Compliance with mobile Ecological Momentary Assessment (mEMA) of health-related behaviours and psychological constructs in adults: Systematic review and meta-analysis (Preprint)

BACKGROUND Mobile Ecological Momentary Assessment (mEMA) permits real time capture of participant behaviours and perceptual experiences. Reporting of mEMA protocols and compliance has been identified as problematic within systematic reviews of child/youth and specific clinical populations of adults. OBJECTIVE In studies of nonclinical and clinical samples of adults, describe: 1) use of mEMA for behaviours and psychological constructs; 2) mEMA protocol and compliance reporting; and 3) associations between key components of mEMA protocols and compliance. METHODS Nine electronic databases were searched (2006 to 2017) for observational studies reporting compliance to mEMA for health-related data from adults (>18 years) in nonclinical and clinical settings. Screening and data extraction were undertaken by independent authors with discrepancies resolved by consensus. Narrative synthesis described participants, mEMA target, protocol and compliance. Random effects meta-analysis explored factors associated with cohort compliance (monitoring duration, daily prompt frequency/schedule, device type, training, incentives and burden score). Random effects ANOVA (P ≤ .05) assessed differences between nonclinical and clinical datasets. RESULTS Of 176 eligible studies, 105 (60%) reported compliance in 115 unique datasets (nonclinical n=69, clinical n=46). The commonest mEMA target was affect (nonclinical 19%, clinical 35%). Median mEMA protocol duration was 10 days (nonclinical 8, clinical 14). The majority of protocols used a single prompt type, random signal (nonclinical 77%, clinical 61%; median prompt frequency 5/d). The median number of items per prompt was similar for nonclinical and clinical datasets (10). Over half of datasets reported mEMA training (78%) and provision of participant incentives (63%). Less than half reported number of prompts delivered (22%), answered (48%), criterion for ‘valid’ mEMA data (33%) or response latency (35%). Meta-analysis (nonclinical n=42, clinical n=31) estimated overall compliance of 83.1% (95%CI 79.7, 89.6) with no difference between before or after data exclusions (P .83) or nonclinical and clinical datasets (P .07). For both nonclinical and clinical datasets, prompts/d and items/prompt were significantly associated with compliance. Compliance was lower with 4-5 prompts/d (nonclinical 77.4%, clinical 81.5%) and with increased items/prompts (clinical 9.5-26 items [71.1%], nonclinical ≥26 items [63%]). For clinical datasets, studies with a 7-day duration reported the lowest compliance (n=10, 77.9%; 95%CI 71.5, 83.6%) compared to studies with duration length <7 days (93.5%), 14 days (84.3%) and >14 days (87.1%). CONCLUSIONS In this 10-year sample of studies using mEMA of health-related behaviours and psychological constructs in adult nonclinical and clinical populations, mEMA was applied across contexts, health conditions and to collect a range of health-related data. There was inconsistent reporting of compliance and key features within protocols, which limited the ability to confidently identify components of mEMA schedules likely to have a specific impact on compliance. CLINICALTRIAL N/A