Validation and selection of criteria for evaluating apps for managing depression: a Delphi study

Despite a lack of evidence relating to effectiveness and safety, the use of apps in the field of mental health is increasing due to their ease of use and accessibility. The aim of the EvalDepApps project is to develop and validate an assessment tool for evaluating depression management apps based on scientific evidence, expert judgement and end-user needs.

The purpose of this study was to determine the most relevant criteria for evaluating apps intended to manage depression through consensus-based assessment.

A total of 43 individuals were invited to participate in an online modified Delphi study of 51 criteria identified from the literature. In Round 1, participants rated criteria according to relevance and three levels of consensus were defined: high level when ≥80% of respondents scored the criterion at 5 or 6; medium when 60%-79% of respondents scored the criterion at 5 or 6 and low when <60% of respondents scored the criterion at 5 or 6. In Round 2, participants were asked to re-rate criteria that had achieved the medium level of consensus in Round 1 under the same parameters. Comments by participants were collected and analysed. The final assessment list consisted of those criteria that had attained the maximum consensus in either round.

The response rate was 59.0% (26/43) in Round 1 and 53.4% (23/43) in Round 2. In Round 1, 24 criteria (47.1%) attained the maximum level of consensus, 20 (39.2%) the medium level and 7 (13.7%) the lowest level. In Round 2, 4 out of 20 criteria (20.0%) attained the maximum consensus. Participant comments reinforced the relevance of the selected criteria. The final list consisted of 28 criteria, the majority relating to Safety and Privacy and Clinical Effectiveness (25.0% each), followed by Usability and Functionality (17.9%).

By prioritising criteria relating to data safety and clinical effectiveness, participants in this study emphasised that the assessment of apps for depression management must take both these aspects into full consideration. Despite some limitations of the study (eg, lack of participant sociodemographic data and its implications for generalisation, not face-to-face inter-round), the results of this study will enable the EvalDepApps project to develop an assessment tool for depression management apps that incorporates the most relevant criteria.

EvalDepApps will support healthcare professionals and users in identifying safe, effective and user-friendly depression management apps.