Association between hospitalisations for ambulatory care-sensitive conditions and primary healthcare physician specialisation: a longitudinal ecological study in Belo Horizonte, Brazil.

OBJECTIVES: Ambulatory care sensitive conditions (ACSCs) are conditions for which the provision of timely and skilled primary care can reduce risks of hospitalisation when preventing, treating or controlling a disease. For this reason, hospitalisations for ACSC have been commonly employed by health systems as an indicator of effectiveness for the primary level of care. This study aims to evaluate whether the provision of primary care services by physicians with residency training in family medicine is associated with rates of general hospitalisations for ACSCs in the Brazilian Unified Health System network in the city of Belo Horizonte, Brazil. DESIGN: Longitudinal ecological study using a Generalised Linear Model for Gamma-distributed variables. SETTING: Primary healthcare centres in Belo Horizonte, Brazil, from January 2017 to December 2021, aggregated at the primary healthcare centres level. PARTICIPANTS: Data aggregated at the primary healthcare centre level, encompassing socioeconomic, professional and health-related variables. PRIMARY OUTCOME MEASURES: Incidence rates of hospitalisations for ACSCs, adjusted for age and sex. RESULTS: After adjusting for age, sex and socioeconomic variables using the Health Vulnerability Index, a higher concentration of family physicians was significantly associated with a lower incidence of hospitalisations for ACSCs. If all physicians in the primary care network were family physicians, compared with a scenario in which none were, an estimated 11.89% reduction in hospitalisations would be expected (95% CI 7.3% to 16.3%, p<0.05). Subgroup analyses confirmed these results across different levels of social vulnerability (low, medium, high, very high), time periods (prepandemic: 9.59% reduction; 95% CI 3.80% to 15.00%, p<0.05; pandemic: 15.93% reduction; 95% CI 9.76% to 20.91%, p<0.05) and groups of causes of hospitalisation (eg, diabetes: 32.80% reduction; 95% CI 22.5% to 41.7%, p<0.05). Furthermore, an estimated 10.58% reduction in hospitalisation costs was observed (95% CI 2.60% to 17.90%, p<0.05). CONCLUSIONS: The findings suggest that specialisation in family medicine positively impacts health outcomes by reducing hospitalisations for ACSCs. These results can inform the development of evidence-based public policies to enhance primary care effectiveness.