Using emergency physicians’ abilities to predict patient admission to decrease admission delay time

Exploiting the Features of Clinical Judgment to Improve Assessment of Disease Severity in the Emergency Department: An Acutelines Study

BACKGROUND

Clinical judgment, also known as gestalt or gut feeling, can predict deterioration and can be easily and rapidly obtained. To date, it is unknown what clinical judgement precisely entails. The aim of this study was to elucidate which features define the clinical impression of health care professionals in the ED.

METHOD

A nominal group technique (NGT) was used to develop a consensus-based instrument to measure the clinical impression score (CIS, scale 1-10) and to identify features associated with either a more severe or less severe estimated disease severity. This single-center observational cohort study included 517 medical patients visiting the ED. The instrument was prospectively validated.. The predictive value of each feature for the clinical impression was assessed using multivariate linear regression analyses to adjust for potential confounders and validated in the infection group.

RESULTS

The CIS at the ED was associated with ICU admission (OR 1.67 [1.37-2.03], p < 0.001), in-hospital mortality (OR 2.25 [1.33-3.81], p < 0.001), and 28-day mortality (OR 1.33 [1.07-1.65], <0.001). Dry mucous membranes, eye glance, red flags during physical examination, results of arterial blood gas analysis, heart and respiratory rate, oxygen modality, triage urgency, and increased age were associated with a higher estimated disease severity (CIS). On the other hand, behavior of family, self-estimation of the patient, systolic blood pressure, and Glascow Coma Scale were associated with a lower estimated disease severity (CIS).

CONCLUSION

We identified several features that were associated with the clinical impression of health care professionals in the ED. Translating the subjective features and objective measurements into quantifiable parameters may aid the development of a novel triage tool to identify patients at risk of deterioration in the ED.

Prospective External Validation of a Commercial Model Predicting the Likelihood of Inpatient Admission From the Emergency Department

STUDY OBJECTIVE

Early notification of admissions from the emergency department (ED) may allow hospitals to plan for inpatient bed demand. This study aimed to assess Epic's ED Likelihood to Occupy an Inpatient Bed predictive model and its application in improving hospital bed planning workflows.

METHODS

All ED adult (18 years and older) visits from September 2021 to August 2022 at a large regional health care system were included. The primary outcome was inpatient admission. The predictive model is a random forest algorithm that uses demographic and clinical features. The model was implemented prospectively, with scores generated every 15 minutes. The area under the receiver operator curves (AUROC) and precision-recall curves (AUPRC) were calculated using the maximum score prior to the outcome and for each prediction independently. Test characteristics and lead time were calculated over a range of model score thresholds.

RESULTS

Over 11 months, 329,194 encounters were evaluated, with an incidence of inpatient admission of 25.4%. The encounter-level AUROC was 0.849 (95% confidence interval [CI], 0.848 to 0.851), and the AUPRC was 0.643 (95% CI, 0.640 to 0.647). With a prediction horizon of 6 hours, the AUROC was 0.758 (95% CI, 0.758 to 0.759,) and the AUPRC was 0.470 (95% CI, 0.469 to 0.471). At a predictive model threshold of 40, the sensitivity was 0.49, the positive predictive value was 0.65, and the median lead-time warning was 127 minutes before the inpatient bed request.

CONCLUSION

The Epic ED Likelihood to Occupy an Inpatient Bed model may improve hospital bed planning workflows. Further study is needed to determine its operational effect.

