Women's Representation in Randomized Controlled Trials Evaluating FDA supervised Medical Devices - A Systematic Review
Description
Underrepresentation of women in clinical trials has been previously reported, even when disease prevalence is similar between sexes. In recent years, guidance for clinical trials of medical devices addressing women's representation was issued. We sought to assess current rate of women representation in clinical trials of FDA type III approved devices. Type III devices are categorized as high-risk devices, usually implanted, and have the potential to cause a severe degree of harm. Methods and Materials Systemic review of Randomized controlled trials (RCTs) evaluating type III FDA devices in various medical fields. The primary outcome was the rate of women randomized (number of women randomized /total number of randomized patients). We also examined the participation-to-prevalence ratio (PPR), which comprises the percentage of participating women/prevalence of device use in women in the general population. PPR<0.8 implies under-representation. Results 195 RCTs published between 2016 and 2022 were included. The majority of trials (67%) were conducted in cardio-vascular devices, with stents being the most prevalent (38%). The percentage of women randomized ranged between 7 to 100%, with a median of 33 (IQR 24-48%). Only 31 trials (23%) included ≥ 50% women (13% of cardiovascular trials). Cardiovascular trials included a significantly lower percentage of women compared to other fields (median 29%), with stents being the device with the lowest percentage (median 27%). The calculated PPR for stents, valves, and orthopedic prostheses was below 0.8 and 1 for pacemakers. Conclusion Despite many efforts to improve the representation of women in clinical trials, women are still under-represented in device evaluating trials. This raises concerns regarding the generalizability of device-related safety and efficacy outcomes to the women population.
Files
Steps to reproduce
Screening for eligible studies and search terms are included in supplements, we used Rayyan software to include appropriate studies, according to the PRISMA guidelines. The participation-to-prevalence ratio (PPR) for devices was calculated as the percentage of women participating / women with the intervention in the general population (See supplement). Categorical data were compared using the Chi-square test, and continuous data was compared using the Mann-Whitney or Kruskal-Wallis tests.