We used data from the North Carolina Trauma Registry (NCTR), a statewide registry and cooperative effort between eighteen North Carolina hospitals, including all 17 North Carolina trauma centers (6 Level I, 3 Level II, and 8 Level III hospitals) and the North Carolina Office of Emergency Medical Services (NCOEMS)(23, 24). This registry, which has been in place since 1987, collects near real-time information using standardized data definitions based off of the National Trauma Registry of the American College of Surgeons and designated NCTR chart abstractors(23, 24). The NCTR includes all hospitalizations where a patient is diagnosed with a traumatic injury (ICD-10-CM: S00-S99, T07, T14, T20-T28, T30-T32, T71, T79.A1-T79.A9), and is admitted to the hospital, taken to the operating room from the emergency department, transferred, or dies due to their injury. Unplanned readmissions within 30 days of the initial injury are also included.
For this study, we included all trauma hospitalizations that occurred between January 1, 2019 and December 31, 2020. To account for variation between weekday and weekend hospitalization rates, we calculated the weekly hospitalization rates for traumatic injuries per 1,000,000 North Carolina residents between January 6, 2019, and December 26, 2020. Admissions that occurred during partial weeks (January 1–5, 2019 and December 27–31, 2020) were excluded from modeling to avoid introducing bias due to underestimation (486 and 393 hospitalizations, respectively; in 2019 there were an average of 682 [SD 61.9] trauma hospitalizations per week). North Carolina population counts for 2019 were obtained from the North Carolina Office of State Budget and Management(25) and were used for both 2019 and 2020 hospitalization rate calculations.
Trauma admissions were classified by injury intent and mechanism into four categories – assault, self-inflicted, unintentional MVC (including MVC-bicyclist and MVC-pedestrian injuries), and unintentional non-MVC – using the ICD-10-CM code framework from the National Center for Health Statistics and National Center for Injury Prevention and Control(26).
Data on race and ethnicity were used to categorize hospitalized patients as non-Hispanic Black/African American, Hispanic/Latino, non-Hispanic White, and non-Hispanic other race. Other race included American Indian (n = 513 hospitalizations), Asian (n = 596 hospitalizations), Pacific Islander (n = 87 hospitalizations), multiracial (n = 202 hospitalizations), and those who listed “other” race (n = 1,048 hospitalizations). Race and ethnicity were self-reported by the patient (or family member) if they were present and capable; otherwise, it was based on staff designation in the electronic medical record.
COVID-related policies of interest included: U.S. declaration of a public health emergency (1/31/2020), the North Carolina statewide Stay-at-Home order (3/30/2020), an initial lifting of the Stay-at-Home order with restrictions (Phase 2: Safer-at-Home, 5/22/2020), and the further lifting of Stay-at-Home restrictions (Phase 2.5: Safer-at-Home, 9/4/2020), Supplementary Table 1. Policies were assigned to the week of their effective date. Other statewide executive orders that were not included in analyses were North Carolina declaring a state of emergency, statewide closure of K-12 public schools, Phase 1 and Phase 3 of lifting the statewide Stay-at-Home orders, and the modified Stay-at-Home order issued before the 2020 holidays. These orders were not included in analyses because either the order made relatively small changes to existing orders (e.g., Phase 1 lifting of Stay-at-Home orders) or it occurred within several weeks of a prior order that we believed would be more salient (e.g., North Carolina declaring a state of emergency).
Differences in patient demographics and clinical characteristics among patients admitted for traumatic injuries between 2019 and 2020 were compared using standardized differences. An absolute difference > 0.20 was considered meaningful.
We conducted a natural experiment using an interrupted-time series design and segmented linear regression(27, 28). Using ordinary least squares, we conducted injury intent and mechanism-specific segmented linear regression models to estimate the trend in trauma hospitalization rates between each pair of interruptions. Our models did not include parameters for level changes (i.e., intercept changes) to focus our analysis on a priori-hypothesized gradual changes in injury hospitalization rates. To reduce error in our model, we used a transformed cosine periodic function to control for potential seasonal fluctuations in hospitalization rates(29). To account for autocorrelation over time, we used Durbin-Watson tests (α = 0.05) to specify autoregressive parameters in our models for lags up to 60 weeks. Similar methods were used to estimate race/ethnicity-specific and age and sex-specific hospitalization rates. Due to low overall rates, race/ethnicity and age/sex-stratified models for self-inflicted injuries were not performed.
All analyses were performed using SAS version 9.4 (SAS Inc., Cary, North Carolina). This study was deemed exempt by the University of North Carolina (IRB# 20-2117) and National Institutes of Health (IRB# 000330) Institutional Review Boards.