Published reports of improved sexual performance have prompted men without erectile dysfunction to use sildenafil inappropriately.^1-4 Sildenafil has also been used to counteract the impotence-inducing effects of “club drugs” such as ecstasy.⁵

Cannabis, another widely abused street drug, is a known inhibitor of the cytochrome P450 3A4 isoenzyme pathway, the same pathway in which sildenafil is metabolized.⁶ Cannabis can thus potentiate the effect of sildenafil. A case report from 2002 has indicated that a young man using sildenafil and cannabis concomitantly suffered a myocardial infarction.⁷

In this study, we sought to answer the following questions:

• What methods did the men use to obtain sildenafil without a prescription?
• Why were the men taking sildenafil?
• Did these men increase the dose without physician supervision?

Methods

Setting, participants, and design

Two senior resident physicians from the Department of Family Practice at The Brooklyn Hospital Center asked male patients between the ages of 18 and 80 if they would be interested in participating in this descriptive study. Two hundred and thirty-one patients agreed to be interviewed during their outpatient clinic appointments in the Departments of Family Practice and Urology. We conducted these interviews between December 1, 2002 and April 30, 2003.

Patients with and without physician-diagnosed erectile dysfunction who were using sildenafil were included in this study. Patients were excluded if they were taking nitrates, had cognitive disabilities, were female, or if they could not read English.

The 1-page, 35-item questionnaire was read to the subjects by a resident, who provided additional explanations if needed. The researchers noted their responses to questions on demographics, medical history, social history, treatment duration of erectile dysfunction, method of procurement of sildenafil, and knowledge about the indications of sildenafil. The questionnaire was pretested on a small sample for comprehension prior to distribution.

We developed descriptive statistics and performed cross-tabulations using SPSS version 11.0 (SPSS, Inc, Chicago, Ill). We used a chi-square test to determine statistical significance between cannabis abuse and illicit sildenafil use. We established statistical significance at P<.05. The Institutional Review Board at the Brooklyn Hospital Center approved our research protocol, and we obtained consent from all the study participants.

Results

Strong link between cannabis use and recreational sildenafil

TABLE 1 shows the demographic information of the entire study population (n=231). Our study found that patient predictors for recreational sildenafil use are a younger unmarried male who smokes cannabis. Of the men in the study, 138 (59.7%) reported erectile function prior to the use of sildenafil.

We decided to examine data from this subgroup of our study population. As shown in TABLE 2, patients with erectile function but with a history of cannabis abuse reported obtaining sildenafil from friends, street vendors, and the Internet significantly more often than those with erectile dysfunction who did not use cannabis.

Discussion: Is there a danger?

Our study showed a strong association between individuals who obtained sildenafil from sources other than a prescribing physician and those who used cannabis. These men purchased this prescription medication from street vendors, friends, family, or via the Internet. Illicit sildenafil users took the medication mainly to improve performance and increase desire–and they often increased the dose of the medication at will. The differences between patients with erectile dysfunction compared with those without were so great that much of our data proved to be statistically significant (TABLE 1).

Data from our subgroup of patients who smoked cannabis supported our suspicion that patients who were able to maintain an erection prior to their use of sildenafil used the medication to improve sexual performance and counteract alterations in libido caused by cannabis.⁸

TABLE 1
Our study group: Who they were, why they were taking sildenafil

TABLE 2
Where did 2 subsets of subjects obtain sildenafil?

Limitations of this study

The main limitation of this study was that the data obtained were self-reported. A chart review could have provided objective data on the patients’ ED diagnosis and medications.

Conclusion

The illicit use of sildenafil raises many issues. Patients with cardiovascular disease, even without the use of nitrates, may be at risk of myocardial infarction. Be aware that younger, male patient with an admitted history of drug abuse may be taking sildenafil without your knowledge, even without a diagnosis of erectile dysfunction.

Funding

Material support was provided by the Department of Family Medicine at The Brooklyn Hospital, Brooklyn, NY.

Acknowledgments

The contents of this manuscript were presented at the New York State Academy of Family Practice and The Albany County Chapter Regional Family Medicine Conference at Lake Placid, NY on September 6, 2003.

Correspondence
Marie L. Eloi-Stiven, MD, Director of Research, The Brooklyn Hospital Center, Department of Family Medicine, 121 Dekalb Ave, Brooklyn, NY 11201; dad9022@nyp.org

Published reports of improved sexual performance have prompted men without erectile dysfunction to use sildenafil inappropriately.^1-4 Sildenafil has also been used to counteract the impotence-inducing effects of “club drugs” such as ecstasy.⁵

Cannabis, another widely abused street drug, is a known inhibitor of the cytochrome P450 3A4 isoenzyme pathway, the same pathway in which sildenafil is metabolized.⁶ Cannabis can thus potentiate the effect of sildenafil. A case report from 2002 has indicated that a young man using sildenafil and cannabis concomitantly suffered a myocardial infarction.⁷

In this study, we sought to answer the following questions:

• What methods did the men use to obtain sildenafil without a prescription?
• Why were the men taking sildenafil?
• Did these men increase the dose without physician supervision?

Methods

Setting, participants, and design

Two senior resident physicians from the Department of Family Practice at The Brooklyn Hospital Center asked male patients between the ages of 18 and 80 if they would be interested in participating in this descriptive study. Two hundred and thirty-one patients agreed to be interviewed during their outpatient clinic appointments in the Departments of Family Practice and Urology. We conducted these interviews between December 1, 2002 and April 30, 2003.

Patients with and without physician-diagnosed erectile dysfunction who were using sildenafil were included in this study. Patients were excluded if they were taking nitrates, had cognitive disabilities, were female, or if they could not read English.

The 1-page, 35-item questionnaire was read to the subjects by a resident, who provided additional explanations if needed. The researchers noted their responses to questions on demographics, medical history, social history, treatment duration of erectile dysfunction, method of procurement of sildenafil, and knowledge about the indications of sildenafil. The questionnaire was pretested on a small sample for comprehension prior to distribution.

We developed descriptive statistics and performed cross-tabulations using SPSS version 11.0 (SPSS, Inc, Chicago, Ill). We used a chi-square test to determine statistical significance between cannabis abuse and illicit sildenafil use. We established statistical significance at P<.05. The Institutional Review Board at the Brooklyn Hospital Center approved our research protocol, and we obtained consent from all the study participants.

Results

Strong link between cannabis use and recreational sildenafil

TABLE 1 shows the demographic information of the entire study population (n=231). Our study found that patient predictors for recreational sildenafil use are a younger unmarried male who smokes cannabis. Of the men in the study, 138 (59.7%) reported erectile function prior to the use of sildenafil.

We decided to examine data from this subgroup of our study population. As shown in TABLE 2, patients with erectile function but with a history of cannabis abuse reported obtaining sildenafil from friends, street vendors, and the Internet significantly more often than those with erectile dysfunction who did not use cannabis.

Discussion: Is there a danger?

Our study showed a strong association between individuals who obtained sildenafil from sources other than a prescribing physician and those who used cannabis. These men purchased this prescription medication from street vendors, friends, family, or via the Internet. Illicit sildenafil users took the medication mainly to improve performance and increase desire–and they often increased the dose of the medication at will. The differences between patients with erectile dysfunction compared with those without were so great that much of our data proved to be statistically significant (TABLE 1).

Data from our subgroup of patients who smoked cannabis supported our suspicion that patients who were able to maintain an erection prior to their use of sildenafil used the medication to improve sexual performance and counteract alterations in libido caused by cannabis.⁸

TABLE 1
Our study group: Who they were, why they were taking sildenafil

TABLE 2
Where did 2 subsets of subjects obtain sildenafil?

Limitations of this study

The main limitation of this study was that the data obtained were self-reported. A chart review could have provided objective data on the patients’ ED diagnosis and medications.

Conclusion

The illicit use of sildenafil raises many issues. Patients with cardiovascular disease, even without the use of nitrates, may be at risk of myocardial infarction. Be aware that younger, male patient with an admitted history of drug abuse may be taking sildenafil without your knowledge, even without a diagnosis of erectile dysfunction.

Funding

Material support was provided by the Department of Family Medicine at The Brooklyn Hospital, Brooklyn, NY.

Acknowledgments

The contents of this manuscript were presented at the New York State Academy of Family Practice and The Albany County Chapter Regional Family Medicine Conference at Lake Placid, NY on September 6, 2003.

