LAUSR

Purpose: Evaluate impact of physician referral to health coaching on patient engagement and health risk reduction. Design: Four-year retrospective, observational cohort study with propensity-matched pair comparisons. Setting: Integrated delivery and finance system in Pittsburgh, Pennsylvania. Sample: 10 457 adult insured members referred to health coaching by their physician; 37 864 other members identified for health coaching through insurer-initiated outreach. Intervention: Practice-based, technology-supported workflow and process for physician prescribing of health coaching during regular office visit, with follow-up on patient’s progress and implementation supports. Measures: Patient engagement based on completion of pre-enrollment assessment, formal enrollment in health coaching, completion of required sessions, health risk levels, and number of health risks pre- and post-health coaching referral. Analysis: Difference-in-difference analysis to assess change in health risk levels and number of health risks pre- and post-health coaching and probability weighting to control for potential confounding between groups. Results: Members referred by a physician were significantly more likely to enroll in a health coaching program (21.0% vs 6.0%, P < .001) and complete the program requirements (8.5% vs 2.7%, P < .001) than when referred by insurer-initiated outreach; significant within group improvement in health risk levels from baseline (P < .001) was observed for both the groups. Conclusions: Patients are significantly more likely to engage in health coaching when a referral is made by a physician; engagement in health coaching significantly improves health risk levels.