Emergency medicine physicians' ability to predict hospital admission at the time of triage Journal Article uri icon
Overview
abstract
  • BACKGROUND: We seek to determine if experienced emergency medicine physicians can accurately predict the likelihood of admission for patients at the time of triage. Such predictions, if proven to be accurate, could decrease the time spent in the ED for patients who will ultimately be admitted by hastening downstream workflow. METHODS: This is a prospective cohort study of experienced physicians at a large urban hospital. Physicians were asked to predict the likelihood of admission for patients based only on information available in the EMR at the time of triage. Physicians also predicted the service to which the patients would be admitted. Physicians provided a confidence level of their prediction. Measures of predictive accuracy were calculated, including sensitivity, specificity, and area under the receiver operating characteristic curve. RESULTS: 35 physicians evaluated 398 patient charts and made predictions. Sensitivity of determining admission for the entire cohort was 51.8%. The specificity was 89.1%. For those predictions made with a confidence level of >90%, sensitivity was 61.5% and specificity was 95.7%. Among physicians correctly predicting admission, the admitting service was predicted accurately 88.6% of the time. CONCLUSION: Physicians performed poorly at predicting which patients would be admitted at the time of triage, even when they were confident in their predictions. Conversely, physicians accurately predicted who would be discharged. Physicians predicted with reasonable accuracy the service to which patients were ultimately admitted. More research and operational assessment needs to be performed to determine if these predictions can help improve ED efficiency.

  • Link to Article
    publication date
  • 2019
  • Research
    keywords
  • Emergency Medicine
  • Forecasting
  • Hospitalization
  • Physicians
  • Prospective Studies
  • Sensitivity and Specificity
  • Additional Document Info
    volume
  • 37
  • issue
  • 3