Ten years of an external quality program control in clinical microbiology: a statistical analysis of the results

Abstract

External quality assessment data of clinical microbiology obtained over ten years by the Colombian company PROASECAL in laboratories of seven countries in Central and South America are analyzed. The objective is to describe the quality of clinical microbiology results of the participants in the study period (2010–2021). This analysis includes the results of genus and species identification and proper reporting of both Gram staining and susceptibility testing. Data from 195 laboratories, which processed 31 bacteria, were analyzed. In this study, 32 % of laboratories were classified as good because they successfully determined the genus and species, 36 % fair, 12 % poor, and 20 % critical, to which recommendations for improvement were sent as part of the reports. In the analysis conducted for concordance between Genus, Morphology, and Gram, the success rates obtained are between 83 % and 94 %. Regarding antibiograms, the performance observed in Gram-positive bacteria ranges from 57 to 98 %, while in Gram-negative bacteria, it is between 82 and 98 %. The descriptive tools to explain external quality assessment results in clinical microbiology are limited to the percentage of successes in each item to be reported. For the analysis of the population data set, in this article, we propose to use contingency tables (including estimates of sensitivity, specificity, and predictive values) and equality of proportions tests as techniques that facilitate the interpretation of the data. On the other hand, the statistical analysis described here can be used as an analysis methodology for similar studies.