Manufacturers need to increase diversity of subpopulations when designing clinical trials to obtain an unbiased estimate of treatment effect, the FDA says.
New draft guidance assists sponsors in developing strategies for evaluating and reporting age, race and ethnicity data on device safety, effectiveness, benefit-risk profile, study design and device labeling.
The guidance shows sponsors what demographic variables should be considered in study designs. It covers collection and consideration of subpopulation data during clinical study design, outlining how to interpret subgroup data, pointers on how to enroll diverse populations and to maintain clinical follow-up.
To better evaluate clinically meaningful differences across patient subpopulations, sponsors need to enroll patients that are representative of the intended-use population of the investigational device, the guidance says.
Barriers to Diversity
The guidance lists several barriers to diversity in trials, including inclusion/exclusion criteria that unintentionally exclude different age, race, or ethnic groups.
Potential barriers also occur during the enrollment process when researchers may run into language, cultural and health literacy gaps between investigators and patients.
Additional barriers can be caused by pressure on trial investigators to quickly recruit regardless of demographic characteristics, increased expense or trial logistic challenges, along with disproportionate dropout rates.
Certain devices or diseases may effect certain subpopulations more than others, the FDA acknowledges.
For example, some orthopedic devices would likely have different considerations for use in specific age groups. And, mortality rates of patients on dialysis have been shown to differ across race and ethnicity groups, the guidance notes.
Similary, older patients and pediatric patients often have co-morbidities, concomitant therapies, or development considerations that could interact with the investigational device effects and impact device performance.
FDA recommends sponsors identify specific information about a subpopulation and the disease associated with the intended use of the device.
Sponsors should also document any known clinically meaningful age, race, and ethnicity differences in disease course, outcomes, or benefit-risk profile to be included in 510(k) summary and labeling.
Unless the investigational device is intended for use in only one age, race or ethnic group, it is important that the variation in data across age, race, and ethnic groups be accounted for both in study design and analysis of results, the guidance states. If co-morbidities or other characteristics are found, sponsors should base their analysis on demographic subgroups.
Although analysis of specific subgroups with inadequate sample sizes can’t be generalized as safe and effective across populations when labeling the device, the data would be useful for considering the overall benefit-risk profile of the device, the FDA said.
More Trials Needed?
If analysis points to a lack of demographically diverse data, negating clinically meaningful differences in outcome, FDA may request additional clinical data and modify trial design to address pre- or post-market questions.
The FDA encourages sponsors to use decision trees when considering various age, race or specific statistical recommendations for different clinical study designs.
Sponsors should also describe how any clinically meaningful differences across subgroups may contribute to differences in benefit-risk profiles in certain subpopulations.
The guidance applies to 510(k), PMA applications, de novo requests, HDE applications, as well as post-approval and post-market surveillance studies.
When finalized, the guidance will extend policy in the FDA’s Evaluation of Sex-Specific Data and complement Collection of Race and Ethnicity Data in Clinical Trials Guidance.
Interested parties can comment on the guidance until Sept. 19.
The draft guidance can be found here: www.fdanews.com/06-20-16-ReportingMedicalDeviceClinicalStudies.pdf. — Joya Patel