Digital health technologies versus traditional methods for cardiovascular risk assessment in asymptomatic adults: a systematic review and network meta-analysis of diagnostic accuracy and clinical outcomes.

Wang, Jing; Li, Xinxiao; Luo, Chuqing; Ren, Qiuping · Int J Med Inform · 2026

meta_analysis · Level I

Where this comes from

Abstract

Cardiovascular disease (CVD) remains the leading cause of mortality worldwide, accounting for approximately 17.9 million deaths annually. Traditional risk assessment tools demonstrate limited discriminatory capacity. Digital health technologies, including AI algorithms, wearable devices, and smartphone applications, offer promising alternatives for improved CVD risk stratification. To evaluate the effectiveness of digital health technologies versus traditional tools in cardiovascular disease risk assessment through a systematic review and network meta-analysis. We conducted a PRISMA-NMA systematic review searching PubMed, Cochrane, and EMBASE through October 2025, comparing digital health technologies with traditional risk scores for CVD assessment in asymptomatic adults, measuring diagnostic accuracy and clinical outcomes via frequentist network meta-analysis. Thirty-two studies encompassing 11,496,837 participants met inclusion criteria. Wearable devices demonstrated the highest pooled AUC for arrhythmia detection (0.92; 95% CI, 0.81-1.00), followed by smartphone applications (0.93; 95% CI, 0.88-0.98) and AI-ML algorithms (0.83; 95% CI, 0.81-0.86). Traditional risk scores exhibited significantly lower discrimination (pooled AUC, 0.75; 95% CI, 0.72-0.78). For clinical outcomes, digital health interventions significantly reduced all-cause mortality (pooled RR, 0.62; 95% CI, 0.48-0.80; I<sup>2</sup> = 91.8%) and major adverse cardiovascular events (pooled RR, 0.70; 95% CI, 0.60-0.83; I<sup>2</sup> = 90.4%). Surface under the cumulative ranking curve analysis ranked wearable devices first for diagnostic accuracy (SUCRA, 92.6%) and AI-ML algorithms first for clinical outcomes (SUCRA, 72.8%). Digital health technologies, wearable devices, and AI-machine learning algorithms surpass traditional cardiovascular risk assessment methods, significantly enhancing diagnostic accuracy, risk stratification, and patient clinical outcomes.