[Reliability of bacteriologic diagnosis of tuberculosis. Results of external quality control 1991-1992]

A report on the results of three proficiency-tests is presented. They were performed to test the efficiency of mycobacteriological laboratories. The incidence of false test-results varied between samples from 2 to 20%. In each survey between 10 and 20% of the participating laboratories arrived at unacceptably poor total assessments. The proportion of institutes where lack of expert competence can be suspected due to repeated failure in external quality tests, was between 3% (microscopy) and 13% (differentiation between tuberculosis bacteria and environmental mycobacteria). Reproducibility of good and poor total assessments in more than 80% of the laboratories indicates even if a laboratory participates in a survey only once, this will already supply quite reliable information on the efficiency of that laboratory. Even though many laboratories with poor performance in surveys try to improve their standards on their own, it is evident from the large number of institutes with repeated poor results, as well as from the irregular participation by many institutes in external quality tests, that external quality assurance of bacterio-logical tuberculosis diagnostics cannot be achieved by proficiency tests alone. Since the provisions of the Federal law on epidemics do not bestow sufficient legal competence on the health authorities in respect of control of the laboratories, and also because eventual sanctions imposed by the sickness insurance bodies cannot be effective in all cases, it is at present imperative that treating physicians take particularly great interest in the efficiency of the laboratories commissioned by them, for their patients' sake.