Effectiveness of Computer-Tailored Health Communication in Increasing Physical Activity in People With or at Risk of Long-Term Conditions: Systematic Review and Meta-Analysis

BACKGROUND: Regular physical activity (PA) is beneficial for enhancing and sustaining both physical and mental well-being as well as for the management of preexisting conditions. Computer-tailored health communication (CTHC) has been shown to be effective in increasing PA and many other health behavior changes in the general population. However, individuals with or at risk of long-term conditions face unique barriers that may limit the applicability of CTHC interventions to this population. Few studies have focused on this cohort, providing limited evidence for the effectiveness of CTHC in promoting PA. OBJECTIVE: This systematic review and meta-analysis aims to assess the effectiveness of CTHC in increasing PA in individuals with or at risk of long-term conditions. METHODS: A systematic review and meta-analysis were conducted to evaluate the effect of CTHC in increasing PA in people with or at risk of long-term conditions. Hedges g was used to calculate the mean effect size. The total effect size was pooled and weighted using inverse variance. When possible, potential moderator variables were synthesized, and their effectiveness was evaluated by subgroups analysis with Q test for between-group heterogeneity Q(b). Potential moderator variables included behavior change theories and models providing the fundamental logic for CTHC design, behavior change techniques and tailoring strategies to compose messages, and computer algorithms to achieve tailoring. Several methods were used to examine potential publication bias in the results, including the funnel plot, Egger test, Begg test, fail-safe N test, and trim-and-fill method. RESULTS: In total, 24 studies were included in the systematic review for qualitative analysis and 18 studies were included in the meta-analysis. Significant small to medium effect size values were found when comparing CTHC to general health information (Hedges g=0.16; P<.001) and to no information sent to participants (Hedges g=0.29; P<.001). Half of the included studies had a low to moderate risk of bias, and the remaining studies had a moderate to high risk of bias. Although the results of the meta-analysis indicated no evidence of publication bias, caution is required when drawing definitive conclusions due to the limited number of studies in each subgroup (N≤10). Message-tailoring strategies, implementation strategies, behavior change theories and models, and behavior change techniques were synthesized from the 24 studies. No strong evidence was found from subgroup analyses on the effectiveness of using particular behavior change theories and models or from using particular message-tailoring and implementation strategies. CONCLUSIONS: This study demonstrates that CTHC is effective in increasing PA for people with or at risk of long-term conditions, with significant small to medium effects compared with general health information or no information. Further studies are needed to guide design decisions for maximizing the effectiveness of CTHC.