Costruzione di un modello decisionale per valutare la costo/efficacia delle strategie alternative nella terapia della polmonite acquisita in comunità

Abstract

Aim of this paper was to analyse the cost effectiveness of the main alternative (and complementary) strategies in the disease menagement of the Community-Acquired Pneumonia (CAP): hospital admission vs home-care, antibiotic parental vs oral therapy, switch vs no-switch therapy, and early discarge vs conventional hospitalization. The cost effectivenessanalysis (CEA) has been performed by implementing a general decision tree model wich describes all the main decisional and change nodes encountered in the clinical course from the firm sign and symptoms of CAP (root) to the final aoutcomes: full recovery or death (terminal nodes). We assumed the perspectives of three main institutional decision-makers: the society, the italian national healthcare system (NACS), and the hospital. In the perspective of society both the direct (health and non-health) costs and the indirect costs have been included, while in the perspective of the NACS only the health-direct costs were considered. In the perspective of the hospital we considered the overall mean expences sustained for each day of staying in the general and in the intensive care unit. Separately, the antibiotic treatement costs to hospital have been accounted. As effectiveness we considered the percentage of recovery for each class of mortality aqccording to fine. Most of the probability data used in the model were obtained or derived from the published literature. The cost were valued according to the Italian NACS charges and prices in use during the year 2000. According to the model structure, the main expenditure factor for the SSN is the hospitalization cost, while the home care is less expensive. The antibiotic parentenal therapy, during hospedalization or home care, is more expensive than the antibiotic oral therapy; but the cost difference between one therapy and the other is clearly lower than the cost difference between the hospitalization and the home care. The optimum expenditure situation for the SSN, the Society and also for the Hospital coul be obtained by decreasing the days of hospital stay in and by choosing to hospitalize the patients according to the death risk. The sensitivity analysies performed confirmed the robusteness of the results obtained with the model. However the model and its usefulness in decision-making will be definitely confirmed when clinical and epidemiological robust data on CAP in Italy will be available.