Análise da sustentabilidade dos beneficários de uma operadora de plano de saúde: evidências para o Norte e Nordeste

Abstract

The Operators of Health Plans (OPS) grew in the Brazilian economic scenario. But they are barred by ANS regulatory rules that prevent them from freely negotiating, in terms of pricing, their plans according to the characteristics of each market. In this way the solvency analysis of your portfolios have become increasingly important for a highly competitive market. Thus, this work sought to identify the best statistical technique for a solvency analysis. The survey sampled about 3.423 beneficiaries of an OPS in the state of North and Northeast. Four techniques were used to predict the sustainability of the portfolio: logistic regression, discriminant analysis, tree classification and regression, and ramdom forest. To analyze the technique with better performance, the Roc Curve was used. The results showed that the statistical technique of logistic regression best identifies the variables that explain the sustainability of a health insurance portfolio.