OBJECTIVES: The 9-item Patient Health Questionnaire (PHQ-9) is used to evaluate depressive symptoms and their severity. While validated in in-person settings, there is limited research on its performance when administered remotely. This study aims to identify meaningful differences in PHQ-9 scores by administration route from a large, U.S-based representative cohort of community-dwelling adults who enrolled in the Project Baseline Health Study (PBHS) from 2017 to 2019.

METHODS: This cross-sectional analysis included PBHS participants completing a PHQ-9 survey either in-person or remotely, via digital application, during their enrollment visit. A directed acyclic graph was used to identify study site as the minimally sufficient adjustment set. Inverse probability of treatment weights were employed to condition on study site. Median PHQ-9 scores were calculated; differences in score distributions were tested using a weighted Mann-Whitney-U test. History of major depressive disorder (MDD) was considered an effect modifier.

RESULTS: Of the 2176 participants who completed PHQ-9 during their PBHS enrollment visit, median (IQR) scores among paper completers (n=1428) and remote application users (n=748) were 2 (0-5) and 3 (1-6), respectively; score distributions differed significantly (U=-6.15, p<0.001]. Among 1813 participants without history of MDD, remote application users reported higher PHQ-9 scores (paper: 2 [0-4]; app: 3 [1-5]; p<0.001). There was no difference in scores among the 349 participants with history of MDD (paper: 6 [3-11]; app: 6 [4-10]; p=0.748).

CONCLUSIONS: The small difference in PHQ-9 scores by route of administration among participants without MDD is not clinically meaningful; no difference was found among participants with MDD. As the prominence of app-based clinical data collection increases, these results suggest remote screening for depression via PHQ-9 in a community-dwelling population is a pragmatic and viable approach.