Structural analysis of causal pathways between adherence, satisfaction and clinical outcomes in hypertensive patients using a mobile adherence intervention: A SEM–NCA–cIPMA approach

Abstract Poor medication adherence remains a major barrier to effective hypertension control, particularly in low-resource settings. This study aimed to explore causal pathways, necessity conditions and optimization priorities of the CareAide mobile application for improving medication adherence and clinical outcomes among hypertensive patients. A prospective, randomised, open-label, two-arm trial was conducted at University Malaya Medical Centre. Adults (N = 275) with hypertension and low Morisky Medication Adherence Scale (MMAS-8) scores were randomised to CareAide plus usual care or usual care alone for six months. Primary outcome was adherence (MMAS-8) at third (F1) and sixth month (F2); secondary outcomes included systolic/diastolic blood pressure (Hypertension F1, F2) and application satisfaction via Mobile Adherence Satisfaction Scale (MASS). Structural Equation Modelling (SEM), Necessary Condition Analysis (NCA) and Combined Importance–Performance Map Analysis (cIPMA) tested hypotheses from an integrated theoretical framework. Intervention showed large positive effects on adherence at (F1 β = 0.58, F2 β = 0.79, P < 0.001 ). Six-month adherence mediated the intervention’s impact on blood pressure (indirect β = −0.16, P = 0.006 ) and together with early adherence, formed a sequential mediation chain (β = −0.05, P = 0.01 ). NCA showed the intervention (d = 0.18) and early adherence (d = 0.28) were necessary for high F2 adherence; the intervention (CE-FDH = 100%) and prior blood pressure control (CE-FDH = 48.58%) were necessary for optimal F2 blood pressure. Feature satisfaction (FS) was both sufficient (β = 0.20, P < 0.001 ) and necessary (d = 0.29) for sustained adherence. cIPMA identified the intervention, early adherence and FS as high-importance, low-performance targets. CareAide significantly improves adherence and blood pressure via a temporally sequenced behavioral pathway with early adherence gains and FS are critical, actionable factors. Trial registration: CAREAide Trial (ClinicalTrials.gov identifier: NCT06068309).