Analysis of exclusion criteria in NIA‐funded Alzheimer’s disease and Alzheimer’s disease‐related dementias clinical trials

Abstract

AbstractBackgroundEligibility criteria are an integral component of clinical trials, ensuring participant safety and reducing factors that may confound results. Research suggests that unnecessarily restrictive exclusion criteria may disphroportionately exclude African Americans (AA) and Hispanics/Latinos (H/L), reduce generalizability, and prevent timely enrollment. This may be of particular concern for Alzheimer’s and related dementias (AD/ADRD), as AA and H/L are at greater risk than non‐Hispanic Whites for AD/ADRD yet underrepresented in clinical trials (CT). This abstract analyzes exclusion criteria proposed within active National Institute on Aging (NIA)‐funded AD/ADRD CTs (N=235) to document the most widely used criteria and to analyze the potential impact on AA and H/L CT participation.MethodCT exclusion criteria and study design were obtained from NIA grant folders and ClinicalTrials.gov. A codebook was developed using a sample of CTs to identify key exclusion categories (N=47 categories; intercoder reliability=92%) and used to code the 235 trials. A review of published literature was also conducted to identify which categories have documented ethno‐racial disparities (e.g., diabetes is more prevalent among H/L).ResultA median of 8 exclusion criteria were listed per CT (range=1‐41). The top categories/conditions were: 1) psychiatric (64% of trials, top ranked: depression, bipolar), 2) neurological (52%, top ranked: head/spinal injury, abnormal EEG/brain MRI), and 3) dementia/cognitive impairment (49%). Literature identifies almost half (47%) of all categories as disproportionately affecting AA and H/L, including psychological disorders and cardiovascular disease. Of 47 exclusion categories, AA have the potential to be disproportionately eliminated by 19 categories (40%) and H/L by 16 (34%). Additional analyses will be presented by trial population, phase, enrollment, and intervention type.ConclusionOur findings demonstrate commonly used exclusion criteria in active NIA‐funded AD/ADRD CTs. We found many criteria that disproportionately affect AA and H/L, which may play a role in their reduced enrollment in AD/ADRD trials. Similar to an effort in the oncology field (Kim, 2017), AD/ADRD stakeholders should undertake a concerted effort to jointly address the use of overly restrictive criteria, advocate for the use of clear rationale when establishing criteria, and implement more equitable, scientifically justified exclusion criteria for AD/ADRD CTs.