October 10, 2019 – Differences exist between acute heart failure patients seen in real-world practice and clinical trials, and these differences may be largest among black patients and women.
Black patients and women are frequently underrepresented in clinical trials. However, aside from trials generally including lower proportions of patients from these groups, new findings from the DCRI and co-authors suggest that black patients and women who are enrolled may not fully reflect black patients and women seen in clinical practice.
An analysis published yesterday in the Journal of the American College of Cardiology: Heart Failure compared outcomes of U.S. patients enrolled in a heart failure trial called ASCEND-HF with outcomes of those from a U.S. heart failure registry called Get With the Guidelines-Heart Failure who were eligible for ASCEND-HF. The study was led by former DCRI fellow Stephen Greene, MD, (pictured top), and the DCRI’s Robert Mentz, MD, (pictured bottom), also contributed.
The study results showed that trial-eligible patients in the registry had worse outcomes than individuals enrolled in the trial. This difference in outcomes was driven largely by differences in black and female patients. For example, 29.4 percent of black patients in the registry experienced 30-day hospital readmission, while only 13.9 percent of black patients in the trial experienced this outcome. The same was true for women; 28.0 percent of women in the registry and 16.5 percent of women in the trial experienced 30-day readmission. The differences between registry patients and trial participants were much greater in these populations than the differences for white patients and men.
These findings suggest that unmeasured social and behavioral factors among patients and site investigators can significantly shape the profile of patients enrolled in heart failure clinical trials, independent of trial eligibility criteria. This selection bias may particularly influence black patients and women, and may result in enrollment of patients who are not necessarily typical of the black patients and women seen in real-world practice.
“Even among patients who meet all the trial eligibility criteria and have otherwise similar disease characteristics, there may be major differences between those who do and do not get enrolled,” Greene said. “Our study suggests that these differences may be greatest among black patients and women.”
“Because heart failure trials often include fewer black patients and women, this can lead to questions of whether trial results apply to these groups,” he continued. “However, our study adds another layer to this. Not only may there be fewer black patients and women in a study, but the ones who are enrolled may not be reflective of the typical black patients and women we see in routine practice. In contrast, white patients and men in the trial may be a closer reflection of these groups seen in practice.”
“Although we already knew underrepresentation was a problem—both in heart failure trials and in trials studying other conditions—these findings suggest that representativeness is also an issue at play,” Mentz said. “To make sure heart failure clinical trials meet the needs of patients with this condition, we will need concentrated efforts to improve both enrollment of black patients and women, and representativeness.”