact, such that the positive consequence of increased independence is likely staving off the need for admission to assisted living and adult family homes, which can be costly.

Despite the role of health behaviors in brain aging and overall health and functioning, research indicates that only a small number of older adults (12.8%) follow recommended guidelines for healthy lifestyle factors.¹⁰ Education has been identified as one factor associated with the likelihood of engaging in positive health behaviors, prompting the delivery of health-education interventions. Most psychoeducational interventions have traditionally focused on one aspect of behavior change at a time (eg, sleep); however, Gross and colleaguesconducted a meta-analysis of cognitive interventions and in addition to the overall positive benefits (effect size 0.38), they also found suggestive evidence that interventions that combined multiple training strategies were associated with larger training gains (P = .04) after adjusting for multiple comparisons.¹¹ For example, Miller and colleagues found a significant improvement on both subjective and objective measures of memory following a multicomponent approach that combined training in memory skills, stress reduction, nutrition, and physical activity.¹²

In addition to the potential positive impacts of health behaviors on brain health, findings suggest that targeted emphasis on health behavior change may have the potential to stave off mild cognitiveimpairment (MCI) or dementia even if for a short time. Given the increasing prevalence rates of MCI with age (6.7% in adults aged 60-64 years, reaching 25.2% in adults aged 80-84 years¹³) and dementia (prevalence of MCI converting to dementia is 18-40%¹⁴), as well as the corresponding emotional, financial, and family-oriented consequences (eg, impact on the well-being of family caregivers), the need for behavioral interventions that seek to optimize brain health is becoming increasingly apparent.

More than 9 million veterans are now aged ≥ 65 years.¹⁵ In addition to representing nearly half of all veterans and a sizable portion of aging adults in the US, older veterans are at increased risk of frailty, mortality, and high rates of chronic medical/mental health conditions that can lead to accelerated cognitive aging.^6-17 Together, these conditions highlight the importance of developing comprehensive psychoeducational and behavioral interventions in this population. To address this need, we developed a novel psychoeducation and behavior change group called the Healthy Aging Project-Brain (HAP-B, pronounced “happy”). The HAP-B intervention was designed to promote healthy brain aging by using empirically supported health behavior change strategies, including education, personalized goal setting, and community support. The primary aim of this project was to develop and implement an intervention that was feasible and acceptable (eg, could be implemented in our setting, was appropriate for a veteran population) and to determine any positive outcomes/preliminary effects on overall health and well-being.

Methods

We recruited veterans aged ≥ 50 years through primary care clinics and self-referrals via flyers in the US Department of Veterans Affairs (VA) Puget Sound Health Care System (VAPSHCS), Seattle Division hospital. We targeted the “worried well” and welcomed veterans with MCI and mental health diagnoses. Notably, if there were significant mental health and/or substance use concerns, we encouraged veterans to seek focused care and stabilization prior to or concurrent with group participation. Exclusion criteria included presence of suicidality/homicidality, untreated or unstable substance use disorder, or a diagnosis of dementia. Exclusion criteria were assessed by the referring health care providers (HCPs), when appropriate, and through a health record review. Group facilitators used their clinical judgment to monitor participants if they began experiencing more severe cognitive impairment or acute mental health concerns. Although we did not encounter any of these instances, facilitators were prepared to discuss any concerns with the veteran and their referring HCP. Participants sampled were from 1 of 5 groups offered between January 2018 and March 2019. A waiver from the institutional review board was obtained after meeting criteria for quality improvement/quality assurance (QI/QA) for this study.

Procedures

At the initial stages of development, our team conducted a needs assessment to identify health-related areas where HCPs felt veterans would benefit from additional education and support. The needs assessment was conducted across primary care, geriatric extended care, and the Geriatric Research, Education, and Clinical Center (GRECC) at VAPSHCS. Combining the needs assessment results with the available research base, we identified sleep, physical activity, social engagement, and cognitive stimulation as areas for focus. Notably, although nutrition has been identified as an important factor in cognitive aging, a diet and nutrition class was already available to older veterans at the Seattle VA; hence, we chose to limit overlap by not covering this topic in our group.

The group was offered on a quarterly basis as six 90-minute psychoeducational classes to allow time for didactics, discussion, and practice without overloading participants with information. Each group consisted of 4 to 9 veterans led by 2 cofacilitators. Group structure allowed for feedback and ideas from group members as well as accountability for engaging in behavior change. Cognitive functioning was not formally evaluated. Attendees were asked but not required to complete questionnaires before the classes began and again at completion. In addition at the completion of each group, feedback was collected from veterans and used to modify group content (Figure).

