COST-EFFECTIVENESS OF SCREENING FOR MICROALBUMINURIA USING IMMUNOCHEMICAL DIPSTICK TESTS OR LABORATORY ASSAYS IN DIABETIC-PATIENTS

To analyse the cost-effectiveness ratio of screening for microalbuminu ria in diabetic patients using either dipstick tests or laboratory ass ays, 506 diabetic patients were screened for microalbuminuria using bo th a traditional laboratory assay (strategy I) or a laboratory assay o nly in the case of a positive dipstick result (strategy II). Dipstick pre-screening was considered positive if at least one of the tests per formed by the two different operators showed an albumin excretion rate > 20 mug min-1. It was performed using a new dipstick, Micral-Test(R) , designed to distinguish low albumin concentrations. Biological assay was the reference method. Costs were related to laboratory assays (st rategy I) or to dipstick tests and laboratory assays for positive resu lts (strategy II). The loss of effectiveness was related to false nega tive results of strategy II. The double dipstick pre-screening showed a sensitivity of 90.8 % and a specificity of 80.1 %. Its predictive va lue was 97 % for a negative result and 55.6 % for a positive result. F alse positive results were associated with elevated urinary volumes. C ompared with strategy I, strategy II showed a sensitivity of 90.8 %, a specificity of 100 %, and predictive values of 100 % and 97.5 %, resp ectively, for positive and negative results. In a fictitious cohort of 10000 patients, strategy II yielded a gain of 16 750 pound on the fir st year, which decreased to 5345 pound after 30 years. The loss in eff ectiveness was estimated at 2.38 quality adjusted life years of a diab etic patient (QALY(d)) initially, and decreased to 0.91 QALY(d) after 30 years, the annual cost-effectiveness ratio being close to 6600 poun d QALY(d-1). Sensitivity analysis found that frequency of false negati ve results, prevalence, and annual incidence of microalbuminuria were the most important factors likely to influence the cost-effectiveness ratio. These results suggest that strategy 11 can be especially cost-e ffective: (1) for annual screening in diabetic patients with urinary e xcretion < 2 ml min-1 and exposed to a low risk of microalbuminuria; ( 2) for frequent screening in patients exposed to a high risk of microa lbuminuria.