Interobserver and intraobserver variability in the histologic assessment of oral lichen planus.

Abstract

The purpose of this study was to evaluate interobserver and intraobserver variability in the histopathological assessment of oral lichen planus (OLP) since this may influence the outcome of studies on epidemiology, treatment and prognosis. Five oral pathologists examined 60 microscopic slides, not being informed about the original histopathological assessment. Forty-five of the cases had been originally signed out as OLP; the remaining 15 cases represented a mixture of other oral white lesions. No clinical information or patient data were provided with the cases. Each reviewing pathologist was asked to apply the WHO definition of OLP and to categorize each case as either: 1) evident OLP, 2) compatible with OLP, or 3) no histological support for OLP. After 2 months, each of the five reviewing pathologists were given 45 slides that were randomly retrieved from the original 60. Interobserver and intraobserver variability were assessed by calculation of unweighted kappa statistics. Interobserver agreement varied from 0.20 (poor) to 0.51 (moderate), while the intraobserver agreement varied from 0.50 (moderate) to 0.67 (substantial). Histopathological assessment of OLP, based on the available WHO definition, is a rather subjective and insufficiently reproducible process. Stricter diagnostic criteria are required in order to obtain a more reproducible diagnosis of OLP.