Abstract
Background: We sought to quantify the association between state trauma funding and (1) in-hospital mortality and (2) transfers of injured patients. Methods: We conducted an observational cross-sectional study of states with publicly available trauma funding data. We analyzed in-hospital mortality using linked data from the Nationwide Inpatient Sample (NIS), American Hospital Association (AHA) Annual Survey, and these State Department of Public Health trauma funding data. Results: A total of 594,797 injured adult patients were admitted to acute care hospitals in 17 states. Patients in states with >$1.00 per capita state trauma funding had 0.82 (95 % CI: 0.78–0.85, p < 0.001) decreased adjusted odds of in-hospital mortality compared to patients in states with less than $1.00 per capita state trauma funding. Conclusions: Increased state trauma funding is associated with decreased adjusted in-hospital mortality.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 133-139 |
| Number of pages | 7 |
| Journal | American journal of surgery |
| Volume | 229 |
| DOIs | |
| State | Published - Mar 2024 |
Funding
This work was supported by the American Association for the Surgery of Trauma and the American College of Surgeons . However, trauma systems development has met many challenges due to a lack of funding. In 1981, the Omnibus Budget Reconciliation Act withdrew federal funding dedicated to trauma systems. The Act consolidated funding into a broader state preventive health block grant, shifting the responsibility to state governments to decide whether and how to fund their trauma systems. 11–13 Up until then, under the EMS Systems Act of 1973, more than $300 million in federal grants had been distributed to establish 300 EMS and trauma systems across the U.S.13 Since then, efforts to institutionalize national leadership on trauma funding and development have been unsuccessful. The Trauma Care Systems Planning and Development Act was passed in 1990 but eliminated five years later. The Healthcare Resources and Services Administration's (HRSA) Trauma/EMS Systems Program was created in 2001 but eliminated five years later.12 These policy changes have resulted in a disjointed process of trauma systems development with little federal guidance and great variation in state funding approaches.14,15This study examined the association between the amount of state trauma funding and (1) distribution of injured patients across hospitals designated as trauma centers (Levels I-IV) and non-trauma centers; (2) adjusted in-hospital mortality of injured patients; and (3) adjusted rates of interfacility transfers to a second acute care hospital versus discharges to home. We hypothesized that increased state trauma funding per capita would be associated with (1) broader distribution of injured patients across trauma centers and non-trauma centers; (2) lower adjusted in-hospital mortality of injured patients; and (3) fewer interfacility of injured patients to a second acute care hospital. This was the first study to test whether amount of state trauma funding was associated with injured patient mortality across a broad range of hospitalized injured patients in many states.The process measure of interest was distribution of injured patients across hospitals designated as trauma centers (Levels I-IV) and non-trauma centers. Trauma center level designations in the AHA Annual Survey were based on state designations. Distribution of injured patients hospitalized at Level I, Level II, Level III, Level IV, and non-trauma centers was characterized by calculating the number and percentage of patients within each per capita state trauma funding quartile. The same calculations were done to determine the distribution of a subset of seriously injured patients [Injury Severity Score (SS) > 15], who should be cared for at Level I or II trauma centers per American College of Surgeons recommendations.2,26,27A sensitivity analysis of adjusted in-hospital mortality was conducted among injured patients hospitalized in the only five states with known trauma system funding that participated in the NIS and AHA Annual Survey (Arkansas, Colorado, Michigan, Minnesota, and Washington), while controlling for age, sex, ISS, weighted ECI, trauma center level designation, teaching hospital status, and medical school affiliation. One large diverse state with known $0.00 of per capita trauma funding (New York) was included as the reference. This analysis was done to exclude confounding from states that had only trauma care funding and no trauma system funding. State trauma care funding could theoretically act as a perverse incentive to keep seriously injured patients at non-trauma center hospitals where they experience higher rates of mortality.This study had several important limitations. First, data on state trauma funding was not easily accessible. This may have introduced a bias towards states that have accessible trauma funding data. Nonetheless, we attempted to minimize this bias with personal communications with individuals working in trauma systems around the country to minimize sampling bias. Furthermore, our findings are similar to those by Lin et al.15 Second, it was often difficult to definitively characterize how funding was allocated. While some states published reports breaking down every dollar of trauma funding, some states only reported a single number and did not detail where each dollar was directed or how it was obtained. In some cases, there were also some minor differences between appropriations (the amount of funding available) and expenditures (the amount of funding spent), making it challenging to narrow state trauma funding to a single number for each state. We were able to gather data for 17 states and verify the data by cross-referencing legislation, budget reports, trauma system planning documents, open records requests, and information from direct communication with public health officials.Third, because patient-level data were obtained from the NIS, this study also did not account for injured patients who were treated in the Emergency Department only and were not admitted to a hospital. As a result, this analysis was not representative of all injured patients, and the results may be biased towards more severely injured patients with higher mortality rates. However, patients who are only treated in the Emergency Department are traditionally patients who are either the most critically injured and unlikely to survive regardless of care, or the least critically injured and unlikely to experience in-hospital mortality. Fourth, since NIS data are primarily captured for administrative purposes, they also do not always reflect the clinical complexity of each patient, possibly resulting in omitted variable bias.36 We attempted to control for clinical complexity by (a) adjusting for patient comorbidities by calculating the weighted ECI29,30 and b) adjusting for injury severity by including patient ISS31 in all multivariable analyses. Fifth, the AHA Annual Survey data was also subject to the limitations of survey data, including but not limited to response rate, responder fatigue, and inaccuracies in reporting. Sixth, the analytical dataset was restricted to the 2010-11 NIS because this was the last year the NIS included the AHA ID hospital identifier required to link NIS patient-level data to AHA Annual Survey hospital-level data. Linking these two datasets allowed us to include trauma center level designation, an essential co-variate of our analyses of the distribution of injured patients. The observed trends may not be generalizable to current day outcomes and funding. However, as part of our data collection, we collected all state trauma funding data since 2010 and found that there has been little to no change during the past decade in the allocation of state trauma funding. We have also found similar associations between state trauma funding and mortality using 2016–2017 data from the Healthcare Cost and Utilization Project (HCUP) State Emergency Department Database (SEDD) and State Inpatient Database (SID).37 Seventh, there is conflation between emergency interfacility transfers from emergency department to emergency department and interfacility transfers from inpatient to inpatient in the NIS. We could not control for these two different types of transfered patients in the current study because the NIS does not contain emergency department data or transfer time, nor can we link NIS patient encounters. Finally, trauma care is a very complex process that depends on trauma system structures, transport time and distance to the nearest trauma center, population density, trauma center density, as well as regional protocols and agreements. These analyses would require patient geographic data that the NIS does not provide.This work was supported by the American Association for the Surgery of Trauma and the American College of Surgeons.
Keywords
- Acute care
- Mortality
- Trauma
- Trauma funding
ASJC Scopus subject areas
- Surgery
