RESUMEN
INTRODUCTION: Burn injuries remain a significant cause of disability, impacting long term quality-of-life and imposing large costs on our health systems. Readmission is a metric of quality and an important contributor to this economic burden. The association of socioeconomic and insurance status with burn readmission is not well established. The aim of our study is to develop a predictive risk model of factors associated with readmission after burns. METHODS: Using the Healthcare Cost and Utilization Project's 2018 Nationwide Readmission Database, we identified patients ≥18 y of age with burns admitted between January and October 2018. We excluded patients who died during index admission. Our primary outcome was readmission within 60 d postdischarge. We performed a Lasso regression analysis with adaptive selection to generate a predictive model with least deviance using patients' demographics and socioeconomic status, burn location and severity, past medical history, and hospital characteristics. Weighted multiple logistic regression was performed to obtain population estimates of adjusted odds ratios (ORs) of each element in the model. RESULTS: Our cohort included 11,380 burn patients. Of those, 1625 (14.3%) were readmitted and 67% were males. Readmitted patients were older (55 ± 17 versus 49 ± 18, P = 0.0001). Weighted logistic regression for the selected model showed higher odds of readmission for patients with lowest income quartile (OR: 1.19, 95% confidence interval [CI]: 1.04-1.36), Medicare or Medicaid insurance (OR: 1.35, 95% CI: 1.17-1.55), history of psychiatric illness (OR:1.19, 95% CI: 1.02-1.39), diabetes (OR: 1.46, 95% CI: 1.25-1.69), chronic kidney disease (OR: 1.66, 95% CI: 1.30-2.11), chronic obstructive pulmonary disease (OR: 1.55, 95% CI:1.26-1.89), and alcohol use disorder (OR: 1.33, 95% CI: 1.13-1.58). Third degree burns and foot burns had higher OR of readmission (OR: 1.21, 95% CI: 1.38-1.98 and 1.66, 95% CI: 1.02-1.45, respectively), while face and hand burns had lower OR of readmission (OR: 0.77, 95% CI: 0.66-0.90 and 0.84, 95% CI: 0.72-0.98, respectively). CONCLUSIONS: Burn readmissions are multifactorial and directly related to the patient's comorbidities, including markers that reflect barriers to care such as socioeconomic characteristics, as well as the anatomical location of burn injuries. Early identification of these high-risk patients may aid in early intervention, resource allocation, and outreach program development in an attempt to reduce readmission rates and improve outcomes. Future prospective validation of these risk factors is warranted.
RESUMEN
BACKGROUND: Whole blood (WB) transfusion has been shown to improve mortality in trauma resuscitation. The optimal ratio of packed red blood cells (pRBC) to WB in emergent transfusion has not been determined. We hypothesized that a low pRBC/WB transfusion ratio is associated with improved survival in trauma patients. METHODS: We analyzed the 2021 Trauma Quality Improvement Program (TQIP) database to identify patients who underwent emergent surgery for hemorrhage control and were transfused within 4 hours of hospital arrival, excluding transfers or deaths in the emergency department. We stratified patients based on pRBC/WB ratios. The primary outcome was mortality at 24 hours. Logistic regression was performed to estimate odds of mortality among ratio groups compared with WB alone, adjusting for injury severity, time to intervention, and demographics. RESULTS: Our cohort included 17,562 patients; of those, 13,678 patients had only pRBC transfused and were excluded. Fresh frozen plasma/pRBC ratio was balanced in all groups. Among those who received WB (n = 3,884), there was a significant increase in 24-hour mortality with higher pRBC/WB ratios (WB alone 5.2%, 1:1 10.9%, 2:1 11.8%, 3:1 14.9%, 4:1 20.9%, 5:1 34.1%, p = 0.0001). Using empirical cutpoint estimation, we identified a 3:1 ratio or less as an optimal cutoff point. Adjusted odds ratios of 24-hour mortality for 4:1 and 5:1 groups were 2.85 (95% confidence interval [CI], 1.19-6.81) and 2.89 (95% CI, 1.29-6.49), respectively. Adjusted hazard ratios of 24-hour mortality were 2.83 (95% CI, 1.18-6.77) for 3:1 ratio, 3.67 (95% CI, 1.57-8.57) for 4:1 ratio, and 1.97 (95% CI, 0.91-4.23) for 5:1 ratio. CONCLUSION: Our analysis shows that higher pRBC/WB ratios at 4 hours diminished survival benefits of WB in trauma resuscitation. Further efforts should emphasize this relationship to optimize trauma resuscitation protocols. LEVEL OF EVIDENCE: Therapeutic/Care Management; Level III.