Correspondence
Marie L. Eloi-Stiven, MD, Director of Research, The Brooklyn Hospital Center, Department of Family Medicine, 121 Dekalb Ave, Brooklyn, NY 11201; dad9022@nyp.org

Published reports of improved sexual performance have prompted men without erectile dysfunction to use sildenafil inappropriately.^1-4 Sildenafil has also been used to counteract the impotence-inducing effects of “club drugs” such as ecstasy.⁵

Cannabis, another widely abused street drug, is a known inhibitor of the cytochrome P450 3A4 isoenzyme pathway, the same pathway in which sildenafil is metabolized.⁶ Cannabis can thus potentiate the effect of sildenafil. A case report from 2002 has indicated that a young man using sildenafil and cannabis concomitantly suffered a myocardial infarction.⁷

In this study, we sought to answer the following questions:

• What methods did the men use to obtain sildenafil without a prescription?
• Why were the men taking sildenafil?
• Did these men increase the dose without physician supervision?

Methods

Setting, participants, and design

Two senior resident physicians from the Department of Family Practice at The Brooklyn Hospital Center asked male patients between the ages of 18 and 80 if they would be interested in participating in this descriptive study. Two hundred and thirty-one patients agreed to be interviewed during their outpatient clinic appointments in the Departments of Family Practice and Urology. We conducted these interviews between December 1, 2002 and April 30, 2003.

Patients with and without physician-diagnosed erectile dysfunction who were using sildenafil were included in this study. Patients were excluded if they were taking nitrates, had cognitive disabilities, were female, or if they could not read English.

The 1-page, 35-item questionnaire was read to the subjects by a resident, who provided additional explanations if needed. The researchers noted their responses to questions on demographics, medical history, social history, treatment duration of erectile dysfunction, method of procurement of sildenafil, and knowledge about the indications of sildenafil. The questionnaire was pretested on a small sample for comprehension prior to distribution.

We developed descriptive statistics and performed cross-tabulations using SPSS version 11.0 (SPSS, Inc, Chicago, Ill). We used a chi-square test to determine statistical significance between cannabis abuse and illicit sildenafil use. We established statistical significance at P<.05. The Institutional Review Board at the Brooklyn Hospital Center approved our research protocol, and we obtained consent from all the study participants.

Results

Strong link between cannabis use and recreational sildenafil

TABLE 1 shows the demographic information of the entire study population (n=231). Our study found that patient predictors for recreational sildenafil use are a younger unmarried male who smokes cannabis. Of the men in the study, 138 (59.7%) reported erectile function prior to the use of sildenafil.

We decided to examine data from this subgroup of our study population. As shown in TABLE 2, patients with erectile function but with a history of cannabis abuse reported obtaining sildenafil from friends, street vendors, and the Internet significantly more often than those with erectile dysfunction who did not use cannabis.

Discussion: Is there a danger?

Our study showed a strong association between individuals who obtained sildenafil from sources other than a prescribing physician and those who used cannabis. These men purchased this prescription medication from street vendors, friends, family, or via the Internet. Illicit sildenafil users took the medication mainly to improve performance and increase desire–and they often increased the dose of the medication at will. The differences between patients with erectile dysfunction compared with those without were so great that much of our data proved to be statistically significant (TABLE 1).

Data from our subgroup of patients who smoked cannabis supported our suspicion that patients who were able to maintain an erection prior to their use of sildenafil used the medication to improve sexual performance and counteract alterations in libido caused by cannabis.⁸

TABLE 1
Our study group: Who they were, why they were taking sildenafil

TABLE 2
Where did 2 subsets of subjects obtain sildenafil?

Limitations of this study

The main limitation of this study was that the data obtained were self-reported. A chart review could have provided objective data on the patients’ ED diagnosis and medications.

Conclusion

The illicit use of sildenafil raises many issues. Patients with cardiovascular disease, even without the use of nitrates, may be at risk of myocardial infarction. Be aware that younger, male patient with an admitted history of drug abuse may be taking sildenafil without your knowledge, even without a diagnosis of erectile dysfunction.

Funding

Material support was provided by the Department of Family Medicine at The Brooklyn Hospital, Brooklyn, NY.

Acknowledgments

The contents of this manuscript were presented at the New York State Academy of Family Practice and The Albany County Chapter Regional Family Medicine Conference at Lake Placid, NY on September 6, 2003.

Correspondence
Marie L. Eloi-Stiven, MD, Director of Research, The Brooklyn Hospital Center, Department of Family Medicine, 121 Dekalb Ave, Brooklyn, NY 11201; dad9022@nyp.org

Illustrative Case

An 85-year-old woman with hypertension and chronic atrial fibrillation has transferred her care to you. She takes an aspirin a day for cardiovascular prevention. You know that warfarin is better than aspirin for preventing stroke but worry about the increased risk of bleeding with warfarin.

Should you recommend that she stay on aspirin or switch to warfarin?

Background: BAFTA: A realistic study

We have been reluctant to use warfarin in elders with atrial fibrillation for good reason: risk of hemorrhage. Since there are few trials looking at use of warfarin among elders in primary care settings, we are uncertain about the balance of benefits and harms.

The BAFTA study¹ is the first trial to compare outcomes of warfarin vs aspirin in elders specifically, in the less-than-ideal conditions of real life.

Guidelines mirror uncertainties

This uncertainty is reflected even in guidelines for anticoagulation in elderly patients with atrial fibrillation.

• The 2004 American College of Chest Physicians Seventh Conference on Antithrombotic and Thrombolytic Therapy recommends treating all patients with atrial fibrillation and high risk of stroke with warfarin. Their definition of high-risk includes any patient with 1 or more of the following risk factors: age >75 years, prior ischemic stroke, transient ischemic attack or systemic embolism, congestive heart failure, impaired left ventricular systolic function, hypertension, or diabetes mellitus.²
• In contrast, the 2006 guidelines for the management of patients with atrial fibrillation from the American College of Cardiology, American Heart Association, and European Society of Cardiology, are more conservative. They recommend that patients with more than 1 risk factor take warfarin, and patients with only 1 risk factor (for example, a patient older than 75 years of age with no other risk factors) take either warfarin or aspirin.³

Clinical context: Reasonable concerns

Fewer than half of the 10% to 12% of people older than 75 with atrial fibrillation are taking warfarin for stroke prevention. In one study, only 35% of patients 85 years and older with no known contraindication to anticoagulation received warfarin.⁴ Possible reasons for this low rate include:

• cost of monitoring warfarin
• concerns about compliance
• increased risk of hemorrhage
• prior studies focused on younger patients, in closely monitored settings.

These factors lead us to speculate that many physicians believe that the risks of warfarin in elderly patients in primary care settings outweigh any potential benefit.

We think this study demonstrates that we should seriously discuss and consider warfarin therapy for most of our elderly patients with atrial fibrillation.

Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation

Study summary: Primary care setting, elders only

This prospective randomized open-label trial was designed to test the effectiveness and safety of warfarin vs aspirin in the elderly, in a realistic primary care setting. The study compared the frequency of stroke, intracranial hemorrhage, and other significant arterial embolism in patients taking either warfarin or aspirin.

Inclusion criteria. Patients were at least 75 years old (average 81.5 years) with an ECG within the previous 2 years showing atrial fibrillation or atrial flutter. Seventy percent of the patients had been previously diagnosed with atrial fibrillation and 30% were identified because they had an irregular pulse on exam.

Exclusion criteria included rheumatic heart disease, major nontraumatic hemorrhage in the past 5 years, intracranial hemorrhage, endoscopically proven peptic ulcer disease in the past year, esophageal varices, allergy to either study drug, terminal illness, surgery in past 3 months, blood pressure greater than 180/110 mm Hg, or if the primary physician judged that a patient should either be on warfarin or not, based on risk factors.

Patient characteristics. The patients were recruited from 260 general practices in England and Wales. At baseline, 39% to 40% of the patients were already taking warfarin, 12% to 13% had had a prior stroke, 53% to 55% had hypertension, 13% to 14% had diabetes, 19% to 20% had heart failure, and 10% to 12% had a history of myocardial infarction. Patients were followed for an average of 2.7 years.

Aspirin and warfarin regimens. Patients were assigned to either aspirin at a dose of 75 mg/day or warfarin with a target international normalized ratio (INR) of 2.5 and an acceptable range of 2 to 3. Because the study aimed to reflect a realistic primary care setting, the frequency and method of INR testing was left to the discretion of participating physicians.

Patients who had been taking aspirin or warfarin prior to the study discontinued that medicine if they were assigned to the other treatment. Sixty-seven percent of the patients assigned to warfarin continued this treatment throughout the study, and 78% of those who either stopped taking warfarin or never started it were put on either aspirin or clopidogrel. Seventy-six percent of the patients assigned to aspirin took the medicine for the entire study period, while 70% of those who stopped taking aspirin or never started it were either switched to or stayed on warfarin.

