Initial emergency department diagnosis and return visits: risk versus perception.

STUDY OBJECTIVES

To identify diagnostic predictors of return emergency department visits, and to compare actual and perceived associations between initial ED diagnosis and revisits to help identify target diagnoses for prevention strategies.

METHODS

This 2-part study involved a retrospective observational study and a health professional survey. The study population consisted of all patients with 2 or more visits to the University of Michigan ED within a 3-day period between July 1995 and June 1996 ("early-return visitors"). Billing records identified the initial diagnoses of subsequent return visitors. The prevalence of each initial diagnosis was determined in the general ED population (n=52,553), early-return population (n=1,422), and early-return population admitted to the hospital ("return-admit," n=313). Surveys were distributed to all ED health professionals to assess their perception of the diagnoses most likely to return within 3 days. Relative risk (RR) ratios and 95% confidence intervals (CIs) were calculated.

RESULTS

Dehydration was the most common diagnosis in the general, early-return, and return-admit populations (prevalence: 7%, 15%, 25%, respectively). Dehydration was also the diagnosis with the highest risk for both early return and subsequent admission on early return (RR [95% CI]: 2.3 [2.0-2.6], 1.8 [1.5-2.3], respectively). Nearly two thirds of health professionals, however, did not identify dehydration as a diagnosis at high risk for return, and almost half did not consider dehydration a high risk for admission.

CONCLUSION

Initial ED diagnosis may be a useful predictor of early ED return and admission. Patients with an initial diagnosis of dehydration are at particularly high risk for early return and admission, yet providers underestimate the risk in this very common group. Screening a return ED population for high-frequency diagnoses may reveal underrecognized target groups for specific prevention strategies.