Predictors of Discharge Disposition in Older Adults With Burns: A Study of the Burn Model Systems

Older patients with burn injury have a greater likelihood for discharge to nursing facilities. Recent research indicates that older patients discharged to nursing facilities are two to three times as likely to die within a 3-year period relative to those discharged to home. In light of these poor long-term outcomes, we conducted this study to identify predictors for discharge to independent vs nonindependent living status in older patients hospitalized for burns. We retrospectively reviewed all older adults (age 55 years) who were prospectively enrolled in a longitudinal multicenter study of outcomes from 1993 to 2011. Patient, injury, and treatment outcomes data were analyzed. Recognizing that transfer to inpatient rehabilitation may have impacted final hospital discharge disposition: we assessed the likelihood of inpatient rehabilitation stay, based on identified predictors of inpatient rehabilitation. We subsequently performed a logistic regression analysis on the clustered, propensity-matched cohort to assess associations of burn and injury characteristics on the primary outcome of final discharge status. A total of 591 patients aged 55 years were treated and discharged alive from three participating U.S. burn centers during the study period. Mean burn size was 14.8% (SD 11.2%) and mean age was 66.7 years (SD 9.3 years). Ninety-three patients had an inpatient rehabilitation stay before discharge (15.7%). Significant factors predictive of inpatient rehabilitation included a burn >20% TBSA, mechanical ventilation, older age, range of motion deficits at acute care discharge, and study site. These factors were included in the propensity model. Four hundred seventy-one patients (80%) were discharged to independent living status. By matched propensity analysis, older age was significantly associated with a higher likelihood of discharge to nonindependent living (P < .01 in both the 65-74 age group and the oldest age group when compared with the 55-64 age group). Comorbidity (P < .01) and history of alcohol abuse (P < 0.01) were also predictive patient factors. Furthermore, clinical practice variations among the three study sites also constituted a significant factor in discharge disposition (both P < .01 when compared with the index study site). Older age remains an important risk factor for discharge to nonindependent living status, even after accounting for inpatient rehabilitation stay. This analysis, however, reveals significant variations in discharge disposition practices among the three participating study sites. We believe that these variations among burn centers need to be elucidated to better understand discharge disposition status in older patients with burns.