INR values. Patients on warfarin had INR values between 2.0 and 3.0 for 67% of the time, below range for 19%, of the time, and above range for 14% of the time. Twenty-two percent of practices had all components of INR monitoring done at the hospital (phlebotomy, INR analysis, and warfarin dosing), 19% of the practices completed all 3 components on site, and the remaining practices had various combinations of onsite and hospital monitoring.

The primary outcomes included disabling stroke (ischemic or hemorrhagic) or clinically significant arterial embolism. There were 24 primary events (1.8% per year) in patients assigned to warfarin compared with 48 primary events (3.8% per year) in those assigned to aspirin, with a relative risk of 0.48 (95% confidence interval [CI], 0.28–0.80 (TABLE). The number needed to treat for 1 year to prevent 1 primary event was 50, when warfarin was compared to aspirin. Warfarin was superior to aspirin in all subgroup analyses, including patients over 85 years old.

Secondary outcomes. There were no significant differences between the warfarin and aspirin groups in the secondary outcomes: hospital admission or death as a result of a non-stroke vascular event (6.1% risk per year with warfarin vs 6.3% risk per year with aspirin), all-cause mortality (8.0% vs 8.4%), and major extracranial hemorrhage (1.4% vs 1.6%). Patients assigned to warfarin, including the subgroup of patients older than 85, did not have an increased risk of a major hemorrhage when compared with those assigned to aspirin (1.9% risk per year with warfarin vs 2.0% risk per year with aspirin; relative risk=0.96; 95% CI, 0.53–1.75).¹

TABLE
BAFTA study: Warfarin was as safe as aspirin and more effective in preventing stroke in the elderly

What’s new?: Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation.

This is the first randomized controlled trial of warfarin for atrial fibrillation that included only patients ages 75 and older, conducted in a primary care setting.⁵

Limitations of earlier studies. The most recent meta-analysis of antithrombotic therapy for stroke prevention in patients with atrial fibrillation included 29 trials with 28,044 patients. This analysis concluded that although both warfarin and aspirin are effective in reducing the risk of stroke in patients with atrial fibrillation (warfarin by 60% and aspirin by 20%), warfarin was more effective than aspirin (relative risk reduction of 39%), with very small (≤0.3% per year) absolute increases in major extracranial hemorrhage.

The average age of patients in those trials, however, was 71. The authors identified the lack of data on older patients (who are at higher risk for serious bleeding events) as a limitation of the meta-analysis. Many of these trials took place in settings with closer monitoring of INR and warfarin dosing than is customary in a primary care setting.⁵

Caveats: Consider the evidence on benefits and risks

Major bleeding from warfarin is a concern, especially in the elderly. A recent cohort study⁶ (summarized as a POEM in this journal⁷) reported high rates of major bleeding (13.1 per hundred person-years or 13.1%) in patients ≥80 years of age during their first year of warfarin therapy. Despite the high risk of bleeding events in this cohort study, there was considerable benefit from warfarin therapy.

None of the patients who remained on warfarin had a thrombotic stroke (personal communication with Dr Hylek by the author). The expected rate of thrombotic stroke is in the range of 5% to 6% per year in this high-risk group.

Furthermore, most of the bleeding events were gastrointestinal and did not lead to catastrophic outcomes.

Do not add warfarin to aspirin in patients >75 years

Dr Hylek also noted that 40% of the patients in their cohort study were taking both warfarin and aspirin, and, although her study did not have sufficient power to detect a difference, prior studies noted increased risk of bleeding with this combination compared to warfarin alone.^8,9 For this reason we think the combination of warfarin and aspirin should be avoided in patients over 75.

Target INR <3

Our caveat is the same as the POEM author’s conclusion:⁷ Patients over 80 should be carefully monitored to keep the INR below 3.0 or for signs of bleeding, especially in the first 90 days of therapy when bleeding is more likely to occur.

A final point that the BAFTA authors make, which is worth repeating here, is that the prior studies showing an increased risk of bleeding complications had INR target rates of 4 to 5, whereas the target in this study was 2 to 3. Two previous studies that also compared aspirin to warfarin with an INR goal of 2 to 3 similarly showed no difference in major bleeding between the 2 groups.^10,11

Challenges to Implementation: Meticulous monitoring, patient education

• Managing warfarin therapy requires meticulous care to avoid complications and optimize treatment effect.
• Patients may be reluctant to take warfarin because they may fear bleeding.
• Patients who do agree to take warfarin need education about possible medication interactions, the need for regular INR monitoring, dosage changes, and dietary issues (eg, maintaining a consistent intake of foods containing vitamin K).

Contraindications

Contraindications to the use of warfarin include hypersensitivity to warfarin, severe hepatic disease, alcoholism, recent trauma or surgery, history of falling or significant risk of falls, and active gastrointestinal, respiratory, or genitourinary bleeding.

INR testing systems

Several randomized trials support the use of monitoring systems such as a pharmacist managed anticoagulation service or decision support software, both of which can improve the percentage of patients with therapeutic INR values.^12,13

Using point-of-care INR tests in the office provides immediate results which allow for more timely adjustments of warfarin dose.¹⁴

PURLs methodology

This study was selected and evaluated using the Family Physician Inquiries Network’s Priority Updates from the Research Literature Surveillance System (PURLs) methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed here.

Illustrative Case

An 85-year-old woman with hypertension and chronic atrial fibrillation has transferred her care to you. She takes an aspirin a day for cardiovascular prevention. You know that warfarin is better than aspirin for preventing stroke but worry about the increased risk of bleeding with warfarin.

Should you recommend that she stay on aspirin or switch to warfarin?

Background: BAFTA: A realistic study

We have been reluctant to use warfarin in elders with atrial fibrillation for good reason: risk of hemorrhage. Since there are few trials looking at use of warfarin among elders in primary care settings, we are uncertain about the balance of benefits and harms.

The BAFTA study¹ is the first trial to compare outcomes of warfarin vs aspirin in elders specifically, in the less-than-ideal conditions of real life.

Guidelines mirror uncertainties

This uncertainty is reflected even in guidelines for anticoagulation in elderly patients with atrial fibrillation.

• The 2004 American College of Chest Physicians Seventh Conference on Antithrombotic and Thrombolytic Therapy recommends treating all patients with atrial fibrillation and high risk of stroke with warfarin. Their definition of high-risk includes any patient with 1 or more of the following risk factors: age >75 years, prior ischemic stroke, transient ischemic attack or systemic embolism, congestive heart failure, impaired left ventricular systolic function, hypertension, or diabetes mellitus.²
• In contrast, the 2006 guidelines for the management of patients with atrial fibrillation from the American College of Cardiology, American Heart Association, and European Society of Cardiology, are more conservative. They recommend that patients with more than 1 risk factor take warfarin, and patients with only 1 risk factor (for example, a patient older than 75 years of age with no other risk factors) take either warfarin or aspirin.³

Clinical context: Reasonable concerns

Fewer than half of the 10% to 12% of people older than 75 with atrial fibrillation are taking warfarin for stroke prevention. In one study, only 35% of patients 85 years and older with no known contraindication to anticoagulation received warfarin.⁴ Possible reasons for this low rate include:

• cost of monitoring warfarin
• concerns about compliance
• increased risk of hemorrhage
• prior studies focused on younger patients, in closely monitored settings.

These factors lead us to speculate that many physicians believe that the risks of warfarin in elderly patients in primary care settings outweigh any potential benefit.

We think this study demonstrates that we should seriously discuss and consider warfarin therapy for most of our elderly patients with atrial fibrillation.

Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation

Study summary: Primary care setting, elders only

This prospective randomized open-label trial was designed to test the effectiveness and safety of warfarin vs aspirin in the elderly, in a realistic primary care setting. The study compared the frequency of stroke, intracranial hemorrhage, and other significant arterial embolism in patients taking either warfarin or aspirin.

Inclusion criteria. Patients were at least 75 years old (average 81.5 years) with an ECG within the previous 2 years showing atrial fibrillation or atrial flutter. Seventy percent of the patients had been previously diagnosed with atrial fibrillation and 30% were identified because they had an irregular pulse on exam.

Exclusion criteria included rheumatic heart disease, major nontraumatic hemorrhage in the past 5 years, intracranial hemorrhage, endoscopically proven peptic ulcer disease in the past year, esophageal varices, allergy to either study drug, terminal illness, surgery in past 3 months, blood pressure greater than 180/110 mm Hg, or if the primary physician judged that a patient should either be on warfarin or not, based on risk factors.

Patient characteristics. The patients were recruited from 260 general practices in England and Wales. At baseline, 39% to 40% of the patients were already taking warfarin, 12% to 13% had had a prior stroke, 53% to 55% had hypertension, 13% to 14% had diabetes, 19% to 20% had heart failure, and 10% to 12% had a history of myocardial infarction. Patients were followed for an average of 2.7 years.

Aspirin and warfarin regimens. Patients were assigned to either aspirin at a dose of 75 mg/day or warfarin with a target international normalized ratio (INR) of 2.5 and an acceptable range of 2 to 3. Because the study aimed to reflect a realistic primary care setting, the frequency and method of INR testing was left to the discretion of participating physicians.

Patients who had been taking aspirin or warfarin prior to the study discontinued that medicine if they were assigned to the other treatment. Sixty-seven percent of the patients assigned to warfarin continued this treatment throughout the study, and 78% of those who either stopped taking warfarin or never started it were put on either aspirin or clopidogrel. Seventy-six percent of the patients assigned to aspirin took the medicine for the entire study period, while 70% of those who stopped taking aspirin or never started it were either switched to or stayed on warfarin.

INR values. Patients on warfarin had INR values between 2.0 and 3.0 for 67% of the time, below range for 19%, of the time, and above range for 14% of the time. Twenty-two percent of practices had all components of INR monitoring done at the hospital (phlebotomy, INR analysis, and warfarin dosing), 19% of the practices completed all 3 components on site, and the remaining practices had various combinations of onsite and hospital monitoring.

The primary outcomes included disabling stroke (ischemic or hemorrhagic) or clinically significant arterial embolism. There were 24 primary events (1.8% per year) in patients assigned to warfarin compared with 48 primary events (3.8% per year) in those assigned to aspirin, with a relative risk of 0.48 (95% confidence interval [CI], 0.28–0.80 (TABLE). The number needed to treat for 1 year to prevent 1 primary event was 50, when warfarin was compared to aspirin. Warfarin was superior to aspirin in all subgroup analyses, including patients over 85 years old.

Secondary outcomes. There were no significant differences between the warfarin and aspirin groups in the secondary outcomes: hospital admission or death as a result of a non-stroke vascular event (6.1% risk per year with warfarin vs 6.3% risk per year with aspirin), all-cause mortality (8.0% vs 8.4%), and major extracranial hemorrhage (1.4% vs 1.6%). Patients assigned to warfarin, including the subgroup of patients older than 85, did not have an increased risk of a major hemorrhage when compared with those assigned to aspirin (1.9% risk per year with warfarin vs 2.0% risk per year with aspirin; relative risk=0.96; 95% CI, 0.53–1.75).¹

TABLE
BAFTA study: Warfarin was as safe as aspirin and more effective in preventing stroke in the elderly

What’s new?: Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation.

This is the first randomized controlled trial of warfarin for atrial fibrillation that included only patients ages 75 and older, conducted in a primary care setting.⁵

Limitations of earlier studies. The most recent meta-analysis of antithrombotic therapy for stroke prevention in patients with atrial fibrillation included 29 trials with 28,044 patients. This analysis concluded that although both warfarin and aspirin are effective in reducing the risk of stroke in patients with atrial fibrillation (warfarin by 60% and aspirin by 20%), warfarin was more effective than aspirin (relative risk reduction of 39%), with very small (≤0.3% per year) absolute increases in major extracranial hemorrhage.

The average age of patients in those trials, however, was 71. The authors identified the lack of data on older patients (who are at higher risk for serious bleeding events) as a limitation of the meta-analysis. Many of these trials took place in settings with closer monitoring of INR and warfarin dosing than is customary in a primary care setting.⁵

Caveats: Consider the evidence on benefits and risks

Major bleeding from warfarin is a concern, especially in the elderly. A recent cohort study⁶ (summarized as a POEM in this journal⁷) reported high rates of major bleeding (13.1 per hundred person-years or 13.1%) in patients ≥80 years of age during their first year of warfarin therapy. Despite the high risk of bleeding events in this cohort study, there was considerable benefit from warfarin therapy.

None of the patients who remained on warfarin had a thrombotic stroke (personal communication with Dr Hylek by the author). The expected rate of thrombotic stroke is in the range of 5% to 6% per year in this high-risk group.

Furthermore, most of the bleeding events were gastrointestinal and did not lead to catastrophic outcomes.

Do not add warfarin to aspirin in patients >75 years

Dr Hylek also noted that 40% of the patients in their cohort study were taking both warfarin and aspirin, and, although her study did not have sufficient power to detect a difference, prior studies noted increased risk of bleeding with this combination compared to warfarin alone.^8,9 For this reason we think the combination of warfarin and aspirin should be avoided in patients over 75.

Target INR <3

Our caveat is the same as the POEM author’s conclusion:⁷ Patients over 80 should be carefully monitored to keep the INR below 3.0 or for signs of bleeding, especially in the first 90 days of therapy when bleeding is more likely to occur.

A final point that the BAFTA authors make, which is worth repeating here, is that the prior studies showing an increased risk of bleeding complications had INR target rates of 4 to 5, whereas the target in this study was 2 to 3. Two previous studies that also compared aspirin to warfarin with an INR goal of 2 to 3 similarly showed no difference in major bleeding between the 2 groups.^10,11

Challenges to Implementation: Meticulous monitoring, patient education

• Managing warfarin therapy requires meticulous care to avoid complications and optimize treatment effect.
• Patients may be reluctant to take warfarin because they may fear bleeding.
• Patients who do agree to take warfarin need education about possible medication interactions, the need for regular INR monitoring, dosage changes, and dietary issues (eg, maintaining a consistent intake of foods containing vitamin K).

Contraindications

Contraindications to the use of warfarin include hypersensitivity to warfarin, severe hepatic disease, alcoholism, recent trauma or surgery, history of falling or significant risk of falls, and active gastrointestinal, respiratory, or genitourinary bleeding.

INR testing systems

Several randomized trials support the use of monitoring systems such as a pharmacist managed anticoagulation service or decision support software, both of which can improve the percentage of patients with therapeutic INR values.^12,13

Using point-of-care INR tests in the office provides immediate results which allow for more timely adjustments of warfarin dose.¹⁴

PURLs methodology

This study was selected and evaluated using the Family Physician Inquiries Network’s Priority Updates from the Research Literature Surveillance System (PURLs) methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed here.

Illustrative Case

An 85-year-old woman with hypertension and chronic atrial fibrillation has transferred her care to you. She takes an aspirin a day for cardiovascular prevention. You know that warfarin is better than aspirin for preventing stroke but worry about the increased risk of bleeding with warfarin.

Should you recommend that she stay on aspirin or switch to warfarin?

Background: BAFTA: A realistic study

We have been reluctant to use warfarin in elders with atrial fibrillation for good reason: risk of hemorrhage. Since there are few trials looking at use of warfarin among elders in primary care settings, we are uncertain about the balance of benefits and harms.

The BAFTA study¹ is the first trial to compare outcomes of warfarin vs aspirin in elders specifically, in the less-than-ideal conditions of real life.

Guidelines mirror uncertainties

This uncertainty is reflected even in guidelines for anticoagulation in elderly patients with atrial fibrillation.

• The 2004 American College of Chest Physicians Seventh Conference on Antithrombotic and Thrombolytic Therapy recommends treating all patients with atrial fibrillation and high risk of stroke with warfarin. Their definition of high-risk includes any patient with 1 or more of the following risk factors: age >75 years, prior ischemic stroke, transient ischemic attack or systemic embolism, congestive heart failure, impaired left ventricular systolic function, hypertension, or diabetes mellitus.²
• In contrast, the 2006 guidelines for the management of patients with atrial fibrillation from the American College of Cardiology, American Heart Association, and European Society of Cardiology, are more conservative. They recommend that patients with more than 1 risk factor take warfarin, and patients with only 1 risk factor (for example, a patient older than 75 years of age with no other risk factors) take either warfarin or aspirin.³

Clinical context: Reasonable concerns

Fewer than half of the 10% to 12% of people older than 75 with atrial fibrillation are taking warfarin for stroke prevention. In one study, only 35% of patients 85 years and older with no known contraindication to anticoagulation received warfarin.⁴ Possible reasons for this low rate include:

• cost of monitoring warfarin
• concerns about compliance
• increased risk of hemorrhage
• prior studies focused on younger patients, in closely monitored settings.

These factors lead us to speculate that many physicians believe that the risks of warfarin in elderly patients in primary care settings outweigh any potential benefit.

We think this study demonstrates that we should seriously discuss and consider warfarin therapy for most of our elderly patients with atrial fibrillation.

Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation

Study summary: Primary care setting, elders only

This prospective randomized open-label trial was designed to test the effectiveness and safety of warfarin vs aspirin in the elderly, in a realistic primary care setting. The study compared the frequency of stroke, intracranial hemorrhage, and other significant arterial embolism in patients taking either warfarin or aspirin.

Inclusion criteria. Patients were at least 75 years old (average 81.5 years) with an ECG within the previous 2 years showing atrial fibrillation or atrial flutter. Seventy percent of the patients had been previously diagnosed with atrial fibrillation and 30% were identified because they had an irregular pulse on exam.

Exclusion criteria included rheumatic heart disease, major nontraumatic hemorrhage in the past 5 years, intracranial hemorrhage, endoscopically proven peptic ulcer disease in the past year, esophageal varices, allergy to either study drug, terminal illness, surgery in past 3 months, blood pressure greater than 180/110 mm Hg, or if the primary physician judged that a patient should either be on warfarin or not, based on risk factors.

Patient characteristics. The patients were recruited from 260 general practices in England and Wales. At baseline, 39% to 40% of the patients were already taking warfarin, 12% to 13% had had a prior stroke, 53% to 55% had hypertension, 13% to 14% had diabetes, 19% to 20% had heart failure, and 10% to 12% had a history of myocardial infarction. Patients were followed for an average of 2.7 years.

Aspirin and warfarin regimens. Patients were assigned to either aspirin at a dose of 75 mg/day or warfarin with a target international normalized ratio (INR) of 2.5 and an acceptable range of 2 to 3. Because the study aimed to reflect a realistic primary care setting, the frequency and method of INR testing was left to the discretion of participating physicians.

Patients who had been taking aspirin or warfarin prior to the study discontinued that medicine if they were assigned to the other treatment. Sixty-seven percent of the patients assigned to warfarin continued this treatment throughout the study, and 78% of those who either stopped taking warfarin or never started it were put on either aspirin or clopidogrel. Seventy-six percent of the patients assigned to aspirin took the medicine for the entire study period, while 70% of those who stopped taking aspirin or never started it were either switched to or stayed on warfarin.

INR values. Patients on warfarin had INR values between 2.0 and 3.0 for 67% of the time, below range for 19%, of the time, and above range for 14% of the time. Twenty-two percent of practices had all components of INR monitoring done at the hospital (phlebotomy, INR analysis, and warfarin dosing), 19% of the practices completed all 3 components on site, and the remaining practices had various combinations of onsite and hospital monitoring.

The primary outcomes included disabling stroke (ischemic or hemorrhagic) or clinically significant arterial embolism. There were 24 primary events (1.8% per year) in patients assigned to warfarin compared with 48 primary events (3.8% per year) in those assigned to aspirin, with a relative risk of 0.48 (95% confidence interval [CI], 0.28–0.80 (TABLE). The number needed to treat for 1 year to prevent 1 primary event was 50, when warfarin was compared to aspirin. Warfarin was superior to aspirin in all subgroup analyses, including patients over 85 years old.

Secondary outcomes. There were no significant differences between the warfarin and aspirin groups in the secondary outcomes: hospital admission or death as a result of a non-stroke vascular event (6.1% risk per year with warfarin vs 6.3% risk per year with aspirin), all-cause mortality (8.0% vs 8.4%), and major extracranial hemorrhage (1.4% vs 1.6%). Patients assigned to warfarin, including the subgroup of patients older than 85, did not have an increased risk of a major hemorrhage when compared with those assigned to aspirin (1.9% risk per year with warfarin vs 2.0% risk per year with aspirin; relative risk=0.96; 95% CI, 0.53–1.75).¹

TABLE
BAFTA study: Warfarin was as safe as aspirin and more effective in preventing stroke in the elderly

What’s new?: Age alone does not preclude warfarin

The key finding from the BAFTA study is that advanced age alone is not a contraindication to the use of warfarin for stroke prevention in elderly patients with atrial fibrillation.

This is the first randomized controlled trial of warfarin for atrial fibrillation that included only patients ages 75 and older, conducted in a primary care setting.⁵

Limitations of earlier studies. The most recent meta-analysis of antithrombotic therapy for stroke prevention in patients with atrial fibrillation included 29 trials with 28,044 patients. This analysis concluded that although both warfarin and aspirin are effective in reducing the risk of stroke in patients with atrial fibrillation (warfarin by 60% and aspirin by 20%), warfarin was more effective than aspirin (relative risk reduction of 39%), with very small (≤0.3% per year) absolute increases in major extracranial hemorrhage.

The average age of patients in those trials, however, was 71. The authors identified the lack of data on older patients (who are at higher risk for serious bleeding events) as a limitation of the meta-analysis. Many of these trials took place in settings with closer monitoring of INR and warfarin dosing than is customary in a primary care setting.⁵

Caveats: Consider the evidence on benefits and risks

Major bleeding from warfarin is a concern, especially in the elderly. A recent cohort study⁶ (summarized as a POEM in this journal⁷) reported high rates of major bleeding (13.1 per hundred person-years or 13.1%) in patients ≥80 years of age during their first year of warfarin therapy. Despite the high risk of bleeding events in this cohort study, there was considerable benefit from warfarin therapy.

None of the patients who remained on warfarin had a thrombotic stroke (personal communication with Dr Hylek by the author). The expected rate of thrombotic stroke is in the range of 5% to 6% per year in this high-risk group.

Furthermore, most of the bleeding events were gastrointestinal and did not lead to catastrophic outcomes.

Do not add warfarin to aspirin in patients >75 years

Dr Hylek also noted that 40% of the patients in their cohort study were taking both warfarin and aspirin, and, although her study did not have sufficient power to detect a difference, prior studies noted increased risk of bleeding with this combination compared to warfarin alone.^8,9 For this reason we think the combination of warfarin and aspirin should be avoided in patients over 75.

Target INR <3

Our caveat is the same as the POEM author’s conclusion:⁷ Patients over 80 should be carefully monitored to keep the INR below 3.0 or for signs of bleeding, especially in the first 90 days of therapy when bleeding is more likely to occur.

A final point that the BAFTA authors make, which is worth repeating here, is that the prior studies showing an increased risk of bleeding complications had INR target rates of 4 to 5, whereas the target in this study was 2 to 3. Two previous studies that also compared aspirin to warfarin with an INR goal of 2 to 3 similarly showed no difference in major bleeding between the 2 groups.^10,11

Challenges to Implementation: Meticulous monitoring, patient education

• Managing warfarin therapy requires meticulous care to avoid complications and optimize treatment effect.
• Patients may be reluctant to take warfarin because they may fear bleeding.
• Patients who do agree to take warfarin need education about possible medication interactions, the need for regular INR monitoring, dosage changes, and dietary issues (eg, maintaining a consistent intake of foods containing vitamin K).

Contraindications

Contraindications to the use of warfarin include hypersensitivity to warfarin, severe hepatic disease, alcoholism, recent trauma or surgery, history of falling or significant risk of falls, and active gastrointestinal, respiratory, or genitourinary bleeding.

INR testing systems

Several randomized trials support the use of monitoring systems such as a pharmacist managed anticoagulation service or decision support software, both of which can improve the percentage of patients with therapeutic INR values.^12,13

Using point-of-care INR tests in the office provides immediate results which allow for more timely adjustments of warfarin dose.¹⁴

PURLs methodology

This study was selected and evaluated using the Family Physician Inquiries Network’s Priority Updates from the Research Literature Surveillance System (PURLs) methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed here.

Why are older African Americans less likely than whites to get a flu vaccination? Despite the existence of an effective flu vaccine, usage rates still remain low: 66% on average,¹ and are even lower among minority groups.² One survey found only 47% of older African Americans were vaccinated against the flu.³

These disparities are not easily explained. Even when controlling for increased risk, age distribution, perceived health status, family size and marital status, poverty level, education, and access to medical care and health insurance, African Americans are still less likely than whites to get vaccinated.^3,4

In light of this disparity, we explored barriers to flu vaccination among this population, including concerns over safety and adverse events, and the role that health care providers can play in overcoming these issues.

What are the barriers?

Structural factors, such as having access to a location that provides vaccinations,⁵ and social factors, such as believing that others support vaccination,⁶ increase vaccination rates. Vaccination increases with age^5,7 and is more likely in those who have previously been vaccinated.⁸

People who believe they aren’t susceptible to a disease are less likely to get vaccinated.^9,10 Fear of side effects is also a significant barrier.^2,6,7,11,12 Fear of illness from the vaccine is often cited by both white and African American Medicare patients, older adults, and most notably by older African Americans.^2,5,12 African Americans and other minority groups are less likely to accept the vaccine as safe.¹³

A physician’s recommendation to get the flu vaccine appears to motivate patients in risk groups to get vaccinated,^6,7,12 and is a significant determinant of vaccine acceptance in surveys of older Americans, high-risk patients, and older African Americans in particular.^6,12

Methods

Focus groups and interviews

The study team conducted qualitative formative research with 2 pertinent audiences: African American adults (hereafter, “public participants”) and clinicians who administer the flu vaccine (hereafter, “providers”). The Saint Louis University Institutional Review Board approved the research.

We recruited older African American public participants through local community contacts, and screened to identify those ambivalent about getting a flu shot. A trained moderator then conducted focus groups and interviews with those who were ambivalent about the flu shot to assess their knowledge, beliefs, norms, and intentions related to vaccination.

We also recruited vaccine providers by calling local hospitals, doctors’ offices, health departments, and clinics that serve African Americans. To identify providers’ perceptions of patients’ concerns, we carried out interviews at the clinicians’ place of business or in a university conference room; a focus group was conducted at a community clinic.

All focus groups and interviews were audiotaped and transcribed. In pairs, research team members (all authors except for HJ) coded each of the transcripts independently, reviewed and discussed their codes, and then came to agreement on the final codes. Coded transcripts were entered into Atlas.ti (Atlas.ti GmbH, Berlin, Germany), a qualitative data analysis software program, and were analyzed with summary reports drafted for each focus group and interview. Findings were then synthesized across groups and interviews, and across segments.

Results

Description of participants

Four focus groups (N=35) and 8 in-depth interviews were conducted with the public participants—African American adults ages 50 and over. As shown in the TABLE, most participants were female, had children, had less than a college degree, and earned less than $30,000 a year. Widows and widowers made up the largest percentage of participants.

One focus group (N=9) and 5 interviews were conducted with vaccine providers—professionals working in clinics that offered the flu vaccine and served African Americans—for a total sample of 14. As the providers were sampled according to their professional affiliation, demographics were not systematically recorded. These informants were nurses, physicians’ assistants, and vaccination program administrators.

TABLE
4 focus groups and 8 interviews included 43 public participants*

How patients and clinicians see things differently

Susceptibility. Most public participants were familiar with only a few of the high-risk groups recommended for vaccination. Many participants said that children needed the flu shot, but did not state that those in their 50s were also recommended to get the flu shot. Those never stricken with the flu didn’t consider themselves susceptible. “Because I have never gotten it before…. I get a lot of other things but I don’t get colds and things like that.”

The health care providers identified senior citizens as a high-risk group for influenza because of the high occurrence of comorbid conditions. Providers were especially concerned that seniors who don’t go out much may incorrectly believe they don’t need to get vaccinated.

Severity of the flu. While some of the public participants didn’t see influenza as a deadly disease, others did.

Providers recognized that some members of the public regarded the flu as more of a nuisance then a serious illness and emphasized the need to raise awareness.

Vaccine efficacy. The opinions that the public participants had about the effectiveness of the influenza vaccine varied. Some said it was effective; others believed the vaccine was not effective because they knew they could still get the flu even after being vaccinated. “I know people who got the flu shot and they still got the flu.” Many believed that home remedies, cleanliness, and staying away from others were more effective means of prevention.

Providers believed the vaccine is effective if it matches the correct strain of flu virus. They noted that even if a patient is vaccinated against the wrong strain, flu symptoms will be milder. Providers recognized patient concerns about vaccine efficacy, but none mentioned that it was important to encourage vaccination along with cleanliness and avoidance measures.

Safety and side effects. The most common concern about vaccine safety that the public participants discussed was that the influenza vaccine causes flu illness. Several respondents substantiated this concern with their own experiences, or those of others, where shortly after vaccination a flulike illness resulted. “It gave them the flu. And they were sick for 2 or 3 weeks.”

Another common safety concern was that the vaccine would interact with prescription medications for chronic illnesses. Many public participants also noted that health care providers neglected to discuss the matter. “[M]y reason for not taking the flu shot is because I’m on other medicine and I do have some concerns.… What are you putting in my body?” Some were also concerned about the safety of vaccine components and receiving a tainted vaccine. These participants also mentioned their fear of an allergic reaction to the flu vaccine.

Overall, providers did not have concerns about vaccine safety; however, they understood patients were afraid the influenza vaccine would give them the flu. They felt they should inform patients that full protection from the influenza vaccine takes up to 2 weeks. Providers were also aware that many African Americans who are 50 years of age and older distrust the medical system.

Main reason for vaccination: Doctor’s advice

All of the African American adults agreed that physicians and other health care professionals were important sources of vaccine information. Though initially ambivalent, a majority also reported receiving the vaccine primarily because of recommendations from their doctor. Some noted that providers can do more to encourage vaccination when patients express concerns. “I would have taken it if he had said, ‘I think you should take it.’” The public participants also got vaccine information from family, friends, broadcast media, and print material. They noted that they wanted to see the following in flu vaccination information: the pros and cons of the vaccine, efficacy of the vaccine, how the vaccine reduces flu severity, vaccine safety, and history and background of the vaccine.

Providers noted the importance of: promoting informed decision making, one-on-one communication, using a matriarchal figure to promote vaccination and using the media to promote vaccination. “You would have to have a mother, or a grandmother, or aunt figure because that is usually who is responsible, who takes care of the family.” Providers also pointed out that patients often voice concerns about not having enough information to make informed decisions, and that patients rely on convenience and doctors’ recommendations when deciding about vaccination.

Discussion

Patients may distrust the system, but they trust their doctor

Our study’s findings from both providers and older African Americans suggest that physicians are the most influential source of information when patients are deciding about flu vaccination. This is true despite the fact that the public makes no secret of its distrust of the medical system and the safety of the vaccine. The African American participants also suggested that physicians do not adequately address patient concerns through discussion and the information they provide.

Providers were concerned that many people do not believe the flu is a severe illness or that they are susceptible. Although the African Americans in the study recognized some high-risk groups, they tended not to consider themselves part of any of those groups.

Doctors didn’t realize that patients fear drug interactions

One of the interesting findings of this study was that armed with the knowledge that the flu shot does not guarantee flu prevention, African American adults were willing to forego the vaccine. This absence of a guarantee also reinforced their beliefs that other prevention methods, such as handwashing and avoidance, are more effective.

Not surprisingly, of course, was the finding that patients continue to avoid the flu shot for fear of getting the flu.^2,6 What was a bit surprising was that providers did not recognize that fear of medication reactions (drug interactions and allergic reactions) was also a barrier to flu vaccination. Providers also missed out on an educational opportunity, since many of the African American participants wanted to discuss the possibility of interactions with them.

Limitations of the study

External validity is limited because the findings cannot be generalized to every African American population in the US. The participants made up a non-random convenience sample of older African Americans in a Midwestern city, although the community-based recruitment strategy succeeded in reaching members of a lower income urban population. This study included only those who were ambivalent about the vaccine and who were open to both the pros and cons of vaccination. Project staff minimized possible interviewer bias by using experienced moderators, ensuring the consistent use of moderator guides, and using consensus coding procedures.

Funding

This research was funded by grant #6465 from the National Immunization Program at the Centers for Disease Control and Prevention, via Special Interest Project 11, to the Prevention research Center at the Saint louis University School of Public Health.

Acknowledgments

We thank Katie Duggan at the School of Public Health and edith Gary and Pascale Wortley at the CDC for their support. This article is dedicated to the memory of Joe D. Wray, MD, who suggested it.

Correspondence
Ricardo J. Wray, PhD, Community Health, Saint Louis University School of Public Health, 3545 Lafayette Avenue, St. Louis, MO 63104; wray@slu.edu

Why are older African Americans less likely than whites to get a flu vaccination? Despite the existence of an effective flu vaccine, usage rates still remain low: 66% on average,¹ and are even lower among minority groups.² One survey found only 47% of older African Americans were vaccinated against the flu.³

These disparities are not easily explained. Even when controlling for increased risk, age distribution, perceived health status, family size and marital status, poverty level, education, and access to medical care and health insurance, African Americans are still less likely than whites to get vaccinated.^3,4

In light of this disparity, we explored barriers to flu vaccination among this population, including concerns over safety and adverse events, and the role that health care providers can play in overcoming these issues.

What are the barriers?

Structural factors, such as having access to a location that provides vaccinations,⁵ and social factors, such as believing that others support vaccination,⁶ increase vaccination rates. Vaccination increases with age^5,7 and is more likely in those who have previously been vaccinated.⁸

People who believe they aren’t susceptible to a disease are less likely to get vaccinated.^9,10 Fear of side effects is also a significant barrier.^2,6,7,11,12 Fear of illness from the vaccine is often cited by both white and African American Medicare patients, older adults, and most notably by older African Americans.^2,5,12 African Americans and other minority groups are less likely to accept the vaccine as safe.¹³

A physician’s recommendation to get the flu vaccine appears to motivate patients in risk groups to get vaccinated,^6,7,12 and is a significant determinant of vaccine acceptance in surveys of older Americans, high-risk patients, and older African Americans in particular.^6,12

Methods

Focus groups and interviews

The study team conducted qualitative formative research with 2 pertinent audiences: African American adults (hereafter, “public participants”) and clinicians who administer the flu vaccine (hereafter, “providers”). The Saint Louis University Institutional Review Board approved the research.

We recruited older African American public participants through local community contacts, and screened to identify those ambivalent about getting a flu shot. A trained moderator then conducted focus groups and interviews with those who were ambivalent about the flu shot to assess their knowledge, beliefs, norms, and intentions related to vaccination.

We also recruited vaccine providers by calling local hospitals, doctors’ offices, health departments, and clinics that serve African Americans. To identify providers’ perceptions of patients’ concerns, we carried out interviews at the clinicians’ place of business or in a university conference room; a focus group was conducted at a community clinic.

All focus groups and interviews were audiotaped and transcribed. In pairs, research team members (all authors except for HJ) coded each of the transcripts independently, reviewed and discussed their codes, and then came to agreement on the final codes. Coded transcripts were entered into Atlas.ti (Atlas.ti GmbH, Berlin, Germany), a qualitative data analysis software program, and were analyzed with summary reports drafted for each focus group and interview. Findings were then synthesized across groups and interviews, and across segments.

Results

Description of participants

Four focus groups (N=35) and 8 in-depth interviews were conducted with the public participants—African American adults ages 50 and over. As shown in the TABLE, most participants were female, had children, had less than a college degree, and earned less than $30,000 a year. Widows and widowers made up the largest percentage of participants.

One focus group (N=9) and 5 interviews were conducted with vaccine providers—professionals working in clinics that offered the flu vaccine and served African Americans—for a total sample of 14. As the providers were sampled according to their professional affiliation, demographics were not systematically recorded. These informants were nurses, physicians’ assistants, and vaccination program administrators.

TABLE
4 focus groups and 8 interviews included 43 public participants*

How patients and clinicians see things differently

Susceptibility. Most public participants were familiar with only a few of the high-risk groups recommended for vaccination. Many participants said that children needed the flu shot, but did not state that those in their 50s were also recommended to get the flu shot. Those never stricken with the flu didn’t consider themselves susceptible. “Because I have never gotten it before…. I get a lot of other things but I don’t get colds and things like that.”

The health care providers identified senior citizens as a high-risk group for influenza because of the high occurrence of comorbid conditions. Providers were especially concerned that seniors who don’t go out much may incorrectly believe they don’t need to get vaccinated.

Severity of the flu. While some of the public participants didn’t see influenza as a deadly disease, others did.

Providers recognized that some members of the public regarded the flu as more of a nuisance then a serious illness and emphasized the need to raise awareness.

Vaccine efficacy. The opinions that the public participants had about the effectiveness of the influenza vaccine varied. Some said it was effective; others believed the vaccine was not effective because they knew they could still get the flu even after being vaccinated. “I know people who got the flu shot and they still got the flu.” Many believed that home remedies, cleanliness, and staying away from others were more effective means of prevention.

Providers believed the vaccine is effective if it matches the correct strain of flu virus. They noted that even if a patient is vaccinated against the wrong strain, flu symptoms will be milder. Providers recognized patient concerns about vaccine efficacy, but none mentioned that it was important to encourage vaccination along with cleanliness and avoidance measures.

Safety and side effects. The most common concern about vaccine safety that the public participants discussed was that the influenza vaccine causes flu illness. Several respondents substantiated this concern with their own experiences, or those of others, where shortly after vaccination a flulike illness resulted. “It gave them the flu. And they were sick for 2 or 3 weeks.”

Another common safety concern was that the vaccine would interact with prescription medications for chronic illnesses. Many public participants also noted that health care providers neglected to discuss the matter. “[M]y reason for not taking the flu shot is because I’m on other medicine and I do have some concerns.… What are you putting in my body?” Some were also concerned about the safety of vaccine components and receiving a tainted vaccine. These participants also mentioned their fear of an allergic reaction to the flu vaccine.

Overall, providers did not have concerns about vaccine safety; however, they understood patients were afraid the influenza vaccine would give them the flu. They felt they should inform patients that full protection from the influenza vaccine takes up to 2 weeks. Providers were also aware that many African Americans who are 50 years of age and older distrust the medical system.

Main reason for vaccination: Doctor’s advice

All of the African American adults agreed that physicians and other health care professionals were important sources of vaccine information. Though initially ambivalent, a majority also reported receiving the vaccine primarily because of recommendations from their doctor. Some noted that providers can do more to encourage vaccination when patients express concerns. “I would have taken it if he had said, ‘I think you should take it.’” The public participants also got vaccine information from family, friends, broadcast media, and print material. They noted that they wanted to see the following in flu vaccination information: the pros and cons of the vaccine, efficacy of the vaccine, how the vaccine reduces flu severity, vaccine safety, and history and background of the vaccine.

Providers noted the importance of: promoting informed decision making, one-on-one communication, using a matriarchal figure to promote vaccination and using the media to promote vaccination. “You would have to have a mother, or a grandmother, or aunt figure because that is usually who is responsible, who takes care of the family.” Providers also pointed out that patients often voice concerns about not having enough information to make informed decisions, and that patients rely on convenience and doctors’ recommendations when deciding about vaccination.

Discussion

Patients may distrust the system, but they trust their doctor

Our study’s findings from both providers and older African Americans suggest that physicians are the most influential source of information when patients are deciding about flu vaccination. This is true despite the fact that the public makes no secret of its distrust of the medical system and the safety of the vaccine. The African American participants also suggested that physicians do not adequately address patient concerns through discussion and the information they provide.

Providers were concerned that many people do not believe the flu is a severe illness or that they are susceptible. Although the African Americans in the study recognized some high-risk groups, they tended not to consider themselves part of any of those groups.

Doctors didn’t realize that patients fear drug interactions

One of the interesting findings of this study was that armed with the knowledge that the flu shot does not guarantee flu prevention, African American adults were willing to forego the vaccine. This absence of a guarantee also reinforced their beliefs that other prevention methods, such as handwashing and avoidance, are more effective.

Not surprisingly, of course, was the finding that patients continue to avoid the flu shot for fear of getting the flu.^2,6 What was a bit surprising was that providers did not recognize that fear of medication reactions (drug interactions and allergic reactions) was also a barrier to flu vaccination. Providers also missed out on an educational opportunity, since many of the African American participants wanted to discuss the possibility of interactions with them.

Limitations of the study

External validity is limited because the findings cannot be generalized to every African American population in the US. The participants made up a non-random convenience sample of older African Americans in a Midwestern city, although the community-based recruitment strategy succeeded in reaching members of a lower income urban population. This study included only those who were ambivalent about the vaccine and who were open to both the pros and cons of vaccination. Project staff minimized possible interviewer bias by using experienced moderators, ensuring the consistent use of moderator guides, and using consensus coding procedures.

Funding

This research was funded by grant #6465 from the National Immunization Program at the Centers for Disease Control and Prevention, via Special Interest Project 11, to the Prevention research Center at the Saint louis University School of Public Health.

Acknowledgments

We thank Katie Duggan at the School of Public Health and edith Gary and Pascale Wortley at the CDC for their support. This article is dedicated to the memory of Joe D. Wray, MD, who suggested it.

Correspondence
Ricardo J. Wray, PhD, Community Health, Saint Louis University School of Public Health, 3545 Lafayette Avenue, St. Louis, MO 63104; wray@slu.edu

Why are older African Americans less likely than whites to get a flu vaccination? Despite the existence of an effective flu vaccine, usage rates still remain low: 66% on average,¹ and are even lower among minority groups.² One survey found only 47% of older African Americans were vaccinated against the flu.³

These disparities are not easily explained. Even when controlling for increased risk, age distribution, perceived health status, family size and marital status, poverty level, education, and access to medical care and health insurance, African Americans are still less likely than whites to get vaccinated.^3,4

In light of this disparity, we explored barriers to flu vaccination among this population, including concerns over safety and adverse events, and the role that health care providers can play in overcoming these issues.

What are the barriers?

Structural factors, such as having access to a location that provides vaccinations,⁵ and social factors, such as believing that others support vaccination,⁶ increase vaccination rates. Vaccination increases with age^5,7 and is more likely in those who have previously been vaccinated.⁸

People who believe they aren’t susceptible to a disease are less likely to get vaccinated.^9,10 Fear of side effects is also a significant barrier.^2,6,7,11,12 Fear of illness from the vaccine is often cited by both white and African American Medicare patients, older adults, and most notably by older African Americans.^2,5,12 African Americans and other minority groups are less likely to accept the vaccine as safe.¹³

A physician’s recommendation to get the flu vaccine appears to motivate patients in risk groups to get vaccinated,^6,7,12 and is a significant determinant of vaccine acceptance in surveys of older Americans, high-risk patients, and older African Americans in particular.^6,12

Methods

Focus groups and interviews

The study team conducted qualitative formative research with 2 pertinent audiences: African American adults (hereafter, “public participants”) and clinicians who administer the flu vaccine (hereafter, “providers”). The Saint Louis University Institutional Review Board approved the research.

We recruited older African American public participants through local community contacts, and screened to identify those ambivalent about getting a flu shot. A trained moderator then conducted focus groups and interviews with those who were ambivalent about the flu shot to assess their knowledge, beliefs, norms, and intentions related to vaccination.

We also recruited vaccine providers by calling local hospitals, doctors’ offices, health departments, and clinics that serve African Americans. To identify providers’ perceptions of patients’ concerns, we carried out interviews at the clinicians’ place of business or in a university conference room; a focus group was conducted at a community clinic.

All focus groups and interviews were audiotaped and transcribed. In pairs, research team members (all authors except for HJ) coded each of the transcripts independently, reviewed and discussed their codes, and then came to agreement on the final codes. Coded transcripts were entered into Atlas.ti (Atlas.ti GmbH, Berlin, Germany), a qualitative data analysis software program, and were analyzed with summary reports drafted for each focus group and interview. Findings were then synthesized across groups and interviews, and across segments.

Results

Description of participants

Four focus groups (N=35) and 8 in-depth interviews were conducted with the public participants—African American adults ages 50 and over. As shown in the TABLE, most participants were female, had children, had less than a college degree, and earned less than $30,000 a year. Widows and widowers made up the largest percentage of participants.

One focus group (N=9) and 5 interviews were conducted with vaccine providers—professionals working in clinics that offered the flu vaccine and served African Americans—for a total sample of 14. As the providers were sampled according to their professional affiliation, demographics were not systematically recorded. These informants were nurses, physicians’ assistants, and vaccination program administrators.

TABLE
4 focus groups and 8 interviews included 43 public participants*

How patients and clinicians see things differently

Susceptibility. Most public participants were familiar with only a few of the high-risk groups recommended for vaccination. Many participants said that children needed the flu shot, but did not state that those in their 50s were also recommended to get the flu shot. Those never stricken with the flu didn’t consider themselves susceptible. “Because I have never gotten it before…. I get a lot of other things but I don’t get colds and things like that.”

The health care providers identified senior citizens as a high-risk group for influenza because of the high occurrence of comorbid conditions. Providers were especially concerned that seniors who don’t go out much may incorrectly believe they don’t need to get vaccinated.

Severity of the flu. While some of the public participants didn’t see influenza as a deadly disease, others did.

Providers recognized that some members of the public regarded the flu as more of a nuisance then a serious illness and emphasized the need to raise awareness.

Vaccine efficacy. The opinions that the public participants had about the effectiveness of the influenza vaccine varied. Some said it was effective; others believed the vaccine was not effective because they knew they could still get the flu even after being vaccinated. “I know people who got the flu shot and they still got the flu.” Many believed that home remedies, cleanliness, and staying away from others were more effective means of prevention.

Providers believed the vaccine is effective if it matches the correct strain of flu virus. They noted that even if a patient is vaccinated against the wrong strain, flu symptoms will be milder. Providers recognized patient concerns about vaccine efficacy, but none mentioned that it was important to encourage vaccination along with cleanliness and avoidance measures.

Safety and side effects. The most common concern about vaccine safety that the public participants discussed was that the influenza vaccine causes flu illness. Several respondents substantiated this concern with their own experiences, or those of others, where shortly after vaccination a flulike illness resulted. “It gave them the flu. And they were sick for 2 or 3 weeks.”

Another common safety concern was that the vaccine would interact with prescription medications for chronic illnesses. Many public participants also noted that health care providers neglected to discuss the matter. “[M]y reason for not taking the flu shot is because I’m on other medicine and I do have some concerns.… What are you putting in my body?” Some were also concerned about the safety of vaccine components and receiving a tainted vaccine. These participants also mentioned their fear of an allergic reaction to the flu vaccine.

Overall, providers did not have concerns about vaccine safety; however, they understood patients were afraid the influenza vaccine would give them the flu. They felt they should inform patients that full protection from the influenza vaccine takes up to 2 weeks. Providers were also aware that many African Americans who are 50 years of age and older distrust the medical system.

Main reason for vaccination: Doctor’s advice

All of the African American adults agreed that physicians and other health care professionals were important sources of vaccine information. Though initially ambivalent, a majority also reported receiving the vaccine primarily because of recommendations from their doctor. Some noted that providers can do more to encourage vaccination when patients express concerns. “I would have taken it if he had said, ‘I think you should take it.’” The public participants also got vaccine information from family, friends, broadcast media, and print material. They noted that they wanted to see the following in flu vaccination information: the pros and cons of the vaccine, efficacy of the vaccine, how the vaccine reduces flu severity, vaccine safety, and history and background of the vaccine.

Providers noted the importance of: promoting informed decision making, one-on-one communication, using a matriarchal figure to promote vaccination and using the media to promote vaccination. “You would have to have a mother, or a grandmother, or aunt figure because that is usually who is responsible, who takes care of the family.” Providers also pointed out that patients often voice concerns about not having enough information to make informed decisions, and that patients rely on convenience and doctors’ recommendations when deciding about vaccination.

Discussion

Patients may distrust the system, but they trust their doctor

Our study’s findings from both providers and older African Americans suggest that physicians are the most influential source of information when patients are deciding about flu vaccination. This is true despite the fact that the public makes no secret of its distrust of the medical system and the safety of the vaccine. The African American participants also suggested that physicians do not adequately address patient concerns through discussion and the information they provide.

Providers were concerned that many people do not believe the flu is a severe illness or that they are susceptible. Although the African Americans in the study recognized some high-risk groups, they tended not to consider themselves part of any of those groups.

Doctors didn’t realize that patients fear drug interactions

One of the interesting findings of this study was that armed with the knowledge that the flu shot does not guarantee flu prevention, African American adults were willing to forego the vaccine. This absence of a guarantee also reinforced their beliefs that other prevention methods, such as handwashing and avoidance, are more effective.

Not surprisingly, of course, was the finding that patients continue to avoid the flu shot for fear of getting the flu.^2,6 What was a bit surprising was that providers did not recognize that fear of medication reactions (drug interactions and allergic reactions) was also a barrier to flu vaccination. Providers also missed out on an educational opportunity, since many of the African American participants wanted to discuss the possibility of interactions with them.

Limitations of the study

External validity is limited because the findings cannot be generalized to every African American population in the US. The participants made up a non-random convenience sample of older African Americans in a Midwestern city, although the community-based recruitment strategy succeeded in reaching members of a lower income urban population. This study included only those who were ambivalent about the vaccine and who were open to both the pros and cons of vaccination. Project staff minimized possible interviewer bias by using experienced moderators, ensuring the consistent use of moderator guides, and using consensus coding procedures.

Funding

This research was funded by grant #6465 from the National Immunization Program at the Centers for Disease Control and Prevention, via Special Interest Project 11, to the Prevention research Center at the Saint louis University School of Public Health.

Acknowledgments

We thank Katie Duggan at the School of Public Health and edith Gary and Pascale Wortley at the CDC for their support. This article is dedicated to the memory of Joe D. Wray, MD, who suggested it.

Correspondence
Ricardo J. Wray, PhD, Community Health, Saint Louis University School of Public Health, 3545 Lafayette Avenue, St. Louis, MO 63104; wray@slu.edu