Study objective: To evaluate the sensitivity and specificity of emergency physician-performed pelvic sonography (EPPPS) and its effect on length of stay (LOS) in the emergency department.
Methods: The study involved a prospective observational convenience sample of women in early pregnancy requiring ED pelvic ultrasound evaluation. Differences in LOS were analyzed with the use of the Mann-Whitney test, with EPPPS and consultation status as classification variables. Results were stratified by diagnosis to identify patient subsets in which EPPPS had the greatest impact. A multivariate model was used to determine the independent associations of EPPPS and consultation status with LOS.
Results: Among 115 pregnant patients, those who underwent EPPPS had a decreased median LOS compared with those who received pelvic sonography administered by radiologists or obstetrics-gynecology consultants (60 versus 180 minutes; P < .001). Obstetrics-gynecology consultation was associated with an increase in LOS from 60 to 170 minutes (P < .001). Stratification revealed that the decreased LOS was most significant in patients with viable intrauterine pregnancies (IUPs) and was not seen in those with abnormal pregnancies. This result was attributed primarily to the need for consultation in the latter group. Multivariate analysis demonstrated that EPPPS remained significantly associated with decreased LOS overall, after adjustment for the potentially confounding effect of consultation status (P < 001). The sensitivity of EPPPS for IUP was 94% (95% confidence interval [CI], 82% to 98%), and the specificity was 100% (95% CI, 83% to 100%). All 10 ectopic pregnancies, six of which were detected by EPPS and are included in the aforementioned sensitivity and specificity analysis, plus four detected by NEPPPS, were correctly identified; none was missed. No patient had an adverse outcome as a result of EPPPS.
Conclusion: Pelvic ultrasound performed by emergency physicians after a brief period of training shortens ED LOS in women in early pregnancy, particularly in those with viable IUPs. This practice appears to be sensitive, specific, and safe.