LAUSR

OBJECTIVE Few reviews and no meta-analyses have explored the utility of investigations, such as laboratory tests, among patients presenting with psychiatric symptoms, and none has explored the yield of history and physical examination. A meta-analysis of studies exploring the utility of "medical clearance" among adult psychiatric patients was conducted. METHODS PubMed, PsycInfo, and Web of Science were systematically searched from inception until February 15, 2021. Primary outcome was detection by investigations (e.g., bloodwork and imaging), history, or physical examination of an illness that caused or aggravated psychiatric symptoms or was comorbid and that resulted in change in the patient's diagnosis or management ("yield"). A mixed-effects meta-analysis with inverse-variance weighting was used to pool results. RESULTS Twenty-five cross-sectional studies were included. Pooled yield of investigations was 1.1% (95% confidence interval [CI]=0.5%-2.2%), although yield was relatively higher among disoriented, agitated, or older patients. Yield was higher in the inpatient setting, compared with the emergency room, with similar results by approach (protocolized versus nonprotocolized). Compared with investigations, yield of history and physical examination was higher (15.6%, 95% CI=9.1%-25.6%; and 14.9%, 95% CI=8.1%-25.9%, respectively), with nonsignificant differences by evaluator (psychiatrist versus nonpsychiatrist) for physical examination. CONCLUSIONS Investigations were of relatively low yield, especially when weighed against cost and potential harm, and they should not be routinely conducted for patients presenting with primarily psychiatric complaints, although certain subgroups may benefit. History and physical examination, by contrast, should be undertaken for all patients, ideally with participation of the consulting psychiatrist.