Antidiabéticos orais e internamentos atribuíveis à diabetes em Portugal

Introduction: The growth of oral antidiabetic drugs spending has led to concerns questioning the health gains and benefits for the health system generated by these drugs. This study contributes to answering these questions. Material and Methods: Initially we present estimates of the three variables central to the analysis: 1) the prevalence of treated diabetes per year and health region, based on the consumption of oral antidiabetic drugs, 2) the number of hospitalization episodes and their costs attributable to diabetes, based on the relative risks and 3) a quantitative characterization of the consumption of oral antidiabetics by calculating average vintages by year and region. Through the use of observations for 10 years and five regions, a total of 50 observations, it was possible to estimate an econometric model explaining statistically both hospitalizations and hospital costs attributable to diabetes by regional characteristics, treated prevalence and the average vintage of oral antidiabetic drugs. Results: The results of the multiple regression models show that hospital costs are proportional to the treated prevalence, all else constant, but that the more recent the oral antidiabetic drugs used, the lower are hospital costs. The effects on the number of hospitalizations are similar. Discussion and Conclusions: For an average observation in the sample, if the average vintage of oral hypoglycemic agents had one additional year, then hospital costs would have a 5.3% reduction (about 11 million € in 2009) and the number of admissions would suffer a 3.8% reduction (about 3965 less episodes in 2009). A counterfactual exercise allows us to estimate that for 2009 the introduction of the class of DPP IV inhibitors reduced the number of hospitalizations attributable to diabetes by 8480 and saved € 23.3 million in hospital costs.