Physician acceptability and clinical impact of a decision support system for secondary cardiovascular disease prevention: findings from the C-QIP trial in India

Abstract

Clinical decision support systems (CDSS) may improve triage, guideline-directed medical therapy (GDMT), and risk factor management in fragmented, resource-constrained health systems. However, evidence on their acceptability and clinical impact among patients with established cardiovascular diseases (CVD) in low- and middle-income countries (LMICs) remains limited. To evaluate physician acceptability of a CDSS and its association with cardiovascular risk factor control within the Collaborative Quality Improvement (C-QIP) trial in India. C-QIP was a multicentre randomised feasibility trial conducted in four tertiary hospitals in India. A CDSS was co-developed with clinical experts to support management of coronary heart disease, ischaemic stroke, and heart failure. In the trial, using patient-specific data including diagnosis, blood pressure, lipid profile, HbA1c, and current medications, the CDSS provided guideline-based treatment prompts across four domains: antihypertensives, lipid-lowering therapy, glucose-lowering therapy, and antiplatelets. Physicians reviewed prompts, while final treatment decisions remained at their discretion. Descriptive analyses assessed CDSS prompt acceptance, reasons for rejection, and associated changes in risk factor control. Over a 2-year period (September 2022-2024), the CDSS generated 1,509 prompts for 198 intervention patients; 67.5% were accepted (n=1,018). Acceptance was highest for glycaemia prompts (89.6%), followed by antiplatelets (69.1%), lipid-lowering therapy (66.5%), and blood pressure prompts (45.8%). Acceptance improved over time across all domains, most notably for blood pressure prompts, increasing from 40.4% at baseline to 51.6% at endline. By diagnosis, acceptance was highest in stroke (81.6%), compared with coronary heart disease (64.7%) and heart failure (59.1%). Greater clinical complexity was associated with lower acceptance, declining from 66.4% in patients with one cardiovascular condition to 63.4% in those with two. Acceptance of blood pressure prompts were associated with significant reductions of 9.5 mmHg (95% CI: 5.2–13.8) in systolic and 3.4 mmHg (95% CI: 0.9–6.0) in diastolic blood pressure. Acceptance of glycaemia prompts were associated with a 0.4% reduction in HbA1c (95% CI: 0.2–0.6). Reasons for CDSS prompt rejection included preference to maintain existing regimens based on cardiac status or recent prescriptions, and contraindications such as statin intolerance, hyperkalaemia, and impaired renal function. In this multicentre feasibility trial, CDSS was acceptable to physicians and associated with improved cardiovascular risk factor control in a resource-limited setting. Providing actionable patient data and guideline-responsive CDSS prompts directly to physicians may improve quality of cardiovascular care in LMICs.C-QIP - Table 1C-QIP - Table 2