Towards selective and improved performance of the mental status examination.

Performance of mental status examinations (MSE) is often useful but is not warranted for all new patients, except for training, research, special request, or policy. Reasons for doing MSE must be weighed against possible detrimental effects on patients or on patient-clinician relationships. Both sides are reviewed. Methods for ameliorating or preventing adverse effects are described. Validity of findings using MSE are frequently doubtful and examiners often misinterpret their significance. A decision tree is presented to help decide whether to use the MSE in particular instances.