Robert C. Marshall, MD, MPH, MISM, FAAFP, FAMIA

ILLUSTRATIVE CASE

A 48-year-old man is admitted to your family medicine service for cellulitis after failed outpatient therapy. He has presumed community-acquired methicillin-resistant Staphylococcus aureus infection of the left lower extremity and is receiving intravenous (IV) vancomycin. His BP this morning is 176/98 mm Hg, and the reading from the previous shift was 168/94 mm Hg. He is asymptomatic from this elevated BP. Based on protocol, his nurse is asking about treatment in response to the multiple elevated readings. How should you address the patient’s elevated BP, knowing that you will see him for a transition management appointment in 2 weeks?

Elevated BP is common in the adult inpatient setting. Prevalence estimates range from 25% to > 50%. Many factors can contribute to elevated BP in the acute illness setting, such as pain, anxiety, medication withdrawal, and volume status.^2,3

Treatment of elevated BP in outpatients is well researched, with evidence-based guidelines for physicians. That is not the case for treatment of asymptomatic elevated BP in the inpatient setting. Most published guidance on inpatient management of acutely elevated BP recommends IV medications, such as hydralazine or labetalol, although there is limited evidence to support such recommendations. There is minimal evidence for outcomes-based benefit in treating acute elevations of inpatient BP, such as reduced myocardial injury or stroke; however, there is some evidence of adverse outcomes, such as hypotension and prolonged hospital stays.^4-8

Although the possibility of intensifying antihypertensive therapy for those with known hypertension or those with presumed “new-onset” hypertension could theoretically lead to improved outcomes over the long term, there is little evidence to support this presumption. Rather, there is evidence that intensification of antihypertensive therapy at discharge is linked to short-term harms. This was demonstrated in a propensity-matched veteran cohort that included 4056 hospitalized older adults with hypertension (mean age, 77 years; 3961 men), equally split between those who received antihypertensive intensification at hospital discharge and those who did not. Within 30 days, patients receiving intensification had a higher risk of readmission (number needed to harm [NNH] = 27) and serious adverse events (NNH = 63).⁹

The current study aimed to put all these pieces together by quantifying the prevalence of hypertension in hospitalized patients, characterizing clinician response to patients’ acutely elevated BP, and comparing both short- and long-term outcomes in patients treated for acute BP elevations while hospitalized vs those who were not. The study also assessed the potential effects of antihypertensive intensification at discharge.

STUDY SUMMARY

Treatment of acute hypertension was associated with end-organ injury

This retrospective, propensity score–matched cohort study (N = 22,834) evaluated the electronic health records of all adult patients (age > 18 years) admitted to a medicine service with a noncardiovascular diagnosis over a 1-year period at 10 Cleveland Clinic hospitals, with 1 year of follow-up data.

Exclusion criteria included hospitalization for a cardiovascular diagnosis; admission for a cerebrovascular event or acute coronary syndrome within the previous 30 days; pregnancy; length of stay of less than 2 days or more than 14 days; and lack of outpatient medication data. Patients were propensity-score matched using BP, demographic features, comorbidities, hospital shift, and time since admission. Exposure was defined as administration of IV antihypertensive medication or a new class of oral antihypertensive medication.

Continue to: Outcomes were defined...

The 22,834 patients had a mean (SD) age of 65.6 (17.9) years; 12,993 (56.9%) were women, and 15,963 (69.9%) were White. Of the 17,821 (78%) who had at least 1 inpatient hypertensive systolic BP (SBP) episode, defined as an SBP ≥ 140 mm Hg, 5904 (33.1%) received a new treatment. Of those receiving a new treatment, 4378 (74.2%) received only oral treatment, and 1516 (25.7%) received at least 1 dose of IV medication with or without oral dosing.

Acute treatment of elevated BP in noncardiac inpatients was not beneficial, and treatment intensification at discharge did not improve BP control over the following year.

Using the propensity-matched sample (4520 treated for elevated BP matched to 4520 who were not treated), treated patients had higher rates of AKI (10.3% vs 7.9%; P < .001) and myocardial injury (1.2% vs 0.6%; P = .003). When assessed by SBP, nontreatment of BP was still superior up to an SBP of 199 mm Hg. At an SBP of ≥ 200 mm Hg, there was no difference in rates of AKI or MI between the treatment and nontreatment groups. There was no difference in stroke in either cohort, although the overall numbers were quite low.

Patients with and without antihypertensive intensification at discharge had similar rates of MI (0.1% vs 0.2%; P > .99) and stroke (0.5% vs 0.4%; P > .99) in a matched cohort at 30 days post discharge. At 1 year, BP control in the intensification vs no-intensification groups was nearly the same: maximum SBP was 157.2 mm Hg vs 157.8 mm Hg, respectively (P = .54) and maximum diastolic BP was 86.5 mm Hg vs 86.1 mm Hg, respectively (P = .49).

WHAT’S NEW

Previous research is confirmed in a more diverse population

Whereas previous research showed no benefit to intensification of treatment among hospitalized older male patients, this large, retrospective, propensity score–matched cohort study demonstrated the short- and long-term effects of treating acute, asymptomatic BP elevations in a younger, more generalizable population that included women. Regardless of treatment modality, there appeared to be more harm than good from treating these BP elevations.

In addition, the study appears to corroborate previous research showing that intensification of BP treatment at discharge did not lead to better outcomes.⁹ At the very least, the study makes a reasonable argument that treating acute BP elevations in noncardiac patients in the hospital setting is not beneficial.

CAVEATS

Impact of existing therapy could be underestimated

This study had several important limitations. First, 23% of treated participants were excluded from the propensity analysis without justification from the authors. Additionally, there was no reporting of missing data and how it was managed. The authors’ definition of treatment excluded dose intensification of existing antihypertensive therapy, which would undercount the number of treated patients. However, this could underestimate the actual harms of the acute antihypertensive therapy. The authors also included patients with atrial fibrillation and heart failure in the study population, even though they already may have been taking antihypertensive agents.

CHALLENGES TO IMPLEMENTATION

Potential delays in translating findings to patient care

Although several recent studies have shown the potential benefit of not treating asymptomatic acute BP elevations in inpatients, incorporating that information into electronic health record order sets or clinical decision support, and disseminating it to clinical end users, will take time. In the interim, despite these findings, patients may continue to receive IV or oral medications to treat acute, asymptomatic BP elevations while hospitalized for noncardiac diagnoses.

ILLUSTRATIVE CASE

A 48-year-old man is admitted to your family medicine service for cellulitis after failed outpatient therapy. He has presumed community-acquired methicillin-resistant Staphylococcus aureus infection of the left lower extremity and is receiving intravenous (IV) vancomycin. His BP this morning is 176/98 mm Hg, and the reading from the previous shift was 168/94 mm Hg. He is asymptomatic from this elevated BP. Based on protocol, his nurse is asking about treatment in response to the multiple elevated readings. How should you address the patient’s elevated BP, knowing that you will see him for a transition management appointment in 2 weeks?

Elevated BP is common in the adult inpatient setting. Prevalence estimates range from 25% to > 50%. Many factors can contribute to elevated BP in the acute illness setting, such as pain, anxiety, medication withdrawal, and volume status.^2,3

Treatment of elevated BP in outpatients is well researched, with evidence-based guidelines for physicians. That is not the case for treatment of asymptomatic elevated BP in the inpatient setting. Most published guidance on inpatient management of acutely elevated BP recommends IV medications, such as hydralazine or labetalol, although there is limited evidence to support such recommendations. There is minimal evidence for outcomes-based benefit in treating acute elevations of inpatient BP, such as reduced myocardial injury or stroke; however, there is some evidence of adverse outcomes, such as hypotension and prolonged hospital stays.^4-8

Although the possibility of intensifying antihypertensive therapy for those with known hypertension or those with presumed “new-onset” hypertension could theoretically lead to improved outcomes over the long term, there is little evidence to support this presumption. Rather, there is evidence that intensification of antihypertensive therapy at discharge is linked to short-term harms. This was demonstrated in a propensity-matched veteran cohort that included 4056 hospitalized older adults with hypertension (mean age, 77 years; 3961 men), equally split between those who received antihypertensive intensification at hospital discharge and those who did not. Within 30 days, patients receiving intensification had a higher risk of readmission (number needed to harm [NNH] = 27) and serious adverse events (NNH = 63).⁹

The current study aimed to put all these pieces together by quantifying the prevalence of hypertension in hospitalized patients, characterizing clinician response to patients’ acutely elevated BP, and comparing both short- and long-term outcomes in patients treated for acute BP elevations while hospitalized vs those who were not. The study also assessed the potential effects of antihypertensive intensification at discharge.

STUDY SUMMARY

Treatment of acute hypertension was associated with end-organ injury

This retrospective, propensity score–matched cohort study (N = 22,834) evaluated the electronic health records of all adult patients (age > 18 years) admitted to a medicine service with a noncardiovascular diagnosis over a 1-year period at 10 Cleveland Clinic hospitals, with 1 year of follow-up data.

Exclusion criteria included hospitalization for a cardiovascular diagnosis; admission for a cerebrovascular event or acute coronary syndrome within the previous 30 days; pregnancy; length of stay of less than 2 days or more than 14 days; and lack of outpatient medication data. Patients were propensity-score matched using BP, demographic features, comorbidities, hospital shift, and time since admission. Exposure was defined as administration of IV antihypertensive medication or a new class of oral antihypertensive medication.

Continue to: Outcomes were defined...

The 22,834 patients had a mean (SD) age of 65.6 (17.9) years; 12,993 (56.9%) were women, and 15,963 (69.9%) were White. Of the 17,821 (78%) who had at least 1 inpatient hypertensive systolic BP (SBP) episode, defined as an SBP ≥ 140 mm Hg, 5904 (33.1%) received a new treatment. Of those receiving a new treatment, 4378 (74.2%) received only oral treatment, and 1516 (25.7%) received at least 1 dose of IV medication with or without oral dosing.

Acute treatment of elevated BP in noncardiac inpatients was not beneficial, and treatment intensification at discharge did not improve BP control over the following year.

Using the propensity-matched sample (4520 treated for elevated BP matched to 4520 who were not treated), treated patients had higher rates of AKI (10.3% vs 7.9%; P < .001) and myocardial injury (1.2% vs 0.6%; P = .003). When assessed by SBP, nontreatment of BP was still superior up to an SBP of 199 mm Hg. At an SBP of ≥ 200 mm Hg, there was no difference in rates of AKI or MI between the treatment and nontreatment groups. There was no difference in stroke in either cohort, although the overall numbers were quite low.

Patients with and without antihypertensive intensification at discharge had similar rates of MI (0.1% vs 0.2%; P > .99) and stroke (0.5% vs 0.4%; P > .99) in a matched cohort at 30 days post discharge. At 1 year, BP control in the intensification vs no-intensification groups was nearly the same: maximum SBP was 157.2 mm Hg vs 157.8 mm Hg, respectively (P = .54) and maximum diastolic BP was 86.5 mm Hg vs 86.1 mm Hg, respectively (P = .49).

WHAT’S NEW

Previous research is confirmed in a more diverse population

Whereas previous research showed no benefit to intensification of treatment among hospitalized older male patients, this large, retrospective, propensity score–matched cohort study demonstrated the short- and long-term effects of treating acute, asymptomatic BP elevations in a younger, more generalizable population that included women. Regardless of treatment modality, there appeared to be more harm than good from treating these BP elevations.

In addition, the study appears to corroborate previous research showing that intensification of BP treatment at discharge did not lead to better outcomes.⁹ At the very least, the study makes a reasonable argument that treating acute BP elevations in noncardiac patients in the hospital setting is not beneficial.

CAVEATS

Impact of existing therapy could be underestimated

This study had several important limitations. First, 23% of treated participants were excluded from the propensity analysis without justification from the authors. Additionally, there was no reporting of missing data and how it was managed. The authors’ definition of treatment excluded dose intensification of existing antihypertensive therapy, which would undercount the number of treated patients. However, this could underestimate the actual harms of the acute antihypertensive therapy. The authors also included patients with atrial fibrillation and heart failure in the study population, even though they already may have been taking antihypertensive agents.

CHALLENGES TO IMPLEMENTATION

Potential delays in translating findings to patient care

Although several recent studies have shown the potential benefit of not treating asymptomatic acute BP elevations in inpatients, incorporating that information into electronic health record order sets or clinical decision support, and disseminating it to clinical end users, will take time. In the interim, despite these findings, patients may continue to receive IV or oral medications to treat acute, asymptomatic BP elevations while hospitalized for noncardiac diagnoses.

ILLUSTRATIVE CASE

A 48-year-old man is admitted to your family medicine service for cellulitis after failed outpatient therapy. He has presumed community-acquired methicillin-resistant Staphylococcus aureus infection of the left lower extremity and is receiving intravenous (IV) vancomycin. His BP this morning is 176/98 mm Hg, and the reading from the previous shift was 168/94 mm Hg. He is asymptomatic from this elevated BP. Based on protocol, his nurse is asking about treatment in response to the multiple elevated readings. How should you address the patient’s elevated BP, knowing that you will see him for a transition management appointment in 2 weeks?

Elevated BP is common in the adult inpatient setting. Prevalence estimates range from 25% to > 50%. Many factors can contribute to elevated BP in the acute illness setting, such as pain, anxiety, medication withdrawal, and volume status.^2,3

Treatment of elevated BP in outpatients is well researched, with evidence-based guidelines for physicians. That is not the case for treatment of asymptomatic elevated BP in the inpatient setting. Most published guidance on inpatient management of acutely elevated BP recommends IV medications, such as hydralazine or labetalol, although there is limited evidence to support such recommendations. There is minimal evidence for outcomes-based benefit in treating acute elevations of inpatient BP, such as reduced myocardial injury or stroke; however, there is some evidence of adverse outcomes, such as hypotension and prolonged hospital stays.^4-8

Although the possibility of intensifying antihypertensive therapy for those with known hypertension or those with presumed “new-onset” hypertension could theoretically lead to improved outcomes over the long term, there is little evidence to support this presumption. Rather, there is evidence that intensification of antihypertensive therapy at discharge is linked to short-term harms. This was demonstrated in a propensity-matched veteran cohort that included 4056 hospitalized older adults with hypertension (mean age, 77 years; 3961 men), equally split between those who received antihypertensive intensification at hospital discharge and those who did not. Within 30 days, patients receiving intensification had a higher risk of readmission (number needed to harm [NNH] = 27) and serious adverse events (NNH = 63).⁹

The current study aimed to put all these pieces together by quantifying the prevalence of hypertension in hospitalized patients, characterizing clinician response to patients’ acutely elevated BP, and comparing both short- and long-term outcomes in patients treated for acute BP elevations while hospitalized vs those who were not. The study also assessed the potential effects of antihypertensive intensification at discharge.

STUDY SUMMARY

Treatment of acute hypertension was associated with end-organ injury

This retrospective, propensity score–matched cohort study (N = 22,834) evaluated the electronic health records of all adult patients (age > 18 years) admitted to a medicine service with a noncardiovascular diagnosis over a 1-year period at 10 Cleveland Clinic hospitals, with 1 year of follow-up data.

Exclusion criteria included hospitalization for a cardiovascular diagnosis; admission for a cerebrovascular event or acute coronary syndrome within the previous 30 days; pregnancy; length of stay of less than 2 days or more than 14 days; and lack of outpatient medication data. Patients were propensity-score matched using BP, demographic features, comorbidities, hospital shift, and time since admission. Exposure was defined as administration of IV antihypertensive medication or a new class of oral antihypertensive medication.

Continue to: Outcomes were defined...

The 22,834 patients had a mean (SD) age of 65.6 (17.9) years; 12,993 (56.9%) were women, and 15,963 (69.9%) were White. Of the 17,821 (78%) who had at least 1 inpatient hypertensive systolic BP (SBP) episode, defined as an SBP ≥ 140 mm Hg, 5904 (33.1%) received a new treatment. Of those receiving a new treatment, 4378 (74.2%) received only oral treatment, and 1516 (25.7%) received at least 1 dose of IV medication with or without oral dosing.

Acute treatment of elevated BP in noncardiac inpatients was not beneficial, and treatment intensification at discharge did not improve BP control over the following year.

Using the propensity-matched sample (4520 treated for elevated BP matched to 4520 who were not treated), treated patients had higher rates of AKI (10.3% vs 7.9%; P < .001) and myocardial injury (1.2% vs 0.6%; P = .003). When assessed by SBP, nontreatment of BP was still superior up to an SBP of 199 mm Hg. At an SBP of ≥ 200 mm Hg, there was no difference in rates of AKI or MI between the treatment and nontreatment groups. There was no difference in stroke in either cohort, although the overall numbers were quite low.

Patients with and without antihypertensive intensification at discharge had similar rates of MI (0.1% vs 0.2%; P > .99) and stroke (0.5% vs 0.4%; P > .99) in a matched cohort at 30 days post discharge. At 1 year, BP control in the intensification vs no-intensification groups was nearly the same: maximum SBP was 157.2 mm Hg vs 157.8 mm Hg, respectively (P = .54) and maximum diastolic BP was 86.5 mm Hg vs 86.1 mm Hg, respectively (P = .49).

WHAT’S NEW

Previous research is confirmed in a more diverse population

Whereas previous research showed no benefit to intensification of treatment among hospitalized older male patients, this large, retrospective, propensity score–matched cohort study demonstrated the short- and long-term effects of treating acute, asymptomatic BP elevations in a younger, more generalizable population that included women. Regardless of treatment modality, there appeared to be more harm than good from treating these BP elevations.

In addition, the study appears to corroborate previous research showing that intensification of BP treatment at discharge did not lead to better outcomes.⁹ At the very least, the study makes a reasonable argument that treating acute BP elevations in noncardiac patients in the hospital setting is not beneficial.

CAVEATS

Impact of existing therapy could be underestimated

This study had several important limitations. First, 23% of treated participants were excluded from the propensity analysis without justification from the authors. Additionally, there was no reporting of missing data and how it was managed. The authors’ definition of treatment excluded dose intensification of existing antihypertensive therapy, which would undercount the number of treated patients. However, this could underestimate the actual harms of the acute antihypertensive therapy. The authors also included patients with atrial fibrillation and heart failure in the study population, even though they already may have been taking antihypertensive agents.

CHALLENGES TO IMPLEMENTATION

Potential delays in translating findings to patient care

Although several recent studies have shown the potential benefit of not treating asymptomatic acute BP elevations in inpatients, incorporating that information into electronic health record order sets or clinical decision support, and disseminating it to clinical end users, will take time. In the interim, despite these findings, patients may continue to receive IV or oral medications to treat acute, asymptomatic BP elevations while hospitalized for noncardiac diagnoses.

ILLUSTRATIVE CASE

You are seeing a 2-month-old female infant for a routine well-child visit. The birth history was unremarkable. The infant is meeting appropriate developmental milestones. Growth is appropriate at the 40th percentile. The infant is exclusively breastfed. The parents report that they have heard confusing information about when to introduce solid foods, including bread, to their child’s diet. There is no known family history of CD. What anticipatory guidance can you offer regarding gluten introduction and the risk of CD?

CD is an inflammatory disease of the small intestine caused by an immune-based reaction to dietary gluten. The worldwide incidence of CD in children younger than 15 years is 21.3 per 100,000 person-years; this incidence has increased by 7.5% per year over the past several decades.² CD has a range of both gastrointestinal and nongastrointestinal manifestations, including diarrhea, weight loss, abdominal pain, abnormal liver function test results, and iron deficiency anemia.

Diagnosis of CD in adults is based on a combination of clinical symptoms, elevated levels of immunoglobulin A anti-tissue transglutaminase antibody (tTG-IgA), and biopsy-confirmed villous atrophy of the duodenum on upper endoscopy.³ European pediatric guidelines suggest that use of certain criteria, including very high results of tTG-IgA antibody testing (> 10 times the upper limit of normal), can help to avoid endoscopic biopsies and/or human leukocyte antigens (HLA) testing for diagnosis in children.⁴

The mainstay of CD management is strict adherence to a gluten-free diet.³ Because this can be difficult, and yield an incomplete disease response, emphasis has been placed on primary prevention by modifying introduction of dietary gluten. Multiple prior studies examining the risk of CD have failed to demonstrate a significant association between timing of gluten introduction and development of CD among high-risk infants (eg, those with HLA-DR3 alleles or first-degree relatives with CD or type 1 diabetes).^5-7 A 2016 meta-analysis concluded that there was not enough evidence to support early introduction of gluten (at 4-6 months).⁸ RCTs have not previously been conducted to examine the timing of gluten introduction on CD prevalence for infants at average risk, using age-appropriate doses of gluten prior to age 6 months.

Current dietary guidelines in the United States and the United Kingdom recommend introduction of nutrient-dense foods, including potentially allergenic foods, at about age 6 months to complement human milk or infant formula feedings.^9,10 These guidelines do not specify the exact timing or quantity of gluten- containing food introduction for infants. A 2016 position paper by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition indicated that gluten could be introduced into the infant’s diet any time between 4 and 12 months. They did indicate that the amount of gluten introduced into the diet should be low to start and then increased, and that infants at high risk for CD should wait longer for gluten introduction (4 vs 6 months or 6 vs 12 months).¹¹

STUDY SUMMARY

Gluten introduced at 4 months may be linked to lower occurrence of CD

The Enquiring About Tolerance (EAT) Study was an open-label RCT (N = 1303) with children from the general population in England and Wales. The EAT Study sought to test the prevention of food allergy by introducing allergenic foods to infants at age 4 months compared with exclusively breastfeeding until age 6 months. The median age at enrollment was 3.4 months, but allergenic food was not started until age 4 months.^1,12 Most patients were White (84.%-85.4%) and lived in an urban area (77.3%-77.4%). The mean gestational age at delivery was 39.7 to 39.9 weeks.¹²

Infants were exclusively breastfed until age 13 weeks, at which time they were randomized into an early introduction group (EIG) or a standard introduction group (SIG). In addition to breast milk, infants in the EIG consumed 6 allergenic foods (peanut, sesame, hen’s egg, cow’s milk, cod fish, and wheat [gluten]) in a specified pattern per protocol, starting at age 4 months. Wheat (gluten) was introduced during Week 5 of the EIG protocol (median age, 20.6 weeks).¹² The recommended minimum dose of gluten was 3.2 g/wk from age 16 weeks, or 4 g/wk of wheat protein (given as 2 cereal biscuits or the equivalent). Infants in the SIG avoided allergenic foods, following UK infant feeding recommendations for exclusive breastfeeding until about age 6 months. The EIG had a significantly higher rate of cesarean births than the SIG, but the study groups were otherwise balanced.¹³

Continue to: Families completed monthly...

Families completed monthly questionnaires on infant gluten intake and symptoms (eg, gastrointestinal, fatigue) through age 1 year, and then every 3 months through age 3 years. All children were tested for anti-transglutaminase type 2 (anti-TG2) antibodies at age 3 years as a screen for CD. Children with antibody levels > 20 IU/L were referred to independent gastroenterologists for further evaluation, which could include HLA (DQ-2/DQ-8) testing and biopsy in accordance with current European diagnostic guidelines.⁴

Introducing gluten as a complement to breast milk or infant formula from age 4 months may reduce the risk of celiac disease at age 3 years.

In an intention-to-treat analysis for the primary outcome, 595 children in the SIG (91.4%) and 567 in the EIG (87.0%) were included. Between ages 4 and 6 months, the mean (SD) quantity of gluten consumed in the SIG was 0.49 (1.40) g/wk; in the EIG, the mean quantity was 2.66 (1.85) g/wk (P < .001). At age 3 years, of a total of 1004 children tested for anti-TG2 antibodies, 9 had anti-TG2 levels requiring referral (7 in the SIG and 2 in the EIG). A diagnosis of CD was confirmed in 7 of 516 children in the SIG (1.4%) vs none of the 488 children in the EIG (P = .02). Using bootstrap resampling, the risk difference between the groups was 1.4% (95% CI, 0.6%-2.6%).

WHAT’S NEW

Findings have potential to change nutritional guidance

This study demonstrated that introduction of age-appropriate portions of gluten-containing products at age 4 months, in addition to breast milk, may reduce the risk of CD at 3 years in children at average risk. This finding has the potential to change anticipatory guidance given to parents regarding infant nutrition recommendations.

CAVEATS

More studies needed to confirm prevention vs delay of CD

The homogeneous study population may limit generalizability. Infants in this study were from England and Wales (84.3% were White), born at term, and were exclusively breastfed until age 13 weeks. Further studies are required to determine whether these findings can be applied to infants who are no longer breastfeeding, are more racially diverse, or are preterm in gestational age at birth. Additionally, the study followed the participants only until age 3 years. Given that the onset of CD after this age is likely, further research is needed to support that CD is truly prevented rather than delayed.

CHALLENGES TO IMPLEMENTATION

Guidance on allergen introduction may be unclear

The EAT Study protocol required parents in the EIG to sequentially introduce a minimum amount of the 6 allergenic foods specified. Only 42% of the EIG cohort reported adherence to the protocol.¹² It is unclear how important this specific regimen is to the study results and whether introduction of all 6 allergenic foods simultaneously modifies the immune response to gluten. Therefore, there may be challenges to implementation if physicians do not know how to provide anticipatory guidance on the appropriate steps for allergen introduction.

ILLUSTRATIVE CASE

You are seeing a 2-month-old female infant for a routine well-child visit. The birth history was unremarkable. The infant is meeting appropriate developmental milestones. Growth is appropriate at the 40th percentile. The infant is exclusively breastfed. The parents report that they have heard confusing information about when to introduce solid foods, including bread, to their child’s diet. There is no known family history of CD. What anticipatory guidance can you offer regarding gluten introduction and the risk of CD?

CD is an inflammatory disease of the small intestine caused by an immune-based reaction to dietary gluten. The worldwide incidence of CD in children younger than 15 years is 21.3 per 100,000 person-years; this incidence has increased by 7.5% per year over the past several decades.² CD has a range of both gastrointestinal and nongastrointestinal manifestations, including diarrhea, weight loss, abdominal pain, abnormal liver function test results, and iron deficiency anemia.

Diagnosis of CD in adults is based on a combination of clinical symptoms, elevated levels of immunoglobulin A anti-tissue transglutaminase antibody (tTG-IgA), and biopsy-confirmed villous atrophy of the duodenum on upper endoscopy.³ European pediatric guidelines suggest that use of certain criteria, including very high results of tTG-IgA antibody testing (> 10 times the upper limit of normal), can help to avoid endoscopic biopsies and/or human leukocyte antigens (HLA) testing for diagnosis in children.⁴

The mainstay of CD management is strict adherence to a gluten-free diet.³ Because this can be difficult, and yield an incomplete disease response, emphasis has been placed on primary prevention by modifying introduction of dietary gluten. Multiple prior studies examining the risk of CD have failed to demonstrate a significant association between timing of gluten introduction and development of CD among high-risk infants (eg, those with HLA-DR3 alleles or first-degree relatives with CD or type 1 diabetes).^5-7 A 2016 meta-analysis concluded that there was not enough evidence to support early introduction of gluten (at 4-6 months).⁸ RCTs have not previously been conducted to examine the timing of gluten introduction on CD prevalence for infants at average risk, using age-appropriate doses of gluten prior to age 6 months.

Current dietary guidelines in the United States and the United Kingdom recommend introduction of nutrient-dense foods, including potentially allergenic foods, at about age 6 months to complement human milk or infant formula feedings.^9,10 These guidelines do not specify the exact timing or quantity of gluten- containing food introduction for infants. A 2016 position paper by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition indicated that gluten could be introduced into the infant’s diet any time between 4 and 12 months. They did indicate that the amount of gluten introduced into the diet should be low to start and then increased, and that infants at high risk for CD should wait longer for gluten introduction (4 vs 6 months or 6 vs 12 months).¹¹

STUDY SUMMARY

Gluten introduced at 4 months may be linked to lower occurrence of CD

The Enquiring About Tolerance (EAT) Study was an open-label RCT (N = 1303) with children from the general population in England and Wales. The EAT Study sought to test the prevention of food allergy by introducing allergenic foods to infants at age 4 months compared with exclusively breastfeeding until age 6 months. The median age at enrollment was 3.4 months, but allergenic food was not started until age 4 months.^1,12 Most patients were White (84.%-85.4%) and lived in an urban area (77.3%-77.4%). The mean gestational age at delivery was 39.7 to 39.9 weeks.¹²

Infants were exclusively breastfed until age 13 weeks, at which time they were randomized into an early introduction group (EIG) or a standard introduction group (SIG). In addition to breast milk, infants in the EIG consumed 6 allergenic foods (peanut, sesame, hen’s egg, cow’s milk, cod fish, and wheat [gluten]) in a specified pattern per protocol, starting at age 4 months. Wheat (gluten) was introduced during Week 5 of the EIG protocol (median age, 20.6 weeks).¹² The recommended minimum dose of gluten was 3.2 g/wk from age 16 weeks, or 4 g/wk of wheat protein (given as 2 cereal biscuits or the equivalent). Infants in the SIG avoided allergenic foods, following UK infant feeding recommendations for exclusive breastfeeding until about age 6 months. The EIG had a significantly higher rate of cesarean births than the SIG, but the study groups were otherwise balanced.¹³

Continue to: Families completed monthly...

Families completed monthly questionnaires on infant gluten intake and symptoms (eg, gastrointestinal, fatigue) through age 1 year, and then every 3 months through age 3 years. All children were tested for anti-transglutaminase type 2 (anti-TG2) antibodies at age 3 years as a screen for CD. Children with antibody levels > 20 IU/L were referred to independent gastroenterologists for further evaluation, which could include HLA (DQ-2/DQ-8) testing and biopsy in accordance with current European diagnostic guidelines.⁴

Introducing gluten as a complement to breast milk or infant formula from age 4 months may reduce the risk of celiac disease at age 3 years.

In an intention-to-treat analysis for the primary outcome, 595 children in the SIG (91.4%) and 567 in the EIG (87.0%) were included. Between ages 4 and 6 months, the mean (SD) quantity of gluten consumed in the SIG was 0.49 (1.40) g/wk; in the EIG, the mean quantity was 2.66 (1.85) g/wk (P < .001). At age 3 years, of a total of 1004 children tested for anti-TG2 antibodies, 9 had anti-TG2 levels requiring referral (7 in the SIG and 2 in the EIG). A diagnosis of CD was confirmed in 7 of 516 children in the SIG (1.4%) vs none of the 488 children in the EIG (P = .02). Using bootstrap resampling, the risk difference between the groups was 1.4% (95% CI, 0.6%-2.6%).

WHAT’S NEW

Findings have potential to change nutritional guidance

This study demonstrated that introduction of age-appropriate portions of gluten-containing products at age 4 months, in addition to breast milk, may reduce the risk of CD at 3 years in children at average risk. This finding has the potential to change anticipatory guidance given to parents regarding infant nutrition recommendations.

CAVEATS

More studies needed to confirm prevention vs delay of CD

The homogeneous study population may limit generalizability. Infants in this study were from England and Wales (84.3% were White), born at term, and were exclusively breastfed until age 13 weeks. Further studies are required to determine whether these findings can be applied to infants who are no longer breastfeeding, are more racially diverse, or are preterm in gestational age at birth. Additionally, the study followed the participants only until age 3 years. Given that the onset of CD after this age is likely, further research is needed to support that CD is truly prevented rather than delayed.

CHALLENGES TO IMPLEMENTATION

Guidance on allergen introduction may be unclear

The EAT Study protocol required parents in the EIG to sequentially introduce a minimum amount of the 6 allergenic foods specified. Only 42% of the EIG cohort reported adherence to the protocol.¹² It is unclear how important this specific regimen is to the study results and whether introduction of all 6 allergenic foods simultaneously modifies the immune response to gluten. Therefore, there may be challenges to implementation if physicians do not know how to provide anticipatory guidance on the appropriate steps for allergen introduction.

ILLUSTRATIVE CASE

You are seeing a 2-month-old female infant for a routine well-child visit. The birth history was unremarkable. The infant is meeting appropriate developmental milestones. Growth is appropriate at the 40th percentile. The infant is exclusively breastfed. The parents report that they have heard confusing information about when to introduce solid foods, including bread, to their child’s diet. There is no known family history of CD. What anticipatory guidance can you offer regarding gluten introduction and the risk of CD?

CD is an inflammatory disease of the small intestine caused by an immune-based reaction to dietary gluten. The worldwide incidence of CD in children younger than 15 years is 21.3 per 100,000 person-years; this incidence has increased by 7.5% per year over the past several decades.² CD has a range of both gastrointestinal and nongastrointestinal manifestations, including diarrhea, weight loss, abdominal pain, abnormal liver function test results, and iron deficiency anemia.

Diagnosis of CD in adults is based on a combination of clinical symptoms, elevated levels of immunoglobulin A anti-tissue transglutaminase antibody (tTG-IgA), and biopsy-confirmed villous atrophy of the duodenum on upper endoscopy.³ European pediatric guidelines suggest that use of certain criteria, including very high results of tTG-IgA antibody testing (> 10 times the upper limit of normal), can help to avoid endoscopic biopsies and/or human leukocyte antigens (HLA) testing for diagnosis in children.⁴

The mainstay of CD management is strict adherence to a gluten-free diet.³ Because this can be difficult, and yield an incomplete disease response, emphasis has been placed on primary prevention by modifying introduction of dietary gluten. Multiple prior studies examining the risk of CD have failed to demonstrate a significant association between timing of gluten introduction and development of CD among high-risk infants (eg, those with HLA-DR3 alleles or first-degree relatives with CD or type 1 diabetes).^5-7 A 2016 meta-analysis concluded that there was not enough evidence to support early introduction of gluten (at 4-6 months).⁸ RCTs have not previously been conducted to examine the timing of gluten introduction on CD prevalence for infants at average risk, using age-appropriate doses of gluten prior to age 6 months.

Current dietary guidelines in the United States and the United Kingdom recommend introduction of nutrient-dense foods, including potentially allergenic foods, at about age 6 months to complement human milk or infant formula feedings.^9,10 These guidelines do not specify the exact timing or quantity of gluten- containing food introduction for infants. A 2016 position paper by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition indicated that gluten could be introduced into the infant’s diet any time between 4 and 12 months. They did indicate that the amount of gluten introduced into the diet should be low to start and then increased, and that infants at high risk for CD should wait longer for gluten introduction (4 vs 6 months or 6 vs 12 months).¹¹

STUDY SUMMARY

Gluten introduced at 4 months may be linked to lower occurrence of CD

The Enquiring About Tolerance (EAT) Study was an open-label RCT (N = 1303) with children from the general population in England and Wales. The EAT Study sought to test the prevention of food allergy by introducing allergenic foods to infants at age 4 months compared with exclusively breastfeeding until age 6 months. The median age at enrollment was 3.4 months, but allergenic food was not started until age 4 months.^1,12 Most patients were White (84.%-85.4%) and lived in an urban area (77.3%-77.4%). The mean gestational age at delivery was 39.7 to 39.9 weeks.¹²

Infants were exclusively breastfed until age 13 weeks, at which time they were randomized into an early introduction group (EIG) or a standard introduction group (SIG). In addition to breast milk, infants in the EIG consumed 6 allergenic foods (peanut, sesame, hen’s egg, cow’s milk, cod fish, and wheat [gluten]) in a specified pattern per protocol, starting at age 4 months. Wheat (gluten) was introduced during Week 5 of the EIG protocol (median age, 20.6 weeks).¹² The recommended minimum dose of gluten was 3.2 g/wk from age 16 weeks, or 4 g/wk of wheat protein (given as 2 cereal biscuits or the equivalent). Infants in the SIG avoided allergenic foods, following UK infant feeding recommendations for exclusive breastfeeding until about age 6 months. The EIG had a significantly higher rate of cesarean births than the SIG, but the study groups were otherwise balanced.¹³

Continue to: Families completed monthly...

Families completed monthly questionnaires on infant gluten intake and symptoms (eg, gastrointestinal, fatigue) through age 1 year, and then every 3 months through age 3 years. All children were tested for anti-transglutaminase type 2 (anti-TG2) antibodies at age 3 years as a screen for CD. Children with antibody levels > 20 IU/L were referred to independent gastroenterologists for further evaluation, which could include HLA (DQ-2/DQ-8) testing and biopsy in accordance with current European diagnostic guidelines.⁴

Introducing gluten as a complement to breast milk or infant formula from age 4 months may reduce the risk of celiac disease at age 3 years.

In an intention-to-treat analysis for the primary outcome, 595 children in the SIG (91.4%) and 567 in the EIG (87.0%) were included. Between ages 4 and 6 months, the mean (SD) quantity of gluten consumed in the SIG was 0.49 (1.40) g/wk; in the EIG, the mean quantity was 2.66 (1.85) g/wk (P < .001). At age 3 years, of a total of 1004 children tested for anti-TG2 antibodies, 9 had anti-TG2 levels requiring referral (7 in the SIG and 2 in the EIG). A diagnosis of CD was confirmed in 7 of 516 children in the SIG (1.4%) vs none of the 488 children in the EIG (P = .02). Using bootstrap resampling, the risk difference between the groups was 1.4% (95% CI, 0.6%-2.6%).

WHAT’S NEW

Findings have potential to change nutritional guidance

This study demonstrated that introduction of age-appropriate portions of gluten-containing products at age 4 months, in addition to breast milk, may reduce the risk of CD at 3 years in children at average risk. This finding has the potential to change anticipatory guidance given to parents regarding infant nutrition recommendations.

CAVEATS

More studies needed to confirm prevention vs delay of CD

The homogeneous study population may limit generalizability. Infants in this study were from England and Wales (84.3% were White), born at term, and were exclusively breastfed until age 13 weeks. Further studies are required to determine whether these findings can be applied to infants who are no longer breastfeeding, are more racially diverse, or are preterm in gestational age at birth. Additionally, the study followed the participants only until age 3 years. Given that the onset of CD after this age is likely, further research is needed to support that CD is truly prevented rather than delayed.

CHALLENGES TO IMPLEMENTATION

Guidance on allergen introduction may be unclear

The EAT Study protocol required parents in the EIG to sequentially introduce a minimum amount of the 6 allergenic foods specified. Only 42% of the EIG cohort reported adherence to the protocol.¹² It is unclear how important this specific regimen is to the study results and whether introduction of all 6 allergenic foods simultaneously modifies the immune response to gluten. Therefore, there may be challenges to implementation if physicians do not know how to provide anticipatory guidance on the appropriate steps for allergen introduction.

ILLUSTRATIVE CASE

A 61-year-old man with hypertension and paroxysmal AF presents to your office shortly after experiencing his third episode of AF in the past 6 months. He describes these episodes, which last for several days, as “just awful,” noting that when he experiences AF, he has fatigue, palpitations, and shortness of breath and “can’t stop paying attention to my heart.” The patient, who has a body mass index of 32, consumes more than 15 alcoholic drinks per week. What can you recommend to him that will decrease his likelihood of experiencing more episodes of AF?

AF is the most common sustained cardiac arrhythmia. It is associated with significant morbidity and mortality. Known risk factors include obesity, physical inactivity, sleep apnea, diabetes, and hypertension.²

According to the Centers for Disease Control and Prevention, an estimated 12.1 million people in the United States will have AF by 2030. In 2018, AF was mentioned on more than 183,000 death certificates and was the underlying cause of more than 26,000 of those deaths.³ AF is the primary diagnosis in 450,000 hospitalizations annually,⁴ and the death rate from AF as the primary or contributing cause of death has been rising for more than 2 decades.³

More than 50% of Americans report alcohol consumption within the past month.⁵ Although alcohol use is associated with new and recurrent AF, only limited prospective data show a clear and causal association between abstaining from alcohol and decreasing AF recurrence.

STUDY SUMMARY

Reduction in AF recurrence and total AF burden following alcohol abstinence

This multicenter, prospective, open-label, randomized controlled trial (N = 140) from 6 sites in Australia evaluated the impact of alcohol abstinence on both the recurrence of AF and the amount of time in AF. Study participants were ages 18 to 85 years, consumed 10 or more standard alcohol-containing drinks per week, had paroxysmal or persistent AF, and were in sinus rhythm at the time of enrollment, regardless of antiarrhythmic therapy. Exclusion criteria included alcohol dependence or abuse, severe left ventricular systolic dysfunction (ejection fraction < 35%), clinically significant noncardiac illness, and/or coexisting psychiatric disorder.¹

After a 4-week run-in period, patients were randomized to either an abstinence or a control group in a 1:1 fashion. Patients enrolled in the abstinence group were encouraged to abstain from alcohol consumption for 6 months and were provided with written and oral instructions to assist with abstaining. Control group patients continued their same level of alcohol consumption. Comprehensive rhythm monitoring occurred for all patients after randomization.

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption.

Alcohol consumption was reported by both groups using a weekly alcohol diary, supplemented with a visual guide showing pictures of standard alcohol drinks. For the abstinence group, random urine testing for ethyl glucuronide (an alcohol metabolite) was possible if no alcohol intake was reported. Primary outcomes during the 6-month study included recurrence of AF and total AF burden (percentage of time in AF).

Continue to: Secondary outcomes included hospitalizations...

Secondary outcomes included hospitalizations for AF, AF symptom severity, and change in weight. Blood pressure, quality-of-life, and depression scores were missing for > 35% of patients.¹

Patients were randomized evenly to the control and abstinence groups. The typical patient was an overweight male in his early 60s with paroxysmal AF, who was taking an antiarrhythmic agent. Patients in the abstinence group decreased their alcohol consumption from 16.8 to 2.1 drinks per week (87.5% reduction; mean difference = –14.7; 95% CI, –12.7 to –16.7). Patients in the control group reduced their intake from 16.4 to 13.2 drinks per week (19.5% reduction; mean difference = –3.2; 95% CI, –1.9 to –4.4).¹

AF recurred in 53% vs 73% of the abstinence and control groups, respectively, with a longer period before recurrence in the abstinence group than in the control group (hazard ratio = 0.55; 95% CI, 0.36-0.84; P = .005; number needed to treat = 5). The AF burden was also lower in the abstinence group (0.5%; interquartile range [IQR] = 0.0-3.0) than in the control group (1.2%; IQR = 0.0-10.3; P = .01). The abstinence group had a lower percentage of AF hospitalizations compared with the control group (9% vs 20%), and fewer patients reporting moderate or severe AF symptoms (10% vs 32%). In addition, the abstinence group lost 3.7 kg more weight than did the control group at 6 months.¹

WHAT’S NEW

Objective new evidence for effective patient counseling

Alcohol consumption and its association with the onset and recurrence of AF has been documented previously.⁶ This study was the first to prospectively examine if abstaining from alcohol reduces paroxysmal AF episodes in moderate drinkers.

The study identified clinically meaningful findings among those who abstained from alcohol, including decreased AF recurrence rates, increased time to recurrence, and lower overall AF burden. This provides objective evidence that can be used for motivational interviewing in patients with paroxysmal AF who may be receptive to reducing or abstaining from alcohol consumption.

Continue to: CAVEATS

The narrow study population may not be widely applicable

The study population was predominantly male, in their seventh decade of life (mean age, 61), and living in Australia. Rates of AF and symptomatology differ by gender and age, making this information challenging to apply to women or older populations. The study excluded patients with alcohol dependence or abuse, left ventricular systolic dysfunction (ejection fraction < 35%), coexisting psychiatric disorders, and clinically significant noncardiac illnesses, limiting the study’s generalizability to these patient populations. Overall, AF recurrence was low in both groups despite the intervention, and the study did not evaluate the efficacy of the counseling method for abstinence.

Since publication of this article, a prospective cohort study of approximately 3800 Swiss patients with AF evaluated the effect of alcohol consumption on the rate of stroke and embolic events. That study did not find statistically significant correlations between patients who drank no alcohol per day, > 0 to < 1, 1 to < 2, or ≥ 2 drinks per day and their rate of stroke.⁷ However, this study did not specifically evaluate the rate of AF recurrence or time spent in AF among the cohort, which is clinically meaningful for patient morbidity.¹

CHALLENGES TO IMPLEMENTATION

Patient willingness to cut alcohol consumption may be limited

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption. In this study population, 697 patients were screened for enrollment and met inclusion criteria; however, 491 patients (70.4%) were not willing to consider abstinence from alcohol, and after the run-in phase, another 17 declined randomization. Many primary care physicians would likely agree that while it is easy to encourage patients to drink less, patient adherence to these recommendations, particularly abstaining, is likely to be limited.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 61-year-old man with hypertension and paroxysmal AF presents to your office shortly after experiencing his third episode of AF in the past 6 months. He describes these episodes, which last for several days, as “just awful,” noting that when he experiences AF, he has fatigue, palpitations, and shortness of breath and “can’t stop paying attention to my heart.” The patient, who has a body mass index of 32, consumes more than 15 alcoholic drinks per week. What can you recommend to him that will decrease his likelihood of experiencing more episodes of AF?

AF is the most common sustained cardiac arrhythmia. It is associated with significant morbidity and mortality. Known risk factors include obesity, physical inactivity, sleep apnea, diabetes, and hypertension.²

According to the Centers for Disease Control and Prevention, an estimated 12.1 million people in the United States will have AF by 2030. In 2018, AF was mentioned on more than 183,000 death certificates and was the underlying cause of more than 26,000 of those deaths.³ AF is the primary diagnosis in 450,000 hospitalizations annually,⁴ and the death rate from AF as the primary or contributing cause of death has been rising for more than 2 decades.³

More than 50% of Americans report alcohol consumption within the past month.⁵ Although alcohol use is associated with new and recurrent AF, only limited prospective data show a clear and causal association between abstaining from alcohol and decreasing AF recurrence.

STUDY SUMMARY

Reduction in AF recurrence and total AF burden following alcohol abstinence

This multicenter, prospective, open-label, randomized controlled trial (N = 140) from 6 sites in Australia evaluated the impact of alcohol abstinence on both the recurrence of AF and the amount of time in AF. Study participants were ages 18 to 85 years, consumed 10 or more standard alcohol-containing drinks per week, had paroxysmal or persistent AF, and were in sinus rhythm at the time of enrollment, regardless of antiarrhythmic therapy. Exclusion criteria included alcohol dependence or abuse, severe left ventricular systolic dysfunction (ejection fraction < 35%), clinically significant noncardiac illness, and/or coexisting psychiatric disorder.¹

After a 4-week run-in period, patients were randomized to either an abstinence or a control group in a 1:1 fashion. Patients enrolled in the abstinence group were encouraged to abstain from alcohol consumption for 6 months and were provided with written and oral instructions to assist with abstaining. Control group patients continued their same level of alcohol consumption. Comprehensive rhythm monitoring occurred for all patients after randomization.

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption.

Alcohol consumption was reported by both groups using a weekly alcohol diary, supplemented with a visual guide showing pictures of standard alcohol drinks. For the abstinence group, random urine testing for ethyl glucuronide (an alcohol metabolite) was possible if no alcohol intake was reported. Primary outcomes during the 6-month study included recurrence of AF and total AF burden (percentage of time in AF).

Continue to: Secondary outcomes included hospitalizations...

Secondary outcomes included hospitalizations for AF, AF symptom severity, and change in weight. Blood pressure, quality-of-life, and depression scores were missing for > 35% of patients.¹

Patients were randomized evenly to the control and abstinence groups. The typical patient was an overweight male in his early 60s with paroxysmal AF, who was taking an antiarrhythmic agent. Patients in the abstinence group decreased their alcohol consumption from 16.8 to 2.1 drinks per week (87.5% reduction; mean difference = –14.7; 95% CI, –12.7 to –16.7). Patients in the control group reduced their intake from 16.4 to 13.2 drinks per week (19.5% reduction; mean difference = –3.2; 95% CI, –1.9 to –4.4).¹

AF recurred in 53% vs 73% of the abstinence and control groups, respectively, with a longer period before recurrence in the abstinence group than in the control group (hazard ratio = 0.55; 95% CI, 0.36-0.84; P = .005; number needed to treat = 5). The AF burden was also lower in the abstinence group (0.5%; interquartile range [IQR] = 0.0-3.0) than in the control group (1.2%; IQR = 0.0-10.3; P = .01). The abstinence group had a lower percentage of AF hospitalizations compared with the control group (9% vs 20%), and fewer patients reporting moderate or severe AF symptoms (10% vs 32%). In addition, the abstinence group lost 3.7 kg more weight than did the control group at 6 months.¹

WHAT’S NEW

Objective new evidence for effective patient counseling

Alcohol consumption and its association with the onset and recurrence of AF has been documented previously.⁶ This study was the first to prospectively examine if abstaining from alcohol reduces paroxysmal AF episodes in moderate drinkers.

The study identified clinically meaningful findings among those who abstained from alcohol, including decreased AF recurrence rates, increased time to recurrence, and lower overall AF burden. This provides objective evidence that can be used for motivational interviewing in patients with paroxysmal AF who may be receptive to reducing or abstaining from alcohol consumption.

Continue to: CAVEATS

The narrow study population may not be widely applicable

The study population was predominantly male, in their seventh decade of life (mean age, 61), and living in Australia. Rates of AF and symptomatology differ by gender and age, making this information challenging to apply to women or older populations. The study excluded patients with alcohol dependence or abuse, left ventricular systolic dysfunction (ejection fraction < 35%), coexisting psychiatric disorders, and clinically significant noncardiac illnesses, limiting the study’s generalizability to these patient populations. Overall, AF recurrence was low in both groups despite the intervention, and the study did not evaluate the efficacy of the counseling method for abstinence.

Since publication of this article, a prospective cohort study of approximately 3800 Swiss patients with AF evaluated the effect of alcohol consumption on the rate of stroke and embolic events. That study did not find statistically significant correlations between patients who drank no alcohol per day, > 0 to < 1, 1 to < 2, or ≥ 2 drinks per day and their rate of stroke.⁷ However, this study did not specifically evaluate the rate of AF recurrence or time spent in AF among the cohort, which is clinically meaningful for patient morbidity.¹

CHALLENGES TO IMPLEMENTATION

Patient willingness to cut alcohol consumption may be limited

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption. In this study population, 697 patients were screened for enrollment and met inclusion criteria; however, 491 patients (70.4%) were not willing to consider abstinence from alcohol, and after the run-in phase, another 17 declined randomization. Many primary care physicians would likely agree that while it is easy to encourage patients to drink less, patient adherence to these recommendations, particularly abstaining, is likely to be limited.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 61-year-old man with hypertension and paroxysmal AF presents to your office shortly after experiencing his third episode of AF in the past 6 months. He describes these episodes, which last for several days, as “just awful,” noting that when he experiences AF, he has fatigue, palpitations, and shortness of breath and “can’t stop paying attention to my heart.” The patient, who has a body mass index of 32, consumes more than 15 alcoholic drinks per week. What can you recommend to him that will decrease his likelihood of experiencing more episodes of AF?

AF is the most common sustained cardiac arrhythmia. It is associated with significant morbidity and mortality. Known risk factors include obesity, physical inactivity, sleep apnea, diabetes, and hypertension.²

According to the Centers for Disease Control and Prevention, an estimated 12.1 million people in the United States will have AF by 2030. In 2018, AF was mentioned on more than 183,000 death certificates and was the underlying cause of more than 26,000 of those deaths.³ AF is the primary diagnosis in 450,000 hospitalizations annually,⁴ and the death rate from AF as the primary or contributing cause of death has been rising for more than 2 decades.³

More than 50% of Americans report alcohol consumption within the past month.⁵ Although alcohol use is associated with new and recurrent AF, only limited prospective data show a clear and causal association between abstaining from alcohol and decreasing AF recurrence.

STUDY SUMMARY

Reduction in AF recurrence and total AF burden following alcohol abstinence

This multicenter, prospective, open-label, randomized controlled trial (N = 140) from 6 sites in Australia evaluated the impact of alcohol abstinence on both the recurrence of AF and the amount of time in AF. Study participants were ages 18 to 85 years, consumed 10 or more standard alcohol-containing drinks per week, had paroxysmal or persistent AF, and were in sinus rhythm at the time of enrollment, regardless of antiarrhythmic therapy. Exclusion criteria included alcohol dependence or abuse, severe left ventricular systolic dysfunction (ejection fraction < 35%), clinically significant noncardiac illness, and/or coexisting psychiatric disorder.¹

After a 4-week run-in period, patients were randomized to either an abstinence or a control group in a 1:1 fashion. Patients enrolled in the abstinence group were encouraged to abstain from alcohol consumption for 6 months and were provided with written and oral instructions to assist with abstaining. Control group patients continued their same level of alcohol consumption. Comprehensive rhythm monitoring occurred for all patients after randomization.

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption.

Alcohol consumption was reported by both groups using a weekly alcohol diary, supplemented with a visual guide showing pictures of standard alcohol drinks. For the abstinence group, random urine testing for ethyl glucuronide (an alcohol metabolite) was possible if no alcohol intake was reported. Primary outcomes during the 6-month study included recurrence of AF and total AF burden (percentage of time in AF).

Continue to: Secondary outcomes included hospitalizations...

Secondary outcomes included hospitalizations for AF, AF symptom severity, and change in weight. Blood pressure, quality-of-life, and depression scores were missing for > 35% of patients.¹

Patients were randomized evenly to the control and abstinence groups. The typical patient was an overweight male in his early 60s with paroxysmal AF, who was taking an antiarrhythmic agent. Patients in the abstinence group decreased their alcohol consumption from 16.8 to 2.1 drinks per week (87.5% reduction; mean difference = –14.7; 95% CI, –12.7 to –16.7). Patients in the control group reduced their intake from 16.4 to 13.2 drinks per week (19.5% reduction; mean difference = –3.2; 95% CI, –1.9 to –4.4).¹

AF recurred in 53% vs 73% of the abstinence and control groups, respectively, with a longer period before recurrence in the abstinence group than in the control group (hazard ratio = 0.55; 95% CI, 0.36-0.84; P = .005; number needed to treat = 5). The AF burden was also lower in the abstinence group (0.5%; interquartile range [IQR] = 0.0-3.0) than in the control group (1.2%; IQR = 0.0-10.3; P = .01). The abstinence group had a lower percentage of AF hospitalizations compared with the control group (9% vs 20%), and fewer patients reporting moderate or severe AF symptoms (10% vs 32%). In addition, the abstinence group lost 3.7 kg more weight than did the control group at 6 months.¹

WHAT’S NEW

Objective new evidence for effective patient counseling

Alcohol consumption and its association with the onset and recurrence of AF has been documented previously.⁶ This study was the first to prospectively examine if abstaining from alcohol reduces paroxysmal AF episodes in moderate drinkers.

The study identified clinically meaningful findings among those who abstained from alcohol, including decreased AF recurrence rates, increased time to recurrence, and lower overall AF burden. This provides objective evidence that can be used for motivational interviewing in patients with paroxysmal AF who may be receptive to reducing or abstaining from alcohol consumption.

Continue to: CAVEATS

The narrow study population may not be widely applicable

The study population was predominantly male, in their seventh decade of life (mean age, 61), and living in Australia. Rates of AF and symptomatology differ by gender and age, making this information challenging to apply to women or older populations. The study excluded patients with alcohol dependence or abuse, left ventricular systolic dysfunction (ejection fraction < 35%), coexisting psychiatric disorders, and clinically significant noncardiac illnesses, limiting the study’s generalizability to these patient populations. Overall, AF recurrence was low in both groups despite the intervention, and the study did not evaluate the efficacy of the counseling method for abstinence.

Since publication of this article, a prospective cohort study of approximately 3800 Swiss patients with AF evaluated the effect of alcohol consumption on the rate of stroke and embolic events. That study did not find statistically significant correlations between patients who drank no alcohol per day, > 0 to < 1, 1 to < 2, or ≥ 2 drinks per day and their rate of stroke.⁷ However, this study did not specifically evaluate the rate of AF recurrence or time spent in AF among the cohort, which is clinically meaningful for patient morbidity.¹

CHALLENGES TO IMPLEMENTATION

Patient willingness to cut alcohol consumption may be limited

The largest challenge to implementation of this intervention is most likely the willingness of patients to cut their alcohol consumption. In this study population, 697 patients were screened for enrollment and met inclusion criteria; however, 491 patients (70.4%) were not willing to consider abstinence from alcohol, and after the run-in phase, another 17 declined randomization. Many primary care physicians would likely agree that while it is easy to encourage patients to drink less, patient adherence to these recommendations, particularly abstaining, is likely to be limited.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 50-year-old woman presents to your office for a well-woman exam. Her past medical history includes a 22-pack-year smoking history (she quit 5 years ago), well-controlled hypertension, and mild obesity. She has no family history of cancer, but she does have a family history of type 2 diabetes and heart disease. Besides age- and risk-appropriate laboratory tests, cervical cancer screening, breast cancer screening, and initial colon cancer screening, are there any other preventive services you would offer her?

Lung cancer is the second most common cancer in both men and women, and it is the leading cause of cancer death in the United States—regardless of gender. The American Cancer Society estimates that 235,760 people will be diagnosed with lung cancer and 131,880 people will die of the disease in 2021.²

In the 2015 National Cancer Institute report on the economic costs of cancer, direct and indirect costs of lung cancer totaled $21.1 billion annually. Lost productivity from lung cancer added another $36.1 billion in annual costs.³ The economic costs increased to $23.8 billion in 2020, with no data on lost productivity.⁴

Smoking tobacco is by far the primary risk factor for lung cancer, and it is estimated to account for 90% of all lung cancer cases. Compared with nonsmokers, the relative risk of lung cancer is approximately 20 times higher for smokers.^5,6

Because the median age of lung cancer diagnosis is 70 years, increasing age is also considered a risk factor for lung cancer.^2,7

Although lung cancer has a relatively poor prognosis—with an average 5-year survival rate of 20.5%—early-stage lung cancer is more amenable to treatment and has a better prognosis (as is true with many cancers).¹

LDCT has a high sensitivity, as well as a reasonable specificity, for lung cancer detection. There is demonstrated benefit in screening patients who are at high risk for lung cancer.^8-11 In 2013, the USPSTF recommended annual lung cancer screening (B recommendation) with LDCT in adults 55 to 80 years of age who have a 30-pack-year smoking history, and who currently smoke or quit within the past 15 years.¹

Continue to: STUDY SUMMARY

Broader eligibility for screening supports mortality benefit

This is an update to the 2013 clinical practice guideline on lung cancer screening. The USPSTF used 2 methods to provide the best possible evidence for the recommendations. The first method was a systematic review of the accuracy of screening for lung cancer with LDCT, evaluating both the benefits and harms of lung cancer screening. The systematic review examined various subgroups, the number and/or frequency of LDCT scans, and various approaches to reducing false-positive results. In addition to the systematic review, they used collaborative modeling studies to determine the optimal age for beginning and ending screening, the optimal screening interval, and the relative benefits and harms of various screening strategies. These modeling studies complemented the evidence review.

This updated guideline nearly doubles eligibility for lung cancer screening using low-dose CT scanning.

The review included 7 randomized controlled trials (RCTs), plus the modeling studies. Only the National Lung Screening Trial (NLST; N = 53,454) and the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trial (N = 15,792) had adequate power to detect a mortality benefit from screening (NLST: relative risk reduction = 16%; 95% CI, 5%-25%; NELSON: incidence rate ratio = 0.75; 95% CI, 0.61-0.90) compared with no screening.

Screening intervals, from the NLST and NELSON trials as well as the modeling studies, revealed the greatest benefit from annual screening (statistics not shared). Evidence also showed that screening those with lighter smoking histories (< 30 pack-years) and at an earlier age (age 50) provided increased mortality benefit. No evidence was found for a benefit of screening past 80 years of age. The modeling studies concluded that the 2013 USPSTF screening program, using a starting age of 55 and a 30-pack-year smoking history, would reduce mortality by 9.8%, but by changing to a starting age of 50, a 20-pack-year smoking history, and annual screening, the mortality benefit was increased to 13%.^1,11

Comparison with computer-based risk prediction models from the Cancer Intervention and Surveillance Modeling Network (CISNET) revealed insufficient evidence at this time to show that prediction model–based screening offered any benefit beyond that of the age and smoking history risk factor model.

The incidence of false-positive results was > 25% in the NLST at baseline and at 1 year. Use of a classification system such as the Lung Imaging Reporting and Data System (Lung-RADS) could reduce that from 26.6% to 12.8%.² Another potential harm from LDCT screening is radiation exposure. Evidence from several RCTs and cohort studies showed the exposure from 1 LDCT scan to be 0.65 to 2.36 mSv, whereas the annual background radiation in the United States is 2.4 mSv. The modeling studies estimated that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

Continue to: WHAT'S NEW

Expanded age range, reduced pack-year history

Annual lung cancer screening is now recommended to begin for patients at age 50 years with a 20-pack-year history instead of age 55 years with a 30-pack-year history. This would nearly double (87% overall) the number of people eligible for screening, and it would include more Black patients and women, who tend to smoke fewer cigarettes than their White male counterparts. The American College of Radiology estimates that the expanded screening criteria could save between 30,000 and 60,000 lives per year.¹²

CAVEATS

Screening criteria for upper age limit, years since smoking remain unchanged

For those patients who quit smoking, the guidelines apply only to those who have stopped smoking within the past 15 years. Furthermore, the benefit does not extend beyond age 80 or where other conditions reduce life expectancy. And, as noted earlier, modeling studies estimate that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

CHALLENGES TO IMPLEMENTATION

Concerns about false-positives, ­radiation exposure may limit acceptance

Challenges would be based mostly on the need for greater, more detailed dialogue between physicians and patients at higher risk for lung cancer in a time-constrained environment. Also, LDCT may not be available in some areas, and patients and physicians may have concerns regarding repeated CT exposure. In addition, false-positive results increase patient stress and may adversely affect both patient and physician acceptance.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 50-year-old woman presents to your office for a well-woman exam. Her past medical history includes a 22-pack-year smoking history (she quit 5 years ago), well-controlled hypertension, and mild obesity. She has no family history of cancer, but she does have a family history of type 2 diabetes and heart disease. Besides age- and risk-appropriate laboratory tests, cervical cancer screening, breast cancer screening, and initial colon cancer screening, are there any other preventive services you would offer her?

Lung cancer is the second most common cancer in both men and women, and it is the leading cause of cancer death in the United States—regardless of gender. The American Cancer Society estimates that 235,760 people will be diagnosed with lung cancer and 131,880 people will die of the disease in 2021.²

In the 2015 National Cancer Institute report on the economic costs of cancer, direct and indirect costs of lung cancer totaled $21.1 billion annually. Lost productivity from lung cancer added another $36.1 billion in annual costs.³ The economic costs increased to $23.8 billion in 2020, with no data on lost productivity.⁴

Smoking tobacco is by far the primary risk factor for lung cancer, and it is estimated to account for 90% of all lung cancer cases. Compared with nonsmokers, the relative risk of lung cancer is approximately 20 times higher for smokers.^5,6

Because the median age of lung cancer diagnosis is 70 years, increasing age is also considered a risk factor for lung cancer.^2,7

Although lung cancer has a relatively poor prognosis—with an average 5-year survival rate of 20.5%—early-stage lung cancer is more amenable to treatment and has a better prognosis (as is true with many cancers).¹

LDCT has a high sensitivity, as well as a reasonable specificity, for lung cancer detection. There is demonstrated benefit in screening patients who are at high risk for lung cancer.^8-11 In 2013, the USPSTF recommended annual lung cancer screening (B recommendation) with LDCT in adults 55 to 80 years of age who have a 30-pack-year smoking history, and who currently smoke or quit within the past 15 years.¹

Continue to: STUDY SUMMARY

Broader eligibility for screening supports mortality benefit

This is an update to the 2013 clinical practice guideline on lung cancer screening. The USPSTF used 2 methods to provide the best possible evidence for the recommendations. The first method was a systematic review of the accuracy of screening for lung cancer with LDCT, evaluating both the benefits and harms of lung cancer screening. The systematic review examined various subgroups, the number and/or frequency of LDCT scans, and various approaches to reducing false-positive results. In addition to the systematic review, they used collaborative modeling studies to determine the optimal age for beginning and ending screening, the optimal screening interval, and the relative benefits and harms of various screening strategies. These modeling studies complemented the evidence review.

This updated guideline nearly doubles eligibility for lung cancer screening using low-dose CT scanning.

The review included 7 randomized controlled trials (RCTs), plus the modeling studies. Only the National Lung Screening Trial (NLST; N = 53,454) and the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trial (N = 15,792) had adequate power to detect a mortality benefit from screening (NLST: relative risk reduction = 16%; 95% CI, 5%-25%; NELSON: incidence rate ratio = 0.75; 95% CI, 0.61-0.90) compared with no screening.

Screening intervals, from the NLST and NELSON trials as well as the modeling studies, revealed the greatest benefit from annual screening (statistics not shared). Evidence also showed that screening those with lighter smoking histories (< 30 pack-years) and at an earlier age (age 50) provided increased mortality benefit. No evidence was found for a benefit of screening past 80 years of age. The modeling studies concluded that the 2013 USPSTF screening program, using a starting age of 55 and a 30-pack-year smoking history, would reduce mortality by 9.8%, but by changing to a starting age of 50, a 20-pack-year smoking history, and annual screening, the mortality benefit was increased to 13%.^1,11

Comparison with computer-based risk prediction models from the Cancer Intervention and Surveillance Modeling Network (CISNET) revealed insufficient evidence at this time to show that prediction model–based screening offered any benefit beyond that of the age and smoking history risk factor model.

The incidence of false-positive results was > 25% in the NLST at baseline and at 1 year. Use of a classification system such as the Lung Imaging Reporting and Data System (Lung-RADS) could reduce that from 26.6% to 12.8%.² Another potential harm from LDCT screening is radiation exposure. Evidence from several RCTs and cohort studies showed the exposure from 1 LDCT scan to be 0.65 to 2.36 mSv, whereas the annual background radiation in the United States is 2.4 mSv. The modeling studies estimated that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

Continue to: WHAT'S NEW

Expanded age range, reduced pack-year history

Annual lung cancer screening is now recommended to begin for patients at age 50 years with a 20-pack-year history instead of age 55 years with a 30-pack-year history. This would nearly double (87% overall) the number of people eligible for screening, and it would include more Black patients and women, who tend to smoke fewer cigarettes than their White male counterparts. The American College of Radiology estimates that the expanded screening criteria could save between 30,000 and 60,000 lives per year.¹²

CAVEATS

Screening criteria for upper age limit, years since smoking remain unchanged

For those patients who quit smoking, the guidelines apply only to those who have stopped smoking within the past 15 years. Furthermore, the benefit does not extend beyond age 80 or where other conditions reduce life expectancy. And, as noted earlier, modeling studies estimate that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

CHALLENGES TO IMPLEMENTATION

Concerns about false-positives, ­radiation exposure may limit acceptance

Challenges would be based mostly on the need for greater, more detailed dialogue between physicians and patients at higher risk for lung cancer in a time-constrained environment. Also, LDCT may not be available in some areas, and patients and physicians may have concerns regarding repeated CT exposure. In addition, false-positive results increase patient stress and may adversely affect both patient and physician acceptance.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 50-year-old woman presents to your office for a well-woman exam. Her past medical history includes a 22-pack-year smoking history (she quit 5 years ago), well-controlled hypertension, and mild obesity. She has no family history of cancer, but she does have a family history of type 2 diabetes and heart disease. Besides age- and risk-appropriate laboratory tests, cervical cancer screening, breast cancer screening, and initial colon cancer screening, are there any other preventive services you would offer her?

Lung cancer is the second most common cancer in both men and women, and it is the leading cause of cancer death in the United States—regardless of gender. The American Cancer Society estimates that 235,760 people will be diagnosed with lung cancer and 131,880 people will die of the disease in 2021.²

In the 2015 National Cancer Institute report on the economic costs of cancer, direct and indirect costs of lung cancer totaled $21.1 billion annually. Lost productivity from lung cancer added another $36.1 billion in annual costs.³ The economic costs increased to $23.8 billion in 2020, with no data on lost productivity.⁴

Smoking tobacco is by far the primary risk factor for lung cancer, and it is estimated to account for 90% of all lung cancer cases. Compared with nonsmokers, the relative risk of lung cancer is approximately 20 times higher for smokers.^5,6

Because the median age of lung cancer diagnosis is 70 years, increasing age is also considered a risk factor for lung cancer.^2,7

Although lung cancer has a relatively poor prognosis—with an average 5-year survival rate of 20.5%—early-stage lung cancer is more amenable to treatment and has a better prognosis (as is true with many cancers).¹

LDCT has a high sensitivity, as well as a reasonable specificity, for lung cancer detection. There is demonstrated benefit in screening patients who are at high risk for lung cancer.^8-11 In 2013, the USPSTF recommended annual lung cancer screening (B recommendation) with LDCT in adults 55 to 80 years of age who have a 30-pack-year smoking history, and who currently smoke or quit within the past 15 years.¹

Continue to: STUDY SUMMARY

Broader eligibility for screening supports mortality benefit

This is an update to the 2013 clinical practice guideline on lung cancer screening. The USPSTF used 2 methods to provide the best possible evidence for the recommendations. The first method was a systematic review of the accuracy of screening for lung cancer with LDCT, evaluating both the benefits and harms of lung cancer screening. The systematic review examined various subgroups, the number and/or frequency of LDCT scans, and various approaches to reducing false-positive results. In addition to the systematic review, they used collaborative modeling studies to determine the optimal age for beginning and ending screening, the optimal screening interval, and the relative benefits and harms of various screening strategies. These modeling studies complemented the evidence review.

This updated guideline nearly doubles eligibility for lung cancer screening using low-dose CT scanning.

The review included 7 randomized controlled trials (RCTs), plus the modeling studies. Only the National Lung Screening Trial (NLST; N = 53,454) and the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trial (N = 15,792) had adequate power to detect a mortality benefit from screening (NLST: relative risk reduction = 16%; 95% CI, 5%-25%; NELSON: incidence rate ratio = 0.75; 95% CI, 0.61-0.90) compared with no screening.

Screening intervals, from the NLST and NELSON trials as well as the modeling studies, revealed the greatest benefit from annual screening (statistics not shared). Evidence also showed that screening those with lighter smoking histories (< 30 pack-years) and at an earlier age (age 50) provided increased mortality benefit. No evidence was found for a benefit of screening past 80 years of age. The modeling studies concluded that the 2013 USPSTF screening program, using a starting age of 55 and a 30-pack-year smoking history, would reduce mortality by 9.8%, but by changing to a starting age of 50, a 20-pack-year smoking history, and annual screening, the mortality benefit was increased to 13%.^1,11

Comparison with computer-based risk prediction models from the Cancer Intervention and Surveillance Modeling Network (CISNET) revealed insufficient evidence at this time to show that prediction model–based screening offered any benefit beyond that of the age and smoking history risk factor model.

The incidence of false-positive results was > 25% in the NLST at baseline and at 1 year. Use of a classification system such as the Lung Imaging Reporting and Data System (Lung-RADS) could reduce that from 26.6% to 12.8%.² Another potential harm from LDCT screening is radiation exposure. Evidence from several RCTs and cohort studies showed the exposure from 1 LDCT scan to be 0.65 to 2.36 mSv, whereas the annual background radiation in the United States is 2.4 mSv. The modeling studies estimated that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

Continue to: WHAT'S NEW

Expanded age range, reduced pack-year history

Annual lung cancer screening is now recommended to begin for patients at age 50 years with a 20-pack-year history instead of age 55 years with a 30-pack-year history. This would nearly double (87% overall) the number of people eligible for screening, and it would include more Black patients and women, who tend to smoke fewer cigarettes than their White male counterparts. The American College of Radiology estimates that the expanded screening criteria could save between 30,000 and 60,000 lives per year.¹²

CAVEATS

Screening criteria for upper age limit, years since smoking remain unchanged

For those patients who quit smoking, the guidelines apply only to those who have stopped smoking within the past 15 years. Furthermore, the benefit does not extend beyond age 80 or where other conditions reduce life expectancy. And, as noted earlier, modeling studies estimate that there would be 1 death caused by LDCT for every 18.5 cancer deaths avoided.^1,11

CHALLENGES TO IMPLEMENTATION

Concerns about false-positives, ­radiation exposure may limit acceptance

Challenges would be based mostly on the need for greater, more detailed dialogue between physicians and patients at higher risk for lung cancer in a time-constrained environment. Also, LDCT may not be available in some areas, and patients and physicians may have concerns regarding repeated CT exposure. In addition, false-positive results increase patient stress and may adversely affect both patient and physician acceptance.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 45-year-old woman with no chronic medical illness presents to your office for her annual physical examination. After a medical assistant (MA) applies an automatic BP cuff to the patient’s left arm, the BP reading is 155/92 mm Hg. The MA then rechecks the BP, and this time it reads 160/98 mm Hg. The MA performs a manual BP reading, which is 158/90 mm Hg (left arm) and 162/100 mm Hg (right arm). The patient denies any headache, visual changes, chest pain, or difficulty breathing and tells the MA that her BP is always high during a doctor visit. You are wondering if she has hypertension or if is this the white-coat effect.

Depending on the definition of hypertension, its prevalence among US adults 18 years or older varies from 46%, based on the American College of Cardiology guideline (≥ 130/80 mm Hg), to 29%, based on the Eighth Joint National Committee (JNC-8) guideline (≥ 140/90 mm Hg for adults ages 18–59 years and ≥ 150/90 mm Hg for adults ≥ 60 years without diabetes and/or chronic kidney disease).^2,3

According to JNC-8, the prevalence is similar among men (30.2%) and women (27.7%) and increases with age: 18 to 39 years, 7.5%; 40 to 59 years, 33.2%; and ≥ 60 years, 63.1%.^3,4 When ranked by risk-attributable ­disability-adjusted life-years (DALYs), high systolic blood pressure (SBP) is the leading risk factor, accounting for 10.4 million deaths and 218 million DALYs globally in 2017.⁵ National medical costs associated with hypertension are estimated to account for about $131 billion in annual health care expenditures, averaged over 12 years from 2003 to 2014.⁶

When performed correctly, the auscultatory method using a mercury sphygmomanometer correlates well with simultaneous intra-arterial BP and was considered the gold standard for office-based measurements for many years.^7,8 However, significant ­observer-related differences in auditory acuity and terminal digit rounding are sources of inaccurate measurement. White-coat hypertension cannot be detected with this method—another significant limitation. The inaccuracy of office-based BP readings leads to concerns about hypertension being inappropriately diagnosed in patients or delays in diagnosis occurring.⁹

A proposed solution to this problem is measurement using an oscillometric sphygmomanometer. This device uses a pressure transducer to assess the oscillations of pressure in a cuff during gradual deflation; it provides accurate BP measurements when fully automated and programmed to complete several BP measurements at appropriate intervals while the patient rests alone in a quiet room.¹⁰

The accuracy of this new method was tested in a 2009 cohort study of 309 patients referred to an ambulatory blood pressure (ABP) monitoring unit at an academic hospital for diagnosis or management of hypertension.¹¹ The study compared mean awake ABP, which continuously measures patients’ BP throughout the day, manual sphygmomanometer readings taken by the patient’s own physician, and an automated office blood pressure (AOBP) device called BpTRU (an automated oscillometric sphygmomanometer) while the patient rested alone in the exam room.¹¹ The awake ABP is a federally approved standard for the diagnosis of white-coat hypertension.¹² In this study, the white-coat response was negated with the use of the automated BpTRU device.¹¹

A 2019 meta-analysis that included 26 studies (N = 7116) comparing AOBP with other BP measurement techniques concluded that the use of automated oscillometric BP readings is more accurate for diagnosing hypertension and assists in negating the white-coat hypertension effect.⁹

Continue to: STUDY SUMMARY

STUDY SUMMARY

Automated office BP devices are just as accurate as more expensive ABP studies

This systematic review and meta-analysis (N = 9279; 23 cross-sectional, 1 cohort, and 7 randomized controlled trials [N = 1304], of which 17 studies overlapped with those included in the previously mentioned meta-analysis⁹) compared SBP and diastolic blood pressure measured by an oscillometric AOBP device to awake or daytime ABP (continuously monitoring BP while awake, used as a standard for BP measurement), routine manual office BP, or research BP measurements.

The study also explored the protocol by which the best AOBP results could be obtained. For AOBP measurement, the included trials had no more than 2 minutes of elapsed time between individual AOBP measurements and had at least 3 AOBP readings to calculate the mean.

Compared with AOBP, in samples with an SBP of ≥ 130 mm Hg, SBP readings were significantly higher for both routine office visits (mean difference [MD] = 14.5 mm Hg; 95% CI, 11.8–17.2) and research (MD = 7 mm Hg; 95% CI, 4.9–9.1). However, no difference was found between AOBP and awake ambulatory SBP values (MD = 0.3 mm Hg; 95% CI, −1.1 to 1.7). In all cases, heterogeneity of the included studies was high (I² was > 75%). There was no evidence of small-study effect or publication bias, and little evidence of potential financial bias. The most accurate methodology for AOBP measurements included multiple BP readings and the patient resting alone in a quiet location.

This meta-analysis supports the use of an automated office blood pressure device to accurately screen for hypertension and avoid the white-coat effect.

Although there was statistical heterogeneity, the results were confirmed in the authors’ analysis of studies with high methodologic quality. In addition, researchers performed multiple meta-regression analyses to evaluate the statistical heterogeneity and found no significant differences based on age, body mass index, number of treated patients, gender, measurement interval, or added rest before AOBP.

WHAT'S NEW

Study confirms unattended, automated office BP as preferred technique

This is the second recent comprehensive systematic review and meta-analysis to directly compare AOBP with other common techniques of BP measurement in screening for and diagnosing hypertension in the clinical setting. ⁹

Continue to: This meta-analysis...

This meta-analysis emphasized the technique (see below) by which to obtain the best AOBP vs ABP results, whereas the other ­meta-analysis⁹ did not. Thus the study provides practice-based settings with the information they need to more closely replicate the results of the studies included in the meta-analysis.

Also, the equivalency comparison with the more expensive and intrusive ABP monitoring may save money, improve patient adherence, and increase patient satisfaction. Given these advantages, along with its demonstrated accuracy, AOBP should be adopted in routine clinical practice to screen patients for hypertension.

CAVEATS

Close adherence to measurementprocedures is a necessity

Effective use of AOBP in clinical practice requires close adherence to the AOBP study procedures described in this meta-analysis. These include taking multiple (at least 3) BP readings, 1 to 2 minutes apart, recorded with a fully automated oscillometric sphygmomanometer while the patient rests alone in a quiet place.

CHALLENGES TO IMPLEMENTATION

Adjusting workflows, addressing cost

Physicians may be reluctant to adopt this technique because they may not be convinced of its advantages compared with the traditional methods of recording BP and because of difficulties with implementing new rooming workflows.¹² The cost of AOBP devices used in this study (Omron 907 and BpTRU; BpTRU ceased operations in 2017) were not disclosed, which may be a hindrance, as devices may cost $1000 or more.

Effective use of automated office blood pressure requires that one take multiple (at least 3) BP readings, 1 to 2 minutes apart, while the patient rests alone in a quiet place.

An online search for “automated oscillometric BP monitor” by one of the PURL authors (RCM) found oscillometric AOBP devices ranging from $150 to > $1000, depending on whether the device was medical grade; a search for “Omron 907” found devices for ≤ $599 on multiple sites. However, none of the lower-cost devices indicated the ability to take multiple, unattended BP readings.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 45-year-old woman with no chronic medical illness presents to your office for her annual physical examination. After a medical assistant (MA) applies an automatic BP cuff to the patient’s left arm, the BP reading is 155/92 mm Hg. The MA then rechecks the BP, and this time it reads 160/98 mm Hg. The MA performs a manual BP reading, which is 158/90 mm Hg (left arm) and 162/100 mm Hg (right arm). The patient denies any headache, visual changes, chest pain, or difficulty breathing and tells the MA that her BP is always high during a doctor visit. You are wondering if she has hypertension or if is this the white-coat effect.

Depending on the definition of hypertension, its prevalence among US adults 18 years or older varies from 46%, based on the American College of Cardiology guideline (≥ 130/80 mm Hg), to 29%, based on the Eighth Joint National Committee (JNC-8) guideline (≥ 140/90 mm Hg for adults ages 18–59 years and ≥ 150/90 mm Hg for adults ≥ 60 years without diabetes and/or chronic kidney disease).^2,3

According to JNC-8, the prevalence is similar among men (30.2%) and women (27.7%) and increases with age: 18 to 39 years, 7.5%; 40 to 59 years, 33.2%; and ≥ 60 years, 63.1%.^3,4 When ranked by risk-attributable ­disability-adjusted life-years (DALYs), high systolic blood pressure (SBP) is the leading risk factor, accounting for 10.4 million deaths and 218 million DALYs globally in 2017.⁵ National medical costs associated with hypertension are estimated to account for about $131 billion in annual health care expenditures, averaged over 12 years from 2003 to 2014.⁶

When performed correctly, the auscultatory method using a mercury sphygmomanometer correlates well with simultaneous intra-arterial BP and was considered the gold standard for office-based measurements for many years.^7,8 However, significant ­observer-related differences in auditory acuity and terminal digit rounding are sources of inaccurate measurement. White-coat hypertension cannot be detected with this method—another significant limitation. The inaccuracy of office-based BP readings leads to concerns about hypertension being inappropriately diagnosed in patients or delays in diagnosis occurring.⁹

A proposed solution to this problem is measurement using an oscillometric sphygmomanometer. This device uses a pressure transducer to assess the oscillations of pressure in a cuff during gradual deflation; it provides accurate BP measurements when fully automated and programmed to complete several BP measurements at appropriate intervals while the patient rests alone in a quiet room.¹⁰

The accuracy of this new method was tested in a 2009 cohort study of 309 patients referred to an ambulatory blood pressure (ABP) monitoring unit at an academic hospital for diagnosis or management of hypertension.¹¹ The study compared mean awake ABP, which continuously measures patients’ BP throughout the day, manual sphygmomanometer readings taken by the patient’s own physician, and an automated office blood pressure (AOBP) device called BpTRU (an automated oscillometric sphygmomanometer) while the patient rested alone in the exam room.¹¹ The awake ABP is a federally approved standard for the diagnosis of white-coat hypertension.¹² In this study, the white-coat response was negated with the use of the automated BpTRU device.¹¹

A 2019 meta-analysis that included 26 studies (N = 7116) comparing AOBP with other BP measurement techniques concluded that the use of automated oscillometric BP readings is more accurate for diagnosing hypertension and assists in negating the white-coat hypertension effect.⁹

Continue to: STUDY SUMMARY

STUDY SUMMARY

Automated office BP devices are just as accurate as more expensive ABP studies

This systematic review and meta-analysis (N = 9279; 23 cross-sectional, 1 cohort, and 7 randomized controlled trials [N = 1304], of which 17 studies overlapped with those included in the previously mentioned meta-analysis⁹) compared SBP and diastolic blood pressure measured by an oscillometric AOBP device to awake or daytime ABP (continuously monitoring BP while awake, used as a standard for BP measurement), routine manual office BP, or research BP measurements.

The study also explored the protocol by which the best AOBP results could be obtained. For AOBP measurement, the included trials had no more than 2 minutes of elapsed time between individual AOBP measurements and had at least 3 AOBP readings to calculate the mean.

Compared with AOBP, in samples with an SBP of ≥ 130 mm Hg, SBP readings were significantly higher for both routine office visits (mean difference [MD] = 14.5 mm Hg; 95% CI, 11.8–17.2) and research (MD = 7 mm Hg; 95% CI, 4.9–9.1). However, no difference was found between AOBP and awake ambulatory SBP values (MD = 0.3 mm Hg; 95% CI, −1.1 to 1.7). In all cases, heterogeneity of the included studies was high (I² was > 75%). There was no evidence of small-study effect or publication bias, and little evidence of potential financial bias. The most accurate methodology for AOBP measurements included multiple BP readings and the patient resting alone in a quiet location.

This meta-analysis supports the use of an automated office blood pressure device to accurately screen for hypertension and avoid the white-coat effect.

Although there was statistical heterogeneity, the results were confirmed in the authors’ analysis of studies with high methodologic quality. In addition, researchers performed multiple meta-regression analyses to evaluate the statistical heterogeneity and found no significant differences based on age, body mass index, number of treated patients, gender, measurement interval, or added rest before AOBP.

WHAT'S NEW

Study confirms unattended, automated office BP as preferred technique

This is the second recent comprehensive systematic review and meta-analysis to directly compare AOBP with other common techniques of BP measurement in screening for and diagnosing hypertension in the clinical setting. ⁹

Continue to: This meta-analysis...

This meta-analysis emphasized the technique (see below) by which to obtain the best AOBP vs ABP results, whereas the other ­meta-analysis⁹ did not. Thus the study provides practice-based settings with the information they need to more closely replicate the results of the studies included in the meta-analysis.

Also, the equivalency comparison with the more expensive and intrusive ABP monitoring may save money, improve patient adherence, and increase patient satisfaction. Given these advantages, along with its demonstrated accuracy, AOBP should be adopted in routine clinical practice to screen patients for hypertension.

CAVEATS

Close adherence to measurementprocedures is a necessity

Effective use of AOBP in clinical practice requires close adherence to the AOBP study procedures described in this meta-analysis. These include taking multiple (at least 3) BP readings, 1 to 2 minutes apart, recorded with a fully automated oscillometric sphygmomanometer while the patient rests alone in a quiet place.

CHALLENGES TO IMPLEMENTATION

Adjusting workflows, addressing cost

Physicians may be reluctant to adopt this technique because they may not be convinced of its advantages compared with the traditional methods of recording BP and because of difficulties with implementing new rooming workflows.¹² The cost of AOBP devices used in this study (Omron 907 and BpTRU; BpTRU ceased operations in 2017) were not disclosed, which may be a hindrance, as devices may cost $1000 or more.

Effective use of automated office blood pressure requires that one take multiple (at least 3) BP readings, 1 to 2 minutes apart, while the patient rests alone in a quiet place.

An online search for “automated oscillometric BP monitor” by one of the PURL authors (RCM) found oscillometric AOBP devices ranging from $150 to > $1000, depending on whether the device was medical grade; a search for “Omron 907” found devices for ≤ $599 on multiple sites. However, none of the lower-cost devices indicated the ability to take multiple, unattended BP readings.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ILLUSTRATIVE CASE

A 45-year-old woman with no chronic medical illness presents to your office for her annual physical examination. After a medical assistant (MA) applies an automatic BP cuff to the patient’s left arm, the BP reading is 155/92 mm Hg. The MA then rechecks the BP, and this time it reads 160/98 mm Hg. The MA performs a manual BP reading, which is 158/90 mm Hg (left arm) and 162/100 mm Hg (right arm). The patient denies any headache, visual changes, chest pain, or difficulty breathing and tells the MA that her BP is always high during a doctor visit. You are wondering if she has hypertension or if is this the white-coat effect.

Depending on the definition of hypertension, its prevalence among US adults 18 years or older varies from 46%, based on the American College of Cardiology guideline (≥ 130/80 mm Hg), to 29%, based on the Eighth Joint National Committee (JNC-8) guideline (≥ 140/90 mm Hg for adults ages 18–59 years and ≥ 150/90 mm Hg for adults ≥ 60 years without diabetes and/or chronic kidney disease).^2,3

According to JNC-8, the prevalence is similar among men (30.2%) and women (27.7%) and increases with age: 18 to 39 years, 7.5%; 40 to 59 years, 33.2%; and ≥ 60 years, 63.1%.^3,4 When ranked by risk-attributable ­disability-adjusted life-years (DALYs), high systolic blood pressure (SBP) is the leading risk factor, accounting for 10.4 million deaths and 218 million DALYs globally in 2017.⁵ National medical costs associated with hypertension are estimated to account for about $131 billion in annual health care expenditures, averaged over 12 years from 2003 to 2014.⁶

When performed correctly, the auscultatory method using a mercury sphygmomanometer correlates well with simultaneous intra-arterial BP and was considered the gold standard for office-based measurements for many years.^7,8 However, significant ­observer-related differences in auditory acuity and terminal digit rounding are sources of inaccurate measurement. White-coat hypertension cannot be detected with this method—another significant limitation. The inaccuracy of office-based BP readings leads to concerns about hypertension being inappropriately diagnosed in patients or delays in diagnosis occurring.⁹

A proposed solution to this problem is measurement using an oscillometric sphygmomanometer. This device uses a pressure transducer to assess the oscillations of pressure in a cuff during gradual deflation; it provides accurate BP measurements when fully automated and programmed to complete several BP measurements at appropriate intervals while the patient rests alone in a quiet room.¹⁰

The accuracy of this new method was tested in a 2009 cohort study of 309 patients referred to an ambulatory blood pressure (ABP) monitoring unit at an academic hospital for diagnosis or management of hypertension.¹¹ The study compared mean awake ABP, which continuously measures patients’ BP throughout the day, manual sphygmomanometer readings taken by the patient’s own physician, and an automated office blood pressure (AOBP) device called BpTRU (an automated oscillometric sphygmomanometer) while the patient rested alone in the exam room.¹¹ The awake ABP is a federally approved standard for the diagnosis of white-coat hypertension.¹² In this study, the white-coat response was negated with the use of the automated BpTRU device.¹¹

A 2019 meta-analysis that included 26 studies (N = 7116) comparing AOBP with other BP measurement techniques concluded that the use of automated oscillometric BP readings is more accurate for diagnosing hypertension and assists in negating the white-coat hypertension effect.⁹

Continue to: STUDY SUMMARY

STUDY SUMMARY

Automated office BP devices are just as accurate as more expensive ABP studies

This systematic review and meta-analysis (N = 9279; 23 cross-sectional, 1 cohort, and 7 randomized controlled trials [N = 1304], of which 17 studies overlapped with those included in the previously mentioned meta-analysis⁹) compared SBP and diastolic blood pressure measured by an oscillometric AOBP device to awake or daytime ABP (continuously monitoring BP while awake, used as a standard for BP measurement), routine manual office BP, or research BP measurements.

The study also explored the protocol by which the best AOBP results could be obtained. For AOBP measurement, the included trials had no more than 2 minutes of elapsed time between individual AOBP measurements and had at least 3 AOBP readings to calculate the mean.

Compared with AOBP, in samples with an SBP of ≥ 130 mm Hg, SBP readings were significantly higher for both routine office visits (mean difference [MD] = 14.5 mm Hg; 95% CI, 11.8–17.2) and research (MD = 7 mm Hg; 95% CI, 4.9–9.1). However, no difference was found between AOBP and awake ambulatory SBP values (MD = 0.3 mm Hg; 95% CI, −1.1 to 1.7). In all cases, heterogeneity of the included studies was high (I² was > 75%). There was no evidence of small-study effect or publication bias, and little evidence of potential financial bias. The most accurate methodology for AOBP measurements included multiple BP readings and the patient resting alone in a quiet location.

This meta-analysis supports the use of an automated office blood pressure device to accurately screen for hypertension and avoid the white-coat effect.

Although there was statistical heterogeneity, the results were confirmed in the authors’ analysis of studies with high methodologic quality. In addition, researchers performed multiple meta-regression analyses to evaluate the statistical heterogeneity and found no significant differences based on age, body mass index, number of treated patients, gender, measurement interval, or added rest before AOBP.

WHAT'S NEW

Study confirms unattended, automated office BP as preferred technique

This is the second recent comprehensive systematic review and meta-analysis to directly compare AOBP with other common techniques of BP measurement in screening for and diagnosing hypertension in the clinical setting. ⁹

Continue to: This meta-analysis...

This meta-analysis emphasized the technique (see below) by which to obtain the best AOBP vs ABP results, whereas the other ­meta-analysis⁹ did not. Thus the study provides practice-based settings with the information they need to more closely replicate the results of the studies included in the meta-analysis.

Also, the equivalency comparison with the more expensive and intrusive ABP monitoring may save money, improve patient adherence, and increase patient satisfaction. Given these advantages, along with its demonstrated accuracy, AOBP should be adopted in routine clinical practice to screen patients for hypertension.

CAVEATS

Close adherence to measurementprocedures is a necessity

Effective use of AOBP in clinical practice requires close adherence to the AOBP study procedures described in this meta-analysis. These include taking multiple (at least 3) BP readings, 1 to 2 minutes apart, recorded with a fully automated oscillometric sphygmomanometer while the patient rests alone in a quiet place.

CHALLENGES TO IMPLEMENTATION

Adjusting workflows, addressing cost

Physicians may be reluctant to adopt this technique because they may not be convinced of its advantages compared with the traditional methods of recording BP and because of difficulties with implementing new rooming workflows.¹² The cost of AOBP devices used in this study (Omron 907 and BpTRU; BpTRU ceased operations in 2017) were not disclosed, which may be a hindrance, as devices may cost $1000 or more.

Effective use of automated office blood pressure requires that one take multiple (at least 3) BP readings, 1 to 2 minutes apart, while the patient rests alone in a quiet place.

An online search for “automated oscillometric BP monitor” by one of the PURL authors (RCM) found oscillometric AOBP devices ranging from $150 to > $1000, depending on whether the device was medical grade; a search for “Omron 907” found devices for ≤ $599 on multiple sites. However, none of the lower-cost devices indicated the ability to take multiple, unattended BP readings.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center for Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.