Case-control analysis with partial knowledge of exposure misclassificationprobabilities

Consider case-control analysis with a dichotomous exposure variable that is subject to misclassification. If the classification probabilities are know n, then methods are available to adjust odds-ratio estimates in light of th e misclassification. We study the realistic scenario where reasonable guess es, but not exact values, are available for the classification probabilitie s. If the analysis proceeds by simply treating the guesses as exact, then e ven small discrepancies between the guesses and the actual probabilities ca n seriously degrade odds-ratio estimates. We show that this problem is miti gated by a Bayes analysis that incorporates uncertainty about the classific ation probabilities as prior information.