Dataset Information

Impact of the Covid-19 pandemic on rates of emergency department utilization and hospital admission due to general surgery conditions.

ABSTRACT:

Background

Recent literature reports a decrease in healthcare-seeking behaviours by adults during the Covid-19 pandemic. Given that emergency general surgery (GS) conditions are often associated with high morbidity and mortality if left untreated, the objective of this study was to describe and quantify the impact of the Covid-19 pandemic on rates of emergency department (ED) utilization and hospital admission due to GS conditions.

Methods

This cohort study involved the analysis of an institutional database and retrospective chart review. We identified adult patients presenting to the ED in a network of three teaching hospitals in Montreal, Canada during the first wave of the Covid-19 pandemic (March13-May13, 2020) and a control pre-pandemic period (March13-May13, 2019). Patients with GS conditions were included in the analysis. ED utilization rates, admission rates and 30-day outcomes were compared between the two periods using multivariate regression analysis.

Results

During the pandemic period, 258 patients presented to ED with a GS diagnosis compared to 351 patients pre-pandemically (adjusted rate ratio (aRR) 0.75; p < 0.001). Rate of hospital admission during the pandemic was also significantly lower (aRR = 0.77, p < 0.001). Patients had a significantly shorter ED stay during the pandemic (adjusted mean difference 5.0 h; p < 0.001). Rates of operative management during the pandemic were preserved compared to the pre-pandemic period. There were no differences in 30-day complications (adjusted odds ratio (aOR) 1.46; p = 0.07), ED revisits (aOR 1.10; p = 0.66) and (re)admissions (aOR 1.42; p = 0.22) between the two periods.

Conclusion

There was a decrease in rates of ED utilization and hospital admissions due to GS conditions during the first wave of the Covid -19 pandemic; however, rates of operative management, complications and healthcare reutilization were unchanged. Although our findings are not generalizable to patients who did not seek healthcare, it was possible to successfully uphold institutional standards of care once patients presented to the ED.

SUBMITTER: Balvardi S

PROVIDER: S-EPMC8722748 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ObjectiveTo predict hospital admission at the time of ED triage using patient history in addition to information collected at triage.MethodsThis retrospective study included all adult ED visits between March 2014 and July 2017 from one academic and two community emergency rooms that resulted in either admission or discharge. A total of 972 variables were extracted per patient visit. Samples were randomly partitioned into training (80%), validation (10%), and test (10%) sets. We trained a series of nine binary classifiers using logistic regression (LR), gradient boosting (XGBoost), and deep neural networks (DNN) on three dataset types: one using only triage information, one using only patient history, and one using the full set of variables. Next, we tested the potential benefit of additional training samples by training models on increasing fractions of our data. Lastly, variables of importance were identified using information gain as a metric to create a low-dimensional model.ResultsA total of 560,486 patient visits were included in the study, with an overall admission risk of 29.7%. Models trained on triage information yielded a test AUC of 0.87 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.88) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on patient history yielded an AUC of 0.86 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.87) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on the full set of variables yielded an AUC of 0.91 for LR (95% CI 0.91-0.91), 0.92 for XGBoost (95% CI 0.92-0.93) and 0.92 for DNN (95% CI 0.92-0.92). All algorithms reached maximum performance at 50% of the training set or less. A low-dimensional XGBoost model built on ESI level, outpatient medication counts, demographics, and hospital usage statistics yielded an AUC of 0.91 (95% CI 0.91-0.91).ConclusionMachine learning can robustly predict hospital admission using triage information and patient history. The addition of historical information improves predictive performance significantly compared to using triage information alone, highlighting the need to incorporate these variables into prediction models.