Validity of clinical diagnosis in dermatology.

OBJECTIVE

We evaluated the accuracy of the clinical diagnostic process performed by dermatologists.

MATERIAL AND

METHODS

We conducted a cross-sectional study, based on a retrospective search of data at a dermatology outpatient department. Fifteen skin diseases were selected. A group of patients with those diseases at their first outpatient visit and a further histopathological study were included. The accuracy of any given clinical diagnosis was studied taking the histopathological diagnosis as the gold standard. Validity indices were calculated.

RESULTS

We studied 2188 patients. Skin diseases were more prevalent among women (p < 0.05). While the clinical diagnostic process displayed a high level of validity in melanocytic naevi and basal cell carcinoma, a poor level was nevertheless observed for Bowen's disease (sensitivity 22.8%, specificity 96.5%). The most frequent sites of appearance of dermatological problems were face and trunk.

CONCLUSIONS

Evaluation of diagnosis in pathologies having an available "gold standard" enables the accuracy of clinical diagnoses to be verified. The validity of clinical diagnosis was high.