Reduction of Recruitment Costs in Preclinical AD Trials: Validation of Automatic Pre-Screening Algorithm for Brain Amyloidosis
This article describes a validated two-step process for recruiting asymptomatic amyloid-positive individuals into clinical trials. The process was tested using cohorts from three Alzheimer’s studies (ADNI-MCI, ADNI-CN and INSIGHT). During a pre-screening phase, researchers pre-selected a subset of individuals who were more likely to be amyloid positive, based on the automatic analysis of data acquired during routine clinical practice, before doing a confirmatory PET scan for these selected individuals only. Using this two-step method can increase the number of participants and reduce the number of PET scans, resulting in a decrease in overall recruitment costs. The researchers tested five different classification algorithms and found that the best results were obtained using solely cognitive, genetic and sociodemographic features; using MRI or longitudinal data slightly increased performance but incurred higher costs. The approach demonstrates how machine learning can be used effectively to optimize recruitment costs in clinical trials.
Ansart M, Epelbaum S, Gagliardi G, et al. Reduction of recruitment costs in preclinical AD trials: Validation of automatic pre-screening algorithm for brain amyloidosis. Statistical Methods in Medical Research 2020;29(1):151-164.