Teleneuropsychology in the memory clinic setting: evidence for reliability and usability of videoconference‐based neuropsychological testing

Abstract

AbstractBackgroundRemote neuropsychological assessment through video teleconferencing (VTC) can improve accessibility to dementia care, even beyond the COVID‐19 pandemic. Evidence for the usability and reliability of VTC is still limited in a memory clinic population, particularly for tests assessing non‐memory and non‐verbal functions. This study aims to investigate stability of performances on neuropsychological tests across in‐person and VTC modality, as well as VTC user experiences, in memory clinic patients.MethodPatients were included at the Alzheimer Center Amsterdam between August 2020 and February 2021. Patients underwent in‐person neuropsychological assessment, and subsequently through VTC within 4 months. Stability of performances on all tests across assessments was investigated with intraclass correlations coefficients/percentage agreement and independent samples t‐tests/Wilcoxon signed‐rank tests. Rank order stability was investigated with Pearson or Spearman correlation/Chi‐square tests. For each test, we converted standardized (T‐)scores into seven performance classes (see Figure 1) and described the proportion of patients showing a clinically relevant difference in performance class. Furthermore, we described stability of patients’ overall assessment performance (no impairment, single‐domain impairment, multi‐domain impairment). User friendliness and system usability were assessed with questionnaires by patients and neuropsychologists. Scores were set against previously defined acceptability cut‐offs, and compared across diagnostic groups (subjective cognitive decline [SCD] vs. MCI/dementia) with independent samples t‐tests.ResultThirty‐one patients (62±6.7 years, 55% male, 58% SCD, 42% MCI/dementia) completed VTC assessment. Intraclass correlation coefficients ranged from moderate (0.5; TMT index) to excellent (>0.9; RAVLT, LDST), see Table 1.Significant differences in mean performances were observed for Digit span and Stroop test (SCD group only). For all tests, the majority of patients showed no clinically relevant difference in performance (Figure 1, 61%‐100% stable). Nineteen patients (61%) had the same overall assessment performance across modalities (Figure 2). User friendliness (mean total scores patients/psychologists:164/143) and usability (mean item scores patients/psychologists:3.0/3.0) reached acceptability and did not differ between diagnostic groups.ConclusionVTC assessment may serve as an accessible and user friendly method to test cognitive functioning in the memory clinic, with largely similar reliability coefficients as in‐person assessments. Implications of VTC assessment for the evaluation of impairment severity, and possibly diagnosis, should be further investigated.