OBJECTIVES: With a system of qualitative indicators, to analyse the pharmaceutical prescription of general practitioners (GPs), and to evaluate the relationship of these indicators to the overall pharmaceutical prescription expenditure per inhabitant. DESIGN: Retrospective descriptive study. SETTING: Primary care. MEASUREMENTS AND MAIN RESULTS: The drugs prescription of 285 GPs from 32 primary care teams was evaluated, with the individual prescription of each doctor as the unit of analysis. The prescription was classified in 3 categories according to its intrinsic value (IV): low (< or = 75%), medium (76-79%) and high (> or = 80%). Selected as tracers of over-prescription were: daily dose per inhabitant (DDI) of antibiotics (AB), DDI of non-steroid anti-inflammatory drugs (NSAID), and DDI of ulcer drugs (ULC). Selected as tracers of selection were: % DDI third-generation cephalosporins/DDI total cephalosporins; % DDI broad-spectre quinolones/DDI total quinolones; and % DDI NSAID/DDI NSAID plus analgesics. Quantitative indicators studied were: total expenditure per allocated population, cost per drugs prescription of doubtful efficacy, and cost per daily dose of AB, NSAID and ULC. Variance analysis, including the Scheffe test for multiple comparisons and Pearson's linear correlation, was applied. 26% of the prescriptions had an IV below 75%, and 34% had an IV above 80%. The means of DDI of AB among the categories of IV were different (p < 0.0001), as were those of DDI of NSAID (p < 0.0001) and of ULC (p = 0.007). Lower consumption of AB, NSAID and ULC was found in prescriptions with the highest IV %. The third-generation cephalosporins and the NSAID + analgesics showed significant differences in the three IV categories (p < 0.0001 and p = 0.041), unlike broad-spectrum quinolones (p = 0.18). The total expenditure per allocated population was less for GPs whose prescriptions had the highest IV %; whereas the cost per prescription and cost per daily dose showed no significant differences for IV categories. CONCLUSIONS: The doctors with the best qualitative profile on these indicators had less expenditure per inhabitant. However, no differences were found in the cost per prescription or cost per treatment between doctors. Therefore, interventions must prioritize improving drug prescription quality rather than just promoting changes to lower-cost drugs.
OBJECTIVES: With a system of qualitative indicators, to analyse the pharmaceutical prescription of general practitioners (GPs), and to evaluate the relationship of these indicators to the overall pharmaceutical prescription expenditure per inhabitant. DESIGN: Retrospective descriptive study. SETTING: Primary care. MEASUREMENTS AND MAIN RESULTS: The drugs prescription of 285 GPs from 32 primary care teams was evaluated, with the individual prescription of each doctor as the unit of analysis. The prescription was classified in 3 categories according to its intrinsic value (IV): low (< or = 75%), medium (76-79%) and high (> or = 80%). Selected as tracers of over-prescription were: daily dose per inhabitant (DDI) of antibiotics (AB), DDI of non-steroid anti-inflammatory drugs (NSAID), and DDI of ulcer drugs (ULC). Selected as tracers of selection were: % DDI third-generation cephalosporins/DDI total cephalosporins; % DDI broad-spectre quinolones/DDI total quinolones; and % DDI NSAID/DDI NSAID plus analgesics. Quantitative indicators studied were: total expenditure per allocated population, cost per drugs prescription of doubtful efficacy, and cost per daily dose of AB, NSAID and ULC. Variance analysis, including the Scheffe test for multiple comparisons and Pearson's linear correlation, was applied. 26% of the prescriptions had an IV below 75%, and 34% had an IV above 80%. The means of DDI of AB among the categories of IV were different (p < 0.0001), as were those of DDI of NSAID (p < 0.0001) and of ULC (p = 0.007). Lower consumption of AB, NSAID and ULC was found in prescriptions with the highest IV %. The third-generation cephalosporins and the NSAID + analgesics showed significant differences in the three IV categories (p < 0.0001 and p = 0.041), unlike broad-spectrum quinolones (p = 0.18). The total expenditure per allocated population was less for GPs whose prescriptions had the highest IV %; whereas the cost per prescription and cost per daily dose showed no significant differences for IV categories. CONCLUSIONS: The doctors with the best qualitative profile on these indicators had less expenditure per inhabitant. However, no differences were found in the cost per prescription or cost per treatment between doctors. Therefore, interventions must prioritize improving drug prescription quality rather than just promoting changes to lower-cost drugs.