Two pilot groups were implemented in early and mid-2018 with iterative changes after each group. Then we revised the assessment battery and implemented the current version (v1.0), which was first offered in the fall of 2018 and was used with the final 3 groups. Noteworthy changes included weekly check-ins to assess use of health behavior logs and progress toward individual goals, additional pre-and postgroup measures, and in vivo skills practice relevant to the topic being discussed that day.

Each session began with a check-in, which included a review of daily logs and SMART (specific, measurable, attainable, relevant/realistic, and timebound) goals from the previous week.¹⁸ This allowed for praise/reinforcement of health behaviors as well as discussion of potential barriers. Second, an overview of research focusing on the relationship between aging, brain health, and the topic of the day was presented. As an example, in the discussion of social engagement, research was presented about the link between social isolation and cognitive decline; the indirect benefits of social support (eg, social support is linked to improved physical and mental health, which, in turn, is associated with less cognitive decline); and the direct benefits of social support (eg, high levels of emotional support are associated with better cognitive function) (Table 1).⁶

Next, facilitators reviewed skills and strategies to improve functioning in the topic of discussion. During the social engagement group, for example, facilitators discussed tips to improve social skills (eg, asking open-ended questions) and how to build social support into a daily routine (eg, scheduling weekly phone calls with family and friends). Following this discussion of skills, an activity was practiced, reinforcing learned material. During the social engagement group, veterans were invited to use small talk strategies with fellow group members. Finally, group sessions ended with each participant identifying a SMART goal for the coming week and troubleshooting potential barriers to success. SMART goals were kept broad, so veterans could choose a goal related to the topic discussed at the group that day (eg, scheduling a phone call with a friend twice in the coming week during the social engagement-focused group) or choose any other goal to focus on (eg, a sleep-related goal). Similarly, goals could change week to week, or could remain the same throughout the 6-week classes.

Measures

The questionnaires used for QI/QA analyses included the Satisfaction with Life Scale (SWLS); Geriatric Depression Scale-Short Form (GDS-S); Social Support Survey Instrument (SSSI); Pittsburg Sleep Quality Index (PSQI); Medical Outcomes Survey-Short Form (MOS-36 SF); and a self-efficacy scale (adapted from Huckans and colleagues for traumatic brain injury).^19-24 Written feedback was collected at the end of the last group to assess perception of progress, self-perceived behavior change, what was helpful or unhelpful, and how likely the participants were to recommend the group to other veterans (0 to 3, very unlikely to very likely).

To promote consistency with other health and behavior change interventions at the VA, HAP-B used resources from the Whole Health model SMART goals. Research supports the use of self-monitoring techniques like SMART goals for behavior change.²⁵

To facilitate skills practice and self-monitoring between classes, veterans were asked to complete 2 homework assignments. First, at the end of each group, each veteran identified a specific SMART goal to focus on and track in the coming week. Goals were unique to each veteran and allowed to change from week to week. Group discussion around SMART goals involved plans for how to address potential barriers; progress toward goals was discussed at the beginning of the following group. Second, veterans were asked to complete a worksheet used to track progress toward the weekly SMART goal and the specific health behaviors related to the 4 domains targeted by HAP-B. For example, when tracking sleep behaviors, veterans noted bedtime, waketime, number of times they woke up during the night, and length of daytime naps if applicable. Tracking logs were provided at the end of each class for personal purposes only. We asked veterans to rate themselves each week on whether they used the tracking sheet to monitor health behaviors; and how successful they were at accomplishing their previously identified SMART goal. We recorded responses on a 0 to 2 scale (0, not good; 1, fair; 2, good). This rating system was developed and implemented in later groups to promote self-monitoring, accountability, and discussion of potential barriers. However, due to the small sample that completed these ratings and the absence of objective corroborating data, these ratings were not included in the current analyses.

Every participant received a manual in binder format, which provided the didactic information for each group session, skills and strategies discussed in each session, and relevant resources in both the VA and community. For example, social engagement resources included information about volunteer opportunities, VA groups that focus on developing interpersonal skills, and recommendations from past group members on social events (eg, dance lessons at a senior center). We also developed a facilitator version of the manual in which we added comments and guidance on topics for discussion. Materials were developed with the goal of optimizing the ease of dissemination to other sites.

