Evaluation of two methods for improving quality of diagnosis of bacteriuria by culture in primary healthcare.

This study evaluates the effect of training on the results from Uricult Trio and an established urine culture when used at primary healthcare laboratories in two Swedish counties, Uppsala and Värmland. Urine cultures and dipslides, Uricult Trio, performed at these laboratories were interpreted a second time at central laboratories. Interpretation errors at the primary healthcare laboratories were calculated. Primary healthcare laboratories also received external control panels with urine cultures and dipslides. There was one study period each year for 3 years in Uppsala and for 2 years in Värmland. A training programme was completed between study periods in Värmland. In Uppsala, primary healthcare laboratory results could be reviewed, as interpretations by the central laboratory were returned to them. The main outcome measures were the percentage of interpretation errors which, in the first study period, was 33-39%. This dropped to 15-19% in the second study period. In the results from the external control panels there were no striking differences between the studied areas and Sweden as a whole, except that Uppsala showed a better result in reporting E. coli and failed in 10% compared to Sweden 46%. A method for both quality assessment and education is to ask the primary healthcare laboratories to send cultures to the central laboratory for interpretation requesting their return to the primary healthcare laboratory with the interpretation from the central laboratory attached.