Difference among frailty assessment tools in predicating postoperative prognosis of elderly patients with mild traumatic brain injury

Abstract

Objectives: Mild Traumatic Brain Injury (mTBI) is quite prevalent in the elderly population, and the authors performed a retrospective analysis regarding the predictive value of frailty assessing tools regarding the prognosis of elderly mTBI patients.

Methods: All the patients underwent assessment of frailty upon admission using five tools including Frailty Phenotype (FP), FRAIL Scale (FS), Edmonton Frailty Scale (EFS), Groningen Frailty Indicator (GFI), and Clinical Frailty Scale (CFS). The predicting potential of tools was analyzed against the prognosis defined by the extended Glasgow Outcome Scale (GOSE).

Results: The incidence of frailty in elderly patients varies widely among the tools. Multivariate logistic regression analysis showed that only frail conditions defined by FP (p-value = 0.014) and FS (p-value = 0.004) could be employed for predicting unfavorable prognosis defined by GOSE, while frailty defined by CFS (p-value = 0.683), EFS (p-value = 0.301) and GFI (p-value = 0.925) could not. The ROC further showed that FP (AUC = 73.2 %) and FS (AUC = 76.2 %) had moderate power in predicting unfavorable conditions, while CFS (AUC = 46.1 %), EFS (AUC = 55.6 %), and GFI (AUC = 51.5 %) only had low or even no power.

Conclusions: FP and FS could be used to predict the unfavorable prognosis associated with mTBI in the elderly population.

Keywords: Clinical frailty scale; Elderly population; Frailty index; Glasgow outcome scale; Mild Traumatic Brain Injury.