Diagnosis of pneumonia in the ED has poor accuracy despite diagnostic uncertainty

In 2007, the Centers for Medicare and Medicaid Services created a measure known as "diagnostic uncertainty" in emergency department (ED) pneumonia admissions. This documentation excludes the antibiotic timing measure, as pressure to quickly diagnose pneumonia may serve to reduce overall accuracy.

The objective of the study was to determine the correlation between ED and final discharge diagnosis of pneumonia and measure the effect of invoking the diagnostic uncertainty documentation on accuracy.

We retrospectively reviewed all ED pneumonia admissions among adults from July to October 2008. We analyzed the effect of invoking the diagnostic uncertainty documentation in the ED by comparing against final outcomes. We then performed a multivariate analysis to adjust for the potential effects of sex, age, Emergency Severity Index (ESI) score, weekend arrival, and level of ED-attending physician staffing.

Of 401 patients who were admitted with pneumonia, 297 (74%) had a discharge diagnosis of pneumonia, with 72 (18%) of those diagnoses being the primary outcome. Diagnostic uncertainty documentation was used in 11% (45/401). This documentation did not significantly alter the odds of a primary pneumonia discharge diagnosis (odds ratio, 0.68; 95% confidence interval, 0.28-1.7) but did reduce the odds of pneumonia being diagnosed (odds ratio, 0.43; 95% confidence interval, 0.23-0.81). Sex, age, day of week, and (ESI) score remained nonsignificant predictors.

Correlation between ED and discharge diagnosis of pneumonia was limited. Use of diagnostic uncertainty documentation decreased the likelihood of a hospital discharge diagnosis of pneumonia. Further analysis of the effects of artificially imposed time constraints on ED diagnoses appears warranted.