Analysis of factors impacting length of stay in thermal and inhalation injury

Abstract

ObjectivesSeveral studies have analyzed single or combinations of variables for impact on length of stay (LOS) in thermally-injured patients. The objective of this study was to evaluate a multitude of established variables and potentially identify novel variables associated with LOS in a single study. MethodsThis two-year, retrospective study included all patients admitted to the burn center between January 2015 and December 2016. Exclusions included death during admission, lack of thermal or inhalation injury, age less than 18 years, readmission(s), and if pregnant or incarcerated. Baseline demographics and pertinent data were collected using electronic medical records. Regression analysis was used to determine the most predictive variables. ResultsSix hundred twenty-nine patients were admitted during the inclusion period and 354 patients remained for analysis after exclusion. Univariable analysis revealed 32 variables significantly associated with LOS. Using multivariable regression, the best-fit baseline demographic model included: percent total body surface area (TBSA) injured, lower/middle socioeconomic status, clotting disorders, anemia, admission serum creatinine, and percent third degree injured (r2 = 0.557). The best-fit combined model (incorporating baseline demographics and early in-hospital variables) included: acute kidney injury, infection and received vasopressor(s), percent TBSA injured, admission serum ethanol level, maximum C-reactive protein, and maximum total bilirubin (r2 = 0.828). ConclusionsThere are multiple factors associated with the increased LOS seen in patients with thermal and inhalation injury. This study confirmed and identified novel factors not previously discussed in the literature that were significantly associated with LOS. Expansion of the data submitted to the National Burn Repository and the Burn Quality Improvement Program may be warranted. This study confirms claims from previous studies on inadequacy of current data submitted for benchmarking and under-reimbursement for the care of such a complex population.