The Impact of Frailty on Traumatic Brain Injury Outcomes: An Analysis of 691 821 Nationwide Cases

Abstract

BACKGROUND:Frailty, a decline in physiological reserve, prognosticates poorer outcomes for several neurosurgical conditions. However, the impact of frailty on traumatic brain injury outcomes is not well characterized.OBJECTIVE:To analyze the association between frailty and traumatic intracranial hemorrhage (tICH) outcomes in a nationwide cohort.METHODS:We identified all adult admissions for tICH in the National Trauma Data Bank from 2007 to 2017. Frailty was quantified using the validated modified 5-item Frailty Index (mFI-5) metric (range = 0-5), with mFI-5 ≥2 denoting frailty. Analyzed outcomes included in-hospital mortality, favorable discharge disposition, complications, ventilator days, and intensive care unit (ICU) and total length of stay (LOS). Multivariable regression assessed the association between mFI-5 and outcomes, adjusting for patient demographics, hospital characteristics, injury severity, and neurosurgical intervention.RESULTS:A total of 691 821 tICH admissions were analyzed. The average age was 57.6 years. 18.0% of patients were frail (mFI-5 ≥ 2). Between 2007 and 2017, the prevalence of frailty grew from 7.9% to 21.7%. Frailty was associated with increased odds of mortality (odds ratio [OR] = 1.36, P < .001) and decreased odds of favorable discharge disposition (OR = 0.72, P < .001). Frail patients exhibited an elevated rate of complications (OR = 1.06, P < .001), including unplanned return to the ICU (OR = 1.55, P < .001) and operating room (OR = 1.17, P = .003). Finally, frail patients experienced increased ventilator days (+12%, P < .001), ICU LOS (+11%, P < .001), and total LOS (+13%, P < .001). All associations with death and disposition remained significant after stratification for age, trauma severity, and neurosurgical intervention.CONCLUSION:For patients with tICH, frailty predicted higher mortality and morbidity, independent of age or injury severity.

Commentary: The Impact of Frailty on Traumatic Brain Injury Outcomes: An Analysis of 691 821 Nationwide Cases

“…Tang et al 1 should be congratulated for an excellent study examining a simple characterization for the baseline health of patients with neurotrauma. Our specialty needs these to evaluate the relatively rare conditions which we treat using large national data sets.…”

“…Tang et al 1 also noted the nearly trebling of the frailty rate among patients with tICH from 2007 to 2017. Although much of this is likely due to an aging and increasingly frail population (living longer means living longer with comorbidities ), the authors acknowledge that much of this increase is also due to better reporting of comorbidities within the data set itself.…”

“…Furthermore, risk-adjustment statistics can be manipulated, with more attention on coding comorbidities, to make patient care seem to be improving when it has, in fact, remained stagnant. 11 Tang et al 1 noted another coding challenge in the primary outcome when the NTDB transitioned from International Classification of Diseases (ICD)-9 to ICD-10 that left many of the tICH subtypes unclearly delineated and difficult to parse.…”

“…Our specialty needs these to evaluate the relatively rare conditions which we treat using large national data sets. We do need to keep in mind their limitations and use appropriate cautions when interpreting results, which Tang et al 1 do with eloquence.…”

“…The mFI-5's simplicity allows for use with the NTDB and likely use with other large data sets, although not investigated here. The longer mFI-11 could not be used by Tang et al 1 because the components were not present in NTDB. Other frailty assessments involve questionnaires or other clinical evaluations that are not available in larger data sets, leaving the optimal frailty measure for large administrative data sets still open for debate.…”