The effects of an innovative integrated care intervention in Brazil on local health service use by dependent older people.

backgroundSince 2011, the Brazilian city of Belo Horizonte has been operating an innovative scheme to support care-dependent older people in disadvantaged communities: Programa Maior Cuidado (PMC - Older Person's Care Program). This paper examines two potential associations between inclusion in PMC on types of outpatient health service utilization by dependent older people. The first is that being in PMC is associated with a higher frequency of outpatient visits for physical rehabilitation. The second is that being in PMC is associated with a higher frequency of planned versus unplanned outpatient visits.

methodsWe apply a quasi-experimental design to a unique set of health administrative data recording visits to outpatient health services. We focus on comparisons of the universe of visits, transformed to ratios of planned/unplanned visits and rehabilitation/other reasons for visiting the outpatient service. First, we preprocess our sample through different matching techniques such as 'coarsened exact matching' (CEM), 'nearest neighbor' based on logit scores (NN), 'optimal pair' (OP) and 'optimal full' (OF) methods. Second, we estimate marginal effects of being in PMC on our outcomes of interest. We use Poisson regressions controlling for individual and community factors and use robust standard errors. Our results are presented as the comparative incidence ratio of PMC on rehabilitation and planned visits.

resultsWe find significant positive incidence rates for belonging to PMC for both outcomes of interest under all matching specifications. Poisson models using CEM shows a higher incidence rate for planned visits in comparison to unplanned visits, 1.3 (95% CI 1.1-1.4), by PMC patients compared to the non-PMC controls, and a higher proportion of visits for rehabilitation, 3.4 (95% CI 1.7-6.8). Similar positive results are found across other matching methods and models.

conclusionsOur analysis reveals significant positive associations between older people included in PMC and a matched set of controls for a greater ratio of making outpatient visits that were planned, rather than unplanned. We find similar associations for the proportion of visits made for rehabilitation, as opposed to other reasons. These findings indicate that PMC influences some elements of outpatient health service utilization by dependent older people.