Epidemiology and Outcomes of Patients Presenting to United States Emergency Departments with Vaginal Bleeding

Introduction: There are significant gaps in knowledge regarding the epidemiology, management, and outcomes of patients presenting to the emergency department (ED) with vaginal bleeding. Methods: This was a retrospective, successional cross-sectional study using data from the National Hospital Ambulatory Medical Care Survey (NHAMCS) examining all adult patients presenting to EDs with vaginal bleeding from 2011–2019. Patients were stratified by age, race/ethnicity, and pregnancy status. Main outcomes were ultimate outcome severity, presenting vital signs, and diagnostic tests performed. We defined high-severity outcome as any patient who was dead on arrival, died in the ED, or during that hospitalization; any patient admitted to the intensive care or stepdown units or to the cardiac catheterization lab or the operating room; or patients transferred to a non-psychiatric hospital. Moderate severity was defined as any patient admitted to floor-level care, held in observation, or transferred to a psychiatric hospital. We defined low-severity outcome as any patient discharged home. Results: Patients presenting with a chief complaint of vaginal bleeding comprised 1.3% (95% CI, 1.2-1.4%,) of all ED visits, representing 14,620,933 total encounters. Of these patients, 53.0% (95% CI, 49.4-56.7%) were identified as pregnant. There was a lower prevalence of White patients presenting with this complaint compared to White patients presenting with any chief complaint (45.6% [95% CI, 41.9-49.4] vs 60.3% [95% CI, 57.7-62.8%]), with a reciprocal higher prevalence of Hispanic patients (21.1% [95% CI,17.7-24.5%] vs 13.2% [95% CI, 11.7-14.8%]). The majority of patients (88.1%, 95% CI, 86.1-90%) were classified as having a low-severity outcome, 10.3% (95% CI, 8.5-12.1%) were classified as moderate-severity, and 1.6% (95% CI,1.0-2.2%) as high-severity. Patients who were ultimately classified with high-severity outcomes had significantly higher shock indices at presentation and shorter wait times than patients with low-severity outcomes (0.75 [95% CI, 0.72-0.78] vs 0.68 [95% CI, 0.67-0.69], and 23.4 minutes [95% CI, 17.1-29.8] vs 41.7 minutes [95% CI, 37.1-46.4], respectively), despite no difference in median Emergency Severity Index triage score (2.5 [IQR 2.1-2.8] v 2.6 [IQR 2.2-2.9]). A quarter of patients (24.3% [95% CI, 20.8-27.7%]) received a pelvic exam: there were no significant differences in pelvic exam rate by age, pregnancy status, race/ethnicity, or ultimate outcome severity. Conclusion: Although most patients presenting to EDs with vaginal bleeding are discharged home, current triage models do not appear to appropriately risk-stratify higher risk patients. Disparities in presentation exist.