S. S. Seng1, E. J. Kaufman2, J. Song2, B. Moran3, S. P. Stawicki4, G. Koenig5, M. Timinski6, A. Ratnasekera1,7 1Crozer Chester Medical Center, Surgery, Upland, PA, USA 2Hospital Of The University Of Pennsylvania, Surgery, Philadelphia, PA, USA 3Einstein Medical Center, Surgery, Philadelphia, PA, USA 4St. Luke’s University Health Network, Surgery, Bethlehem, PA, USA 5Thomas Jefferson University, Surgery, Philadelphia, PA, USA 6Geisinger Wyoming Valley, Surgery, Wilkes Barre, PA, USA 7Christiana Care, Surgery, Newark, DE, USA
Introduction: The COVID pandemic and social isolation resulted in major disruptions to social life and access to healthcare. Healthcare providers and the public have raised concern that the effect of social isolation would result in worse population mental health and lead to increased suicides and self-inflicted injuries (SIIs). In this study, we analyzed SIIs treated at Pennsylvania trauma centers with the hypothesis that these would be more common and would have worse outcomes during the COVID pandemic.
Methods: A retrospective review from 2018-2021 was conducted of the Pennsylvania Trauma Outcome Study (PTOS) database examining adult patients (≥18 years) treated for SIIs. Patients were separated into pre-COVID-era (1/1/2018 to 3/31/2020) and COVID-era (4/1/2020 to 3/31/2021). Patient demographics, comorbidities, injury characteristics and severity (ISS), lengths of stay (LOS), and outcomes were tabulated. Specific behavioral health diagnoses were abstracted from available PTOS data. Univariate comparisons were accomplished using chi-squared tests for categorical variables and Kruskal Wallis tests for continuous variables. A multivariable logistic regression (MLR) model was used to determine if COVID-era presentation was a significant predictor of injury mortality.
Results: We identified 1,075 SIIs in the pre-COVID cohort and 482 in the COVID-era cohort. Monthly counts of SIIs did not change significantly between time periods, nor did the proportion of trauma patients whose injuries were due to SIIs. There were no significant differences in demographics or injury characteristics between time periods. Those in the pre-COVID-era cohort had higher rates of pre-existing behavioral health diagnosis than the COVID-era cohort (59.2% vs. 52.3%, p<0.01). The COVID-era cohort had longer hospital LOS (4 vs 3 days, p=0.001), higher rates of cardiac arrest (5.0% vs. 2.7%, p=0.022), unplanned visits to the operating room (2.9% vs. 0.2%, p<0.001) and delirium (1.9% vs. 0.2%, p<0.001) than their pre-COVID-era counterparts. MLR revealed the COVID pandemic was not significantly associated with mortality (Table 1).
Conclusion: Contrary to the concerns raised; the COVID pandemic was not associated with an increase in SIIs. Mortality was equivalent, though complications were slightly more common and length of stay slightly longer during the COVID time period. The mental health consequences of the pandemic remain a widespread public health concern. While the impact of social isolation was not seen with regards to SIIs, future studies should evaluate other markers of mental health to ensure appropriate care for trauma patients.
