Enrollment disparities in cancer clinical trials: A single institution experience.

Abstract

e18128 Background: The importance of population heterogeneity in clinical cancer research is universally recognized. This study aims to identify possible disparities in trial participation at UF Health Cancer Center at Orlando Health. Methods: Data on patients’ enrollment, during the period from 01/01/2016 to 12/31/2016, was obtained from our research C-base. We utilized descriptive statistics to characterize the main demographic variables. Pearson's Chi-squared and Z-tests were utilized to compare percentages of enrolled versus not enrolled patients by ethnicity, race and gender. Results: A total of 3704 cases were established as new patients during the study period, of which 369 (10%) were enrolled in cancer research studies (Table). Our analysis showed statistically significant differences in enrollment by racial category (χ2= 95.09, df = 3, p < 0.0000). American Indian (n = 6) and Native Hawaiian (n = 4), were excluded from the analysis due to small sample sizes and no enrollments. In addition we tested with the Z-stat whether the enrollment in each group was different from the population average (p = 0.10). The Caucasian group did not differ from the average (z = 0.3944; p < 0.6933), whereas both the African American (z = -4.2286; p < 0.0001) and the Asian group (z = -2.3102; p < 0.0209) were under enrolled. The Others group was above the average (z = 8.4632; p < 0.0001). There were also differences in enrollment by gender and ethnicity. Females were under enrolled as compared to Males (z = -7.0716; p < 0.0001). The Hispanic group enrollment was significantly below the population average (z = -4.17; p < 0.0001). Conclusions: We found significant differences in the enrollment of minorities in cancer trials in our institution. African Americans, Asians, Hispanics and Females were all under represented. The group labeled as Others requires further attention. [Table: see text]