Objectives: To study the frequency of medical complaints and need for routine ED medical, laboratory, and toxicologic clearance for patients presenting with psychiatric chief complaints.
Methods: A retrospective, observational analysis of psychiatric patients seen in an urban teaching hospital ED over a 2-month period was performed. The individual sensitivities of history, physical examination, vital signs, and complete blood counts and chemistry panels for identifying medical problems were determined. The sensitivities and predictive values of patient self-reporting of recent illicit drug and ethanol use were also determined.
Results: 352 patients were seen with psychiatric chief complaints. A complete data set was available for 345 patients (98%). Of those with complete data, 65 (19%) had medical problems of any type. History, physical examination, vital signs, and laboratory testing had sensitivities of 94%, 51%, 17%, and 20%, respectively, for identifying these medical problems. Screening without universal laboratory testing would have missed 2 asymptomatic patients with mild hypokalemia. Patient self-reporting had a 92% sensitivity, a 91% specificity, an 88% positive predictive value (PPV), and a 94% negative predictive value (NPV) for identifying those with a positive drug screen, and a 96% sensitivity, an 87% specificity, a 73% PPV, and a 98% NPV for identifying those with a positive ethanol level.
Conclusion: The vast majority of medical problems and substance abuse in ED psychiatric patients can be identified by initial vital signs and a basic history and physical examination. Universal laboratory and toxicologic screening of all patients with psychiatric complaints is of low yield.