Results

Across the 5 groups, 31 veterans enrolled as participants and completed the initial intake measures, with an average of 6 participants per group (range 4-9). The majority (80%) attended at least 5 of the 6 classes. The mean age was 70.7 years, and 90% of participants were men. Seventy percent of participants self-identified as white, 32% African American, and 3% Native American, which is consistent with VAPSHCS demographics. Of the 31 participants, 16 had a mental health diagnosis, and 6 had a cognitive diagnosis.

At the start of the class, the mean (SD) reports of participants were mild depressive symptoms 5.96 (3.8) on the GDS scale, moderate levels of self-efficacy 3.69 (0.5) on the self-efficacy scale, and moderate levels of satisfaction with life 18.08 (6.8) on the SWLS scale (Table 2). Data from 25 of 31 veterans who completed both pregroup and postgroup surveys were analyzed and paired samples t tests without corrections indicated a reduction in depressive symptoms (P = .01), improved self-efficacy (P = .08), and improved satisfaction with life (P = .03). There were no significant differences in self-reported sleep quality or perceived social support from pregroup to postgroup evaluations. Because the sample size was smaller for the MOS-36, which was not used until group 3, and the subscales are composed of few items each, we conducted exploratory analyses of the 8 MOS-36 subscales and found that well-being, physical functioning, role limitations due to physical and emotional functioning, and energy/fatigue significantly improved over time (Ps < .04).

Discussion

The ability and need to promote brain health with age is an emerging priority as our aging population grows. A growing body of evidence supports the role of health behaviors in healthy brain aging. Education and skills training in a group setting provides a supportive, cost-effective approach for increasing overall health in aging adults. Yet older adults are statistically less likely to engage in these behaviors on a regular basis. The current investigation provides preliminary support for a model of care that uses a comprehensive, experiential psychoeducational approach to facilitate behavior change in older adults. Our aim was to develop and implement an intervention that was feasible and acceptable to our older veterans and to determine any positive outcomes/preliminary effects on overall health and well-being.

Participants indicated that they enjoyed the group, learned new skills (per participant feedback and facilitator observation), and experienced improvements in mood, self-efficacy, and life satisfaction. Given the participants’ positive response to the group and its content, as well as continued referrals by HCPs to this group and low difficulty with ongoing recruitment, this program was deemed both feasible and acceptable in our veteran health care setting. Questions remain about the extent to which participants modified their health behaviors given that we did not collect objective measurements of behaviors (eg, time spent exercising), the duration of behavior change (ie, how long during and after the group were behaviors maintained), and the role of premorbid or concurrent characteristics that may moderate the effect of the intervention on health-related outcomes (eg, sleep quality, perceived social support, overall functioning, concurrent interventions, medications).

Strengths and Limitations

This study had a limited sample size and no control group. However, evidence of significant improvements in depressive symptoms, self-efficacy, and life satisfaction in the development groups without a control group is encouraging. This is particularly noteworthy given that older veterans as a group have higher rates of frailty and mortality than do other similarly aged counterparts.¹⁷An additional weakness is the absence of a brief cognitive assessment or other formal assessment as part of the inclusion/exclusion criteria. However, this program development project provides data from a realistic condition (recruited broadly and with few exclusions, offered in similar format as other VA classes), thus adding strength to the interpretation and possibly the generalizability of these findings.

Conclusions

Future directions include disseminating HAP-B materials and procedures across a variety of sites, both VA and non-VA. In line with this goal, we hope to increase sample size and sample diversity while optimizing protocol integrity during the exportation phase. With a greater sample size and power, we aim to examine the role of self-efficacy and other premorbid factors (eg, cognitive functioning at baseline) as mediators for observed changes in pre-/postmeasures and outcomes. We also hope to incorporate objective measures of behavior change, such as fitness trackers, heart rate/pulse monitors, and actigraphy for monitoring sleep. Finally, we are interested in conducting follow-up with past and future participants to detect changes that may occur with learning new skills following the completion of the group (eg, changes in sleep behavior that take time to take effect) and the extent to which participants continue to use the health behavior skills and strategies to maintain or enhance progress in behavioral goals. Finally, although this intervention was initially designed for use with older veterans receiving health care through the VA, we believe the concepts and work products described here can be used with older adults across a wide range of health care settings. Providers interested in trialing HAP-B at their local site are encouraged to contact the authors.