Utility of inoculum counting (Walshe and English criteria) in clinical diagnosis of onychomycosis caused by nondermatophytic filamentous fungi

Opportunistic onychomycosis caused by nondermatophytic molds may differ in treatment from tinea unguium. Confirmed diagnosis of opportunistic onychomy cosis classically requires more than one laboratory analysis to show consis tency of fungal outgrowth. Walshe and English in 1966 proposed to extract s ufficient diagnostic information from a single patient consultation by coun ting the number of nail fragments positive for inoculum of the suspected fu ngus. Twenty fragments were plated per patient, and each case in which five or more fragments grew the same mold was considered an infection by that m old, provided that compatible filaments were also seen invading the nail ti ssue by direct microscopy. This widely used and often recommended method ha s never been validated. Therefore, the validity of substituting any techniq ue based on inoculum counting for conventional follow-up study in the diagn osis of opportunistic onychomycosis was investigated. Sampling of 473 patie nts was performed repeatedly. Nail specimens were examined by direct micros copy, and 15 pieces were plated on standard growth media. After 3 weeks, ou tgrowing dermatophytes were recorded, and pieces growing any nondermatophyt e mold were counted. Patients returned on two to eight additional occasions over a 1- to 3-year period for similar examinations. Onychomycosis was eti ologically classified based on long-term study. Opportunistic onychomycosis was definitively established for 86 patients. Counts of nondermatophyte mo lds in initial examinations were analyzed to determine if they successfully predicted both true cases of opportunistic onychomycosis and cases of insi gnificant mold contamination. There was a strong positive statistical assoc iation between mold colony counts and true opportunistic onychomycosis. Log istic regression analysis, however, determined that even the highest counts predicted true cases of opportunistic onychomycosis only 89.7% of the time . The counting criterion suggested by Walshe and English was correct only 2 3.2% of the time. Acremonium infections were especially likely to be correc tly predicted by inoculum counting. Inoculum counting could be used to indi cate a need for repeat studies in cases of false-negative results from labo ratory direct microscopy. Inoculum counting cannot serve as a valid substit ute for follow-up study in the diagnosis of opportunistic onychomycosis. It may, nonetheless, provide useful information both to the physician and to the laboratory, and it may be especially valuable when the patient does not present for follow-up sampling.