Abstract
<title>Abstract</title> <p>Poor medication adherence contributes substantially to treatment failure and preventable morbidity in chronic non-communicable diseases, yet objective monitoring remains limited in resource-constrained primary-care settings. We conducted a 10-month mixed-methods pilot study of STAMP, a digital adherence system, involving 194 patients and 200 concurrent controls across five urban primary health centres in Tamil Nadu, India. Adherence among STAMP users was 97.0%, with only 3.0% non-adherence. Nearly all delayed doses were recovered within 75 minutes through an automated escalation protocol, with 77.0% resolved before healthcare-worker involvement and a median recovery time of 18 minutes. Clinical outcomes improved more among STAMP users than controls, including a large effect size for post-prandial blood glucose improvement (Cohen's d = 0.85). These findings demonstrate the feasibility of real-time digital adherence monitoring in government primary care and support further evaluation in a fully powered randomized controlled trial.</p>