Predictors of Delayed Emergency Department Throughput Among Blunt Trauma Patients

A Pilot Machine Learning Study Using Trauma Admission Data to Identify Risk for High Length of Stay

INTRODUCTION

Trauma patients have diverse resource needs due to variable mechanisms and injury patterns. The aim of this study was to build a tool that uses only data available at time of admission to predict prolonged hospital length of stay (LOS).

METHODS

Data was collected from the trauma registry at an urban level one adult trauma center and included patients from 1/1/2014 to 3/31/2019. Trauma patients with one or fewer days LOS were excluded. Single layer and deep artificial neural networks were trained to identify patients in the top quartile of LOS and optimized on area under the receiver operator characteristic curve (AUROC). The predictive performance of the model was assessed on a separate test set using binary classification measures of accuracy, precision, and error.

RESULTS

2953 admitted trauma patients with more than one-day LOS were included in this study. They were 70% male, 60% white, and averaged 47 years-old (SD: 21). 28% were penetrating trauma. Median length of stay was 5 days (IQR 3-9). For prediction of prolonged LOS, the deep neural network achieved an AUROC of 0.80 (95% CI: 0.786-0.814) specificity was 0.95, sensitivity was 0.32, with an overall accuracy of 0.79.

CONCLUSION

Machine learning can predict, with excellent specificity, trauma patients who will have prolonged length of stay with only physiologic and demographic data available at the time of admission. These patients may benefit from additional resources with respect to disposition planning at the time of admission.

Hispanic Farmers Experience Shorter EMS Response Times but Longer Emergency Department Length of Stay Following Occupational Injuries

BACKGROUND

Agriculture has the highest rate of fatal injuries by sector. Hispanic workers also experience more fatal work injuries than every other minority group combined. Pre-hospital and initial trauma evaluation represent an important marker to understand the impact of a trauma system. We sought to investigate whether Hispanic agricultural workers in the United States (US) experience disparities following traumatic occupational injuries in terms of pre-hospital and emergency department care.

METHODS

We retrospectively analyzed the National Trauma Data Bank from 2012-2016 to understand differences between Hispanic and non-Hispanic farmers in Emergency Medical Services (EMS) response and transport times (minutes), transport mode, transfer rates, presentation to University or Level I trauma hospitals, Injury Severity Scores (ISS), length of stay (LOS) in the emergency department (ED, minutes) or hospital (days), need for the operating room (OR), admittance to the Intensive Care Unit (ICU), and mortality.

RESULTS

A total of 6,161 farmers were included in our analyses (median age 47 years, females 7.0%). Multivariable analyses indicate differences regarding EMS response, EMS transport, and LOS in the ED. Rates of admission to the ICU, surgical operations, days on a ventilator, discharge from the hospital with supportive care, and mortality did not differ between groups.

CONCLUSIONS

Non-Hispanic patients have longer median EMS response and total transport times. Hispanic patients have longer median LOS in the ED. However, the lack of significant differences in management variables other than EMS times and ED LOS indicate an equitable delivery of trauma care once patients were transferred from the ED.

Emergency Department Boarding of Non-Trauma Patients Adversely Affects Trauma Patient Length of Stay

Introduction Emergency Department (ED) boarding delays initiation of time-sensitive protocols for trauma patients and makes them susceptible to increased mortality and morbidity. In this study, we compared the ED boarding times of non-trauma patients and ED length of stay (LOS) of trauma patients. Methods This was a single-center retrospective cohort study in a Level 1 trauma center. The median boarding time among non-trauma patients and ED LOS among trauma patients was determined by month between the period of April 2018 to March 2019. Linear regression and Pearson correlation coefficient were used to express the magnitude and direction of the relationship between these two variables. Results During the study period, the mean number of non-trauma patients admitted in our ED per month was 1,154 and trauma patients was 89. The mean of the median boarding time per month for non-trauma patients was 76 minutes, and the mean of the median ED LOS per month for trauma patients was 198 minutes. There was a significant positive correlation between boarding time for non-trauma patients and ED LOS for trauma patients (Pearson correlation coefficient: 0.73; p = 0.007). Conclusion The long boarding times for non-trauma patients is associated with ED LOS for trauma patients, indicating that the total patient volume in the hospital contributes to the trauma patient's stay in the ED. Thus, ED LOS of trauma patients can be minimized by improving overall ED and hospital flow, including non-trauma patients.