Maximizing fidelity of neuropsychology assessments in fully remote studies

Remote, interdisciplinary, observational clinical studies and clinical trials are becoming increasingly common, yet they pose challenges for maintaining data fidelity-the accuracy, completeness, and consistency of data. This study aimed to (1) evaluate the data fidelity of a remote videoconference-administered neuropsychology protocol within the context of an observational clinical trial; (2) develop three methodological roadmaps to support data fidelity for remote neuropsychological research; and (3) characterize the sample and their performance. One hundred forty eight participants (62% female; median age = 67, mean education = 15.1 years) completed the protocol. Data were analyzed using descriptive statistics. On average, 5.59 queries (i.e., data quality flags) per participant were generated during the data monitoring phase, decreasing to 0.30 during the cleaning and curation pipeline phase. Overall, missing data were minimal (0.34%). Virtual environment factors (e.g., internet connectivity and distractions) had minimal impact on data quality; only 8 participants (5.4%) had 1-2 tasks affected resulting in missing data. The fidelity of the remote videoconference-administered protocol was comparable to that of in-person assessments. These findings suggest that remote videoconference-administered neuropsychological assessment can yield high-fidelity data when supported by rigorous quality assurance and quality control procedures. The proposed roadmaps provide practical guidance for future remote neuropsychological research and support the broader adoption of standardized telehealth methodologies in the study of neurodegenerative diseases.