A Prospective Cohort Study of a Community-based Primary Care Program's Effects on Pharmacotherapy Quality in Low-income Peruvians With Type 2 Diabetes and Hypertension

A pre-post study of visit-level data gathered prospectively during a community-based primary care program evaluated its effects on pharmacotherapy of type 2 diabetes and hypertension. The program was conducted by Asociación Siempre Salud, a Peruvian non-profit organization, in three neighborhoods ("the community") in Pueblo Nuevo, Chincha District, Peru. Door-to-door surveys identified all persons residing in the study community. Persons > 35 years old were eligible, without exclusions, to participate in type 2 diabetes and hypertension screening and diagnosis. Most participated in mass screenings between September and December 2011, others in later ad hoc screenings. Patients with negative screening results on medications were re-screened following medication withdrawal. Those with confirmed diagnoses were eligible to participate in the program.

Three binary exposures were evaluated for their effects on adherence to four treatment standards and one composite standard: 1) 'program exposure' (post- vs. pre-exposure); 2) 'program retention in care' (>50% of allowable time in both home and clinic care periods, compared to <50% in either care period (but not both)), and 3) 'care period', (clinic compared to home care). 'Care period treatment time' (>50% of allowable time versus less) was evaluated as an independent variable in the study of the care period exposure.

During the pre-exposure period, usual care was provided by two public hospitals, several Ministry of Health clinics, private clinics and pharmacies, and Siempre Salud. The program, to which patients were then exposed, had components of four CCM elements (delivery system design, self-management, decision support, and community resources), CHWs, and no out-of-pocket costs for visits and medications. Self-management was based on national standards and utilized educational materials from professional societies and government agencies in the USA. Decision support consisted of guidelines-based standards adapted to our low-resource setting and simplified medication treatment protocols for diabetes, hypertension, and primary prevention of CVD. The delivery system consisted of two care models employed sequentially: first home care, then clinic care (see care period exposure).

During the home care period, the physician made an initial visit, after which CHWs made weekly visits to patients' homes. CHWs monitored clinical parameters, provided self-management education and support, tracked self-care behaviors, documented visits, acted on clinical alerts, and delivered medications. They entered encounter data into a spreadsheet and filled prescriptions under physician supervision. The physician made treatment decisions reliant on home visit data and patient care conferences. During the subsequent clinic period, patients made monthly visits to the clinic physician who provided all care.