[Stratification of the population in the Basque Country: results in the first year of implementation]

Abstract

Predictive models allow populations to be stratified according to their health requirements for the following year. They offer health care organizations the opportunity to act proactively, designing specific interventions adapted to the level of need of different groups of people. The "Strategy for tackling the challenge of chronic illness in the Basque Country" proposes the use of such models, integrating them with other policies. The prospective categorization of all the population assigned to Osakidetza was performed for the first time in 2010 using the Johns Hopkins Adjusted Clinical Groups predictive model (ACG-PM). For this purpose, already recorded information extracted from electronic health records of primary care and hospital discharge reports was used. This article discusses the advantages of the combined use of various sources of information, and describes the application of the stratification in three programs, targeted at chronic patients who suffer different burdens of comorbidity.

Figura 1

Grupos generados por el sistema ACG-PM para clasificar diagnósticos y algunos ejemplos.

Figura 2

Variables incluidas por el modelo predictivo ACG-PM.