OBJECTIVE: To design a mathematical model which permits pharmacy budgets to be assigned to Health Centres (HC), examining the social and demographic variables and health service usage which explain the variability of the pharmaceutical expenditure (PE) of the HCs. DESIGN: A descriptive, crossover study. SETTING: 17 HC of the Insitut Català de la Salut (Catalan Health Service) in the city of Barcelona during 1994. MEASUREMENTS AND MAIN RESULTS: Relationships among the following variables at the 17 HCs were studied: pharmaceutical expenditure per inhabitant (PEi), frequency of attendance (FRA), care pressure (CP), percentage of the population 65 or over (P65), percentage of the population with medical records (PPR), index of family economic capacity (IFEC), ratio of comparative mortality (RCM) and ratio of potential years of life lost (RPYLL). In the bivariant analysis, those variables with a statistically significant linear correlation with PEi were FRA (r = 0.67; p < 0.01), PPR (r = 0.56; p <0.01), IFEC (r = -0.68; p < 0.01), RCM (r = 0.61; p < 0.01) and RPYLL (r = 0.62; p <0.01). In the multivariant analysis, IFEC, P65 and FRA enabled 94% of the PEi variability to be explained (r2 = 0.94; p < 0.001). Through multiple regression a mathematical formula for calculating the PE of HCs was obtained. CONCLUSION: The model presented enables pharmacy budgets to be assigned to the HCs if the IFEC, P65 and FRA of the population attended is known.
OBJECTIVE: To design a mathematical model which permits pharmacy budgets to be assigned to Health Centres (HC), examining the social and demographic variables and health service usage which explain the variability of the pharmaceutical expenditure (PE) of the HCs. DESIGN: A descriptive, crossover study. SETTING: 17 HC of the Insitut Català de la Salut (Catalan Health Service) in the city of Barcelona during 1994. MEASUREMENTS AND MAIN RESULTS: Relationships among the following variables at the 17 HCs were studied: pharmaceutical expenditure per inhabitant (PEi), frequency of attendance (FRA), care pressure (CP), percentage of the population 65 or over (P65), percentage of the population with medical records (PPR), index of family economic capacity (IFEC), ratio of comparative mortality (RCM) and ratio of potential years of life lost (RPYLL). In the bivariant analysis, those variables with a statistically significant linear correlation with PEi were FRA (r = 0.67; p < 0.01), PPR (r = 0.56; p <0.01), IFEC (r = -0.68; p < 0.01), RCM (r = 0.61; p < 0.01) and RPYLL (r = 0.62; p <0.01). In the multivariant analysis, IFEC, P65 and FRA enabled 94% of the PEi variability to be explained (r2 = 0.94; p < 0.001). Through multiple regression a mathematical formula for calculating the PE of HCs was obtained. CONCLUSION: The model presented enables pharmacy budgets to be assigned to the HCs if the IFEC, P65 and FRA of the population attended is known.