Evaluation of mycology laboratory proficiency testing

Changes over the last decade in overt proficiency testing (OPT) regulations have been ostensibly directed at improving laboratory performance on patie nt samples. However, the overt (unblinded) format of the tests and regulato ry penalties associated with incorrect values allow and encourage laborator ians to take extra precautions with OPT analytes. As a result OPT may measu re optimal laboratory performance instead of the intended target of typical performance attained during routine patient testing. This study addresses this issue by evaluating medical mycology OPT and comparing its fungal spec imen identification error rates to those obtained in a covert (blinded) pro ficiency testing (CPT) program. Identifications from 188 laboratories parti cipating in the New York State mycology OPT from 1982 to 1994 were compared with the identifications of the same fungi recovered from patient specimen s in 1989 and 1991 as part of the routine procedures of 88 of these laborat ories. The consistency in the identification of OPT specimens was sufficien t to make accurate predictions of OPT error rates. However, while the error rates in OPT and CPT mere similar for Candida albicans, significantly high er error rates were found in CPT for Candida tropicalis, Candida glabrata, and other common pathogenic fungi, These differences may, in part, be due t o OPT's use of ideal organism representatives cultured under optimum growth conditions. This difference, as well as the organism-dependent error rate differences, reflects the limitations of OPT as a means of assessing the qu ality of routine laboratory performance in medical mycology.