OBJECTIVES:
To estimate an econometric model for the apportionment of costs related to the litigation of medicines not covered by the Unified Health System (acronym in Portuguese SUS) among the federated entities of Brazil. METHODS:
The analysis was based on the method developed by the National Treasury of Brazil to assess the fiscal situation of subnational entities, based on indicators of debt, current savings and liquidity, known as "Payment Capacity” (acronym in Portuguese CAPAG), categorized as high, medium, low, and very low. The econometric method used was multivariate ordinal logistic regression, using CAPAG for each federated entity as a dependent variable. The pre-selected independent variables were GDP, GDP per capita, population, Social Vulnerability Index (IVS), Human Development Index (HDI), debt, current savings, and liquidity, with 2022 data obtained from government sources and UNDP. To select the most appropriate model, the stepwise method and the R software were used.
RESULTS:
The statistical tests applied to the set of independent variables showed that some variables in the model (GDP, GDP per capita, IVS, HDI, and Population) presented multicollinearity evidenced by the Variance Inflation Factors (VIF) and the tolerance measures. Furthermore, these variables did not present statistical significance and did not increase the explanatory power of the CAPAG variable, according to Nagelkerke's Pseudo-R² analysis. The best adjustment of the model was achieved with the variables debt, current savings and relative liquidity, which were validated by absence of multicollinearity, low VIF values (≈1.000), high tolerance values (above 0.9), statistical significance (p ≤ 0.05), and adequate explanatory power (Nagelkerke's Pseudo-R² = 0.783).
CONCLUSIONS:
The linear apportionment percentages of the costs related to the litigation of medicines were determined using the matrix of linear combinations of CAPAG, derived from the three explanatory variables of the model. Thus, the proposed model can be employed in public policy planning.
