Is frailty a prognostic factor for critically ill elderly patients?

Aging Clin Exp Res. 2017 Apr;29(2):247-255. doi: 10.1007/s40520-016-0557-y. Epub 2016 Mar 22.

Abstract

Aim: The study aimed to investigate the effects of frailty on clinical outcomes of patients in an intensive care unit (ICU).

Methods: In this prospective study, 122 patients (59 frail, 37 pre-frail and 26 robust) were included. A frailty index (FI) derived from comprehensive geriatric assessment parameters was used for the evaluation. The FI score of ≤0.25 was considered as robust, 0.25-0.40 as pre-frail and >0.40 as frail. The prognostic effects of FI were investigated and FI and APACHE II and SOFA scores, the prognostic scores using in ICU, were compared.

Results: Median age of the patients was 71 years old and 50.8 % were male. ICU mortality rate and median length of stay (LOS) were 51.6 % and 8 days (min-max: 1-148), respectively. ICU mortality was higher (69.2, 56.8 and 40.7 %, respectively, p = 0.040) and median overall survival was lower in frail group compared to pre-frail and robust subjects (23, 31 and 140 days, p = 0.013, respectively). Long term mortality over 3 and 6 months in frail patients were 80.8 and 84.6 %, respectively and significantly higher than others. Multivariate analysis showed that LOS in ICU (HR 1.067, 95 % CI 1.021-1.114), SOFA score (HR 1.272, 95 % CI 1.096-1.476) and FI (HR 39.019, 95 % CI 1.235-1232.537) were the independent correlates for ICU mortality (p = 0.004, p = 0.002 and 0.038, respectively). There was a weak but statistically significant positive correlation between APACHE II and FI scores (r = 0.190, p = 0.036).

Conclusions: FI may be used as a predictor for the evaluation of elderly patients' clinical outcomes in ICUs.

Keywