Purpose: This study aimed to develop a context-sensitive conceptual framework for health systems responsiveness in chronic care by integrating empirical evidence and literature. Methodology: A quasi-experimental design was employed across three Kenyan hospitals, incorporating baseline, intervention (training), and end-line evaluations. From a total of 853 patients, 323 were sampled using Fisher’s formula, with 258 retained at end line. Responsiveness was measured using structured five-point Likert-scale questionnaires, later dichotomized into favorable and unfavorable categories using a demarcation threshold. Quantitative analysis included logistic regression to identify predictors. Qualitative data were collected through three provider focus group discussions (27 participants), three client focus groups (20 participants), and three key informant interviews, and analyzed thematically. Results: The intervention improved overall responsiveness from 63.7% to 67.4%, while the proportion of patients experiencing responsive care rose from 38.3% to 52.7%. Odds of receiving responsive care nearly doubled. Significant predictors included structural capacity (OR=2.171), accountability (OR=2.730), organizational culture (OR=2.267), and justice perceptions (OR=2.909). Model explanatory power increased from 15.7% at baseline to 32.8% post-intervention, and 40.4% when socio-demographic variables were included. Qualitative findings revealed four themes: resource constraints, weak accountability, limited patient agency, and structural inequalities shaping responsiveness outcomes. Conclusion: It views responsiveness as an outcome shaped by three elements: context (policy, resources, sociocultural and environmental factors), drivers (valuations, accountability, access, structure, culture, justice and perceptions), and actors (engaged clients and competent providers). This dynamic interaction emphasizes systemic investment in policy, training, and equity to enhance patient-centered care and health outcomes.
Hillary et al. (Mon,) studied this question.