Impact of COVID-19 on Emergency Department psychiatric hospitalization patterns in rural South Georgia : a retrospective analysis

Moultrie, GA

Introduction:

The COVID-19 pandemic has significantly impacted healthcare systems worldwide, including psychiatric care in the emergency setting. This study examines psychiatric admission trends before and during the pandemic, analyzing differences in overall admissions, gender distribution, age group variations, and specific psychiatric diagnoses. We aim to determine if the pandemic resulted in significant changes in psychiatric hospitalization patterns in the Colquitt Regional Medical Center (CRMC) emergency department.

Methods:

This retrospective study utilized electronic medical record (EMR) data from CRMC. Admissions from pre-COVID (April 1, 2019 – December 31, 2019) were compared to those during COVID ( April 1, 2020 to December 31, 2020). Statistical analyses included chi-square tests for categorical comparisons, logistic regression for predictive modeling, and Seasonal Autoregressive Integrated Moving Average (SARIMA) time-series modeling for forecasting admission trends. Power analyses were conducted to ensure sufficient statistical strength. Subgroup analyses were also performed to explore potential variations in specific psychiatric diagnoses and other patient demographics.

Results:

A total of 182 psychiatric admissions were recorded pre-COVID compared to 179 during COVID. Chi-square tests showed no significant difference in overall admissions (p = 1.0), but there is a significant difference within gender distribution across age groups (p = 0.026). Younger females (0-18, 19-30) had increased admissions during COVID, whereas female admissions in middle-aged adults (31-45) declined. Male admissions among older adults (61-75) rose during the pandemic.

Regarding diagnoses, "Anxiety disorder, unspecified" remained the most common psychiatric condition across both periods, with an increase in younger age groups (0-30). Suicidal ideation remained prevalent but did not significantly increase during COVID. Alcohol-related disorders were more common in middle-aged and older adult patients (31-60).

Time-series analysis using SARIMA modeling indicated cyclical fluctuations rather than a steady increase or decrease in psychiatric admissions, with predicted peaks in March and May 2021. The logistic regression model found no significant predictors of admission trends based on gender, age, or diagnosis type.

Discussion:

Contrary to expectations, psychiatric admissions did not rise during COVID-19, though sample size limitations may have influenced results. However, mental health-related hospitalizations changed during COVID, with increased admissions among younger females and older males. The cyclical nature of admissions suggests external seasonal, geographic, or sociocultural factors outside of measured parameters that may influence hospital visits.

Power analysis confirmed that the chi-square test for age group and gender had sufficient statistical power (0.997), reinforcing the significance of observed differences. However, logistic regression did not identify strong predictive factors, indicating that other variables (e.g., policy changes, outpatient mental health access) may better explain admission trends.

Future research with larger sample sizes and multi-center collaborations could provide more robust insights into the potential influence of COVID-19 on psychiatric presentations in rural emergency settings. SARIMA modeling forecasting can also be compared to actual data from 2021 in a future study. It is crucial to understand these patterns to improve resource allocation and enhanced preparedness for heightened psychiatric demand in the emergency department in the midst of a pandemic.