The Value of the First Clinical Impression as Assessed by 18 Observations in Patients Presenting to the Emergency Department

The first clinical impression of emergency patients conveys a myriad of information that has been incompletely elucidated. In this prospective, observational study, the value of the first clinical impression, assessed by 18 observations, to predict the need for timely medical attention, the need for hospital admission, and in-hospital mortality in 1506 adult patients presenting to the triage desk of an emergency department was determined. Machine learning models were used for statistical analysis. The first clinical impression could predict the need for timely medical attention [area under the receiver operating characteristic curve (AUC ROC), 0.73; p = 0.01] and hospital admission (AUC ROC, 0.8; p = 0.004), but not in-hospital mortality (AUC ROC, 0.72; p = 0.13). The five most important features informing the prediction models were age, ability to walk, admission by emergency medical services, lying on a stretcher, breathing pattern, and bringing a suitcase. The inability to walk at triage presentation was highly predictive of both the need for timely medical attention (p < 0.001) and the need for hospital admission (p < 0.001). In conclusion, the first clinical impression of emergency patients presenting to the triage desk can predict the need for timely medical attention and hospital admission. Important components of the first clinical impression were identified.

Interventions to improve consultations in the emergency department: A systematic review

OBJECTIVES

Emergency department (ED) consultations with specialists are necessary for safe and effective patient care. Delays in the ED consultation process, however, have been shown to increase ED length of stay (LOS) and contribute to ED crowding. This review aims to describe and evaluate the effectiveness of interventions to improve the ED consultation process.

METHODS

Eight primary literature databases and the gray literature were searched to identify comparative studies assessing ED-based interventions to improve the specialist consultation process. Two independent reviewers identified eligible studies, assessed study quality, and extracted data. Individual or pooled meta-analysis for continuous outcomes were calculated as mean differences (MDs) with 95% confidence intervals (CIs) using a random-effects model was conducted.

RESULTS

Thirty-five unique comparative intervention studies were included. While the interventions varied, four common components/themes were identified including interventions to improve consultant responsiveness (n = 11), improve access to consultants in the ED (n = 9), expedite ED consultations (n = 8), and bypass ED consultations (n = 7). Studies on interventions to improve consult responsiveness consistently reported a decrease in consult response times in the intervention group with percent changes between 10% and 71%. Studies implementing interventions to improve consult responsiveness (MD -2.55, 95% CI -4.88 to -0.22) and interventions to bypass ED consultations (MD -0.99, 95% CI -1.43 to -0.56) consistently reported a decrease in ED LOS; however, heterogeneity was high (I²  = 99%). Evidence on whether any of the interventions were effective at reducing the proportion of patients consulted or subsequently admitted varied.

CONCLUSIONS

The various interventions impacting the consultation process were predominately successful in reducing ED LOS, with evidence suggesting that interventions improving consult responsiveness and improving access to consultants in the ED also improve consult response times. Health care providers looking to implement interventions to improve the ED consultation process should identify key areas in their setting that could be targeted.

Development and Validation of Machine Learning Models to Predict Admission From Emergency Department to Inpatient and Intensive Care Units

STUDY OBJECTIVE

This study aimed to develop and validate 2 machine learning models that use historical and current-visit patient data from electronic health records to predict the probability of patient admission to either an inpatient unit or ICU at each hour (up to 24 hours) of an emergency department (ED) encounter. The secondary goal was to provide a framework for the operational implementation of these machine learning models.

METHODS

Data were curated from 468,167 adult patient encounters in 3 EDs (1 academic and 2 community-based EDs) of a large academic health system from August 1, 2015, to October 31, 2018. The models were validated using encounter data from January 1, 2019, to December 31, 2019. An operational user dashboard was developed, and the models were run on real-time encounter data.

RESULTS

For the intermediate admission model, the area under the receiver operating characteristic curve was 0.873 and the area under the precision-recall curve was 0.636. For the ICU admission model, the area under the receiver operating characteristic curve was 0.951 and the area under the precision-recall curve was 0.461. The models had similar performance in both the academic- and community-based settings as well as across the 2019 and real-time encounter data.

CONCLUSION

Machine learning models were developed to accurately make predictions regarding the probability of inpatient or ICU admission throughout the entire duration of a patient's encounter in ED and not just at the time of triage. These models remained accurate for a patient cohort beyond the time period of the initial training data and were integrated to run on live electronic health record data, with similar performance.