JMIR Publications

ABSTRACT

Background:

Digital technologies have enormous potential for supporting the general population to monitor their mental health across time via self-report. The core challenge in achieving this goal is to develop a rational framework for the measurement of relevant mental health constructs. The present project introduces such a framework, and links it to existing literature via a systematic review of the plethora of existing measures.

Objective:

This systematic review represents the conceptualisation stage in developing a novel integrated digital instrument for repeated measurements. We interrogated the literature as the first step towards optimal measurement of the multi-faceted mental health concept in the context of digital repeated measurement. This review identified (1) digital self-report instruments administered repeatedly to measure the mental health of the general adult population; (2) their structure and format; (3) their psychometric properties; (4) their usage frequencies in empirical studies; (5) the constructs these instruments have been used to measure in empirical studies in comparison to the constructs they were developed to measure.

Methods:

Five major electronic databases were searched. Studies administering digital mental health instruments (in English) repeatedly to community dwellers in the general adult population were eligible. Preliminary screening for eligible studies was conducted by one reviewer, before two reviewers independently screened the eligible articles (in full-text). Data were extracted from the final included articles and later grouped by instruments. The risk of bias of these instruments was assessed through their psychometric properties. This review was synthesised using a narrative approach and guided by the PRISMA checklist.

Results:

The preliminary screening of 95,849 articles identified 8,460 eligible articles, amongst which 1,000 articles (indicative of the instrument’s usage) were randomly selected for full-text screening in four iterations. A total of 223 articles were included in this process. The top 30 most used instruments measured constructs related to mental illness. The format and structure of these instruments conformed to the recognised criteria for repeated measurement. They were administered through different modalities (online, mobile apps) and frequencies (e.g., on-demand, daily, monthly, yearly). Most instruments exhibited excellent content validity and internal consistency but the evidence for other psychometric properties was either scarce or indeterminate. The Patient Health Questionnaire (PHQ-9) and Generalised Anxiety Disorder (GAD-7) were the most used instruments with consistent yearly usage between 2010 and 2021. No significant discrepancies were identified between the constructs measured by empirical studies and the target constructs described in the instruments’ original publications.

Conclusions:

This review highlights that digital instruments developed for general population use predominantly measure mental illness, despite growing interest in positive mental health and well-being. The findings are consistent with past reviews on usage patterns and construct coverage. The wide range of evidence for instruments’ psychometric properties was a barrier to synthesis and comparison of quality. The generalisability of this review is limited by our restricted focus on English language instruments, and the pragmatic strategy employed to select articles for full-text screening. Implications of this review for development of a novel adaptive digital self-report measure are discussed. Clinical Trial: CRD42022306547 (PROSPERO)