Digital Health Interventions Providing Behavioral Assessment and Goal Prioritization Support: Scoping Review

Affordability of health care systems depends on populations' engagement in preventive health behavior and appropriate self-management of long-term conditions. Digital health interventions (DHIs) could facilitate this by prompting and supporting individual health behavior change. Behavior change is often undermined by suboptimal prioritization of goals. Therefore, DHIs aiming to promote behavior change should help users identify behavior patterns that need changing and scaffold goal prioritization.

This scoping review explores the extent to which DHIs are supporting users to identify and prioritize goals relevant to managing and improving health.

The review followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines. Web of Science (Core Collection), Scopus, Ovid (Embase, MEDLINE, PsycINFO, and Global Health), and EBSCOHOST (Academic Search Complete and CINAHL Complete) were searched for literature on the development and evaluation of digital interventions that (1) assess users' current health or health-related behaviors and (2) offer support on prioritization of health-related goals.

Fifty-six papers were included. These identified 19 unique DHIs. Targeted populations included the general population (n=10), those at risk of or diagnosed with cardiovascular disease (n=4), those at risk of or diagnosed with diabetes (n=2), those diagnosed with cancer (n=2), or those diagnosed with HIV (n=1). One DHI targeted preconception among African American women. All DHIs targeted physical activity and most (n=17) targeted diet and smoking, closely followed by alcohol use (n=15) and mental health (n=13). Social wellbeing (n=5), sleep (n=4), and pain (n=1) were less commonly included. All 19 DHIs included a health risk assessment with feedback identifying health domains in need of improvement, but only four asked users to select a prioritized change goal. Outcome evaluations were conducted for most (n=14), with nine DHIs evaluated using at least one randomized control trial (RCT). Almost half of all DHIs (n=9) reported at least one evaluation of behavioral outcomes, mostly employing RCTs (7/9). Six of 19 reported at least one evaluation of psychological health outcomes, again mostly employing RCTs (5/6). Among the seven DHIs for which behavioral outcomes were evaluated using a RCT, effects were mixed, with only one DHI showing significant effects across all assessed behavioral outcomes. Three found significant effects for some, but not all, outcomes or timepoints, and three found no significant effects.

Although all 19 DHIs provided some advice about which health-related goals to prioritize, most did not actively prompt users to set such priorities. DHIs showing the most promise in terms of health behavior change were those that explicitly promoted users to prioritize changing specified health behaviors. This review highlights how DHIs could provide greater behavior change support and provides the basis for designing more effective DHIs.