Personalized Smartphone Messaging for Secondary Prevention After Percutaneous Coronary Intervention: Randomized Controlled Trial

Background: Cardiac rehabilitation improves outcomes after percutaneous coronary intervention (PCI), but participation remains suboptimal due to barriers such as distance, cost, and lack of referral. Mobile health interventions may enhance accessibility and engagement. Objective: This study evaluated the effectiveness and user acceptability of a smartphone messaging app (AnSim) in improving cardiovascular risk factors among patients who recently underwent PCI. Methods: A 2-arm randomized controlled trial was conducted at 2 Korean hospitals; 120 patients who had undergone PCI within the prior month were randomized (1:1) to receive AnSim plus usual care (intervention) or usual care alone (control). The intervention comprised personalized messages tailored by behavioral stage and risk profile, delivered 6 times/week for 6 months. Outcomes were assessed at baseline, 6 months, and 9 months. The primary outcome was change in blood pressure (BP) at 6 and 9 months. Secondary outcomes were lipid profiles, hemoglobin A, BMI, smoking status, physical activity, and achievement of guideline-recommended risk-factor targets. User acceptability was assessed by questionnaire. Results: Overall, 115 (95.8%; intervention: n=58, control: n=57) participants completed follow-up at 6 months and 110 (91.7%; intervention: n=54, control: n=56) at 9 months. At 6 months, systolic BP was 124.9 (SD 16.7) mmHg in the intervention group and 124.8 (SD 13.1) mmHg in the control group (between-group difference 0.1 mmHg, 95% CI −5.4 to 5.6), and diastolic BP was 80.2 (SD 11.4) mmHg vs 78.4 (SD 10.6) mmHg (difference 1.8 mmHg, 95% CI −2.3 to 5.9). At 9 months, systolic BP was 126 (SD 14.2) mmHg vs 128 (SD 15.4) mmHg (difference −2 mmHg, 95% CI −7.6 to 3.6), and diastolic BP was 80.8 (SD 13.2) mmHg vs 81.3 (SD 12.0) mmHg (difference −0.5 mmHg, 95% CI −5.3 to 4.3). No significant between-group differences were observed in secondary outcomes. In post hoc analyses, participants with above-median message reading were more likely to achieve ≥4 of 5 guideline-recommended risk-factor targets at 9 months (69% vs 19.2%; relative risk 3.59, 95% CI 1.57‐8.18; <.001) and logged more diary entries (mean difference 203.4 entries, 95% CI 79.3‐327.5; =.001). BP responders at 9 months read messages on more days than nonresponders (110.3 vs 83.3 d; mean difference 27 d, 95% CI 4.6‐49.3; =.02). Among survey respondents (54/60, 90%), 87% (47/54) found the messages helpful and 81.4% (44/54) wished to continue. Two events occurred in the control group (=.50). Conclusions: This trial found no between-group improvement in BP with low-intensity personalized messaging compared to an attention-matched diary app, but exploratory analyses suggest engagement may be a key determinant of potential benefit. Unlike comprehensive app-based cardiac rehabilitation programs, AnSim provides scalable behavioral reinforcement that can be integrated into post-PCI follow-up with minimal burden. These findings support future pragmatic trials optimizing engagement strategies and evaluating longer-term clinical and implementation outcomes. Trial Registration: Clinical Research Information Service (CRIS) KCT0002361; https://cris.nih.go.kr/cris/search/detailSearch.do?seq=7230