A, Reporting of patients’ sex and racial groups over time. B, Percentage of female patients in the RCTs of AF, HF and ACS guidelines. C, Percentage of different racial groups in the AF, HF, and ACS guidelines. ACS indicates acute coronary syndromes; AF, atrial fibrillation; HF, heart failure; and RCTs, randomized controlled trials.
Sardar MR, Badri M, Prince CT, Seltzer J, Kowey PR. Underrepresentation of Women, Elderly Patients, and Racial Minorities in the Randomized Trials Used for Cardiovascular Guidelines. JAMA Intern Med. 2014;174(11):1868-1870. doi:10.1001/jamainternmed.2014.4758
The American College of Cardiology/American Heart Association (ACC/AHA) guidelines are frequently used to guide cardiovascular care. Randomized controlled trials (RCTs) are considered to be the highest level of evidence and are cited in these guidelines whenever available. However, RCTs can have limited external validity. To examine their generalizability, we examined the percentage of enrollment and reporting of sexes, races, geographic distribution, and age groups in the RCTs cited in the ACC/AHA guidelines for atrial fibrillation (AF),1 heart failure (HF),2 and unstable angina/non–ST-segment elevation myocardial infarction (acute coronary syndromes [ACS]).3
All guidelines’ references were screened for RCTs. Meta-analyses, review articles, guidelines, articles in a language other than English, abstracts, studies reporting follow-up results, or subgroup analyses of already included RCTs and other non-RCT references were excluded.
Data extracted from the RCTs included sample size, mean age, percentage of male and female patients, racial category when reported, trials’ publication year, whether it was a single or multicenter trial, and continents from which subjects were recruited. Elderly patients were defined as those 75 years or older because this is the cutoff used in ACC/AHA guidelines.3
Randomized controlled trials constituted the highest proportion of HF references (28% [230 of 814]) compared with AF (25% [229 of 910]) and ACS (20% [195 of 970]). The majority (72%) of the 653 RCTs from all 3 guidelines were multicenter trials. Mean sample size across all studies was 2047 patients. There was a continuous trend toward increase in study sample size over time.
Female representation was highest in AF RCTs (33%) followed by ACS (29%) and HF (29%) (Table), which is lower than that of US registries of AF (55%),4 ACS (42%),4 and HF (47%).4 Reporting of study patients’ sex has increased over time (Figure, A). After excluding trials with sex-specific intervention (therefore necessitating single-sex inclusion), female representation across all trials was 30%. This increased from 24% in the 1980s, to 28% in the 1990s, and to 31% for the years 2000 through 2009. Inclusion of women in HF and ACS clinical trials improved, while that in AF RCTs decreased steadily over time (Figure, B).
Less than a quarter (153 of 653 [23.4%]) of all RCTs reported patient race (36% of HF studies, 27% of ACS studies, and only 7% of AF studies). Reporting of race improved over time, particularly in the last decade, when approximately twice as many studies reported race as those in the prior decade (Figure, A).
When race was reported, the majority of patients were white. Over 86% of patients enrolled in AF and ACS trials were white compared with 73% in HF trials. Consequently, other races had a relatively low representation. Black patients constituted 19% of those studied in HF RCTs and 6% in AF and ACS RCTs (Figure, C). These percentages are close to their ratios in US registries of the same diseases (6%,4 21%,5 and 11%,6 respectively). Multicenter trials had better reporting of racial groups (29%) compared with single center trials (9%). Mean patients’ age was 75 years or older in only 2% (10 of 653) of all RCTs. Most trials (94%) enrolled patients from North America or Europe. Approximately 9% of RCTs included patients from Asia, South America, and Australia. Patient recruitment from Africa was limited to 4%, all from South Africa.
Our results suggest that women, elderly patients, and those of nonwhite racial backgrounds are underrepresented in the RCTs of the ACC/AHA guidelines for AF, HF, and ACS. These findings raise concerns about clinical trial enrollment and the applicability of the guidelines in these underrepresented populations. Investigators should enroll and report more women, elderly patients, and minorities in future trials to improve the evidence base for patient care as well as the professional society guidelines.
Corresponding Author: Marwan Badri, MBChB, MRCP, Lankenau Medical Center, 100 Lancaster Ave, Ste 558, Wynnewood, PA 19096 (email@example.com).
Published Online: September 29, 2014. doi:10.1001/jamainternmed.2014.4758.
Author Contributions: Dr Badri had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Sardar, Badri, Seltzer, Kowey.
Acquisition, analysis, or interpretation of data: Sardar, Badri, Prince.
Drafting of the manuscript: Sardar, Badri.
Critical revision of the manuscript for important intellectual content: Sardar, Prince, Seltzer, Kowey,
Statistical analysis: Sardar, Prince.
Administrative, technical, or material support: Badri, Seltzer.
Study supervision: Sardar, Badri, Kowey.
Conflict of Interest Disclosures: None of the authors are currently or have been involved in a guidelines-writing task force or recommendation and advisory board of such committees. Dr Kowey was a reviewer for the 2014 ACC/AHA guidelines for atrial fibrillation. No other disclosures are reported.