The Effects of Nonclinician Guidance on Effectiveness and Process Outcomes in Digital Mental Health Interventions: Systematic Review and Meta-analysis

BACKGROUND: Digital mental health interventions are increasingly prevalent in the current context of rapidly evolving technology, and research indicates that they yield effectiveness outcomes comparable to in-person treatment. Integrating professionals (ie, psychologists and physicians) into digital mental health interventions has become common, and the inclusion of guidance within programs can increase adherence to interventions. However, employing professionals to enhance mental health programs may undermine the scalability of digital interventions. Therefore, delegating guidance tasks to paraprofessionals (peer supporters, technicians, lay counsellors, or other nonclinicians) can help reduce costs and increase accessibility. OBJECTIVE: This systematic review and meta-analysis evaluates the effectiveness, adherence, and other process outcomes of nonclinician-guided digital mental health interventions. METHODS: Four databases (MEDLINE, Embase, CINAHL, and PsycINFO) were searched for randomized controlled trials published between 2010 and 2020 examining digital mental health interventions. Three journals that focus on digital intervention were hand searched; gray literature was searched using ProQuest and the Cochrane Central Register of Control Trials (CENTRAL). Two researchers independently assessed risk of bias using the Cochrane risk-of-bias tool version 2. Data were collected on effectiveness, adherence, and other process outcomes, and meta-analyses were conducted for effectiveness and adherence outcomes. Nonclinician-guided interventions were compared with treatment as usual, clinician-guided interventions, and unguided interventions. RESULTS: Thirteen studies qualified for inclusion. Nonclinician-guided interventions yielded higher posttreatment effectiveness outcomes when compared to conditions involving control programs (eg, online psychoeducation and monitored attention control) or wait-list controls (k=7, Hedges g=–0.73; 95% CI –1.08 to –0.38). There were also significant differences between nonclinician-guided interventions and unguided interventions (k=6, Hedges g=–0.17; 95% CI –0.23 to –0.11). In addition, nonclinician-guided interventions did not differ in effectiveness from clinician-guided interventions (k=3, Hedges g=0.08; 95% CI –0.01 to 0.17). These results suggest that guided digital mental health interventions are helpful to improve mental health outcomes regardless of the qualifications of the individual performing the intervention, and that the presence of a nonclinician guide improves effectiveness outcomes compared to having no guide. Nonclinician-guided interventions did not yield significantly different adherence outcomes when compared with unguided interventions (k=3, odds ratio 1.58; 95% CI 0.51 to 4.92), although a general trend of improved adherence was observed within nonclinician-guided interventions. CONCLUSIONS: Integrating paraprofessionals and nonclinicians appears to improve the outcomes of digital mental health interventions, and may also enhance adherence outcomes (though this trend was nonsignificant). Further research should focus on the specific types of tasks these paraprofessionals can successfully provide (ie, psychosocial support, therapeutic alliance, and technical augmentation) and their associated outcomes. TRIAL REGISTRATION: PROSPERO International Prospective Register of Systematic Reviews CRD42020191226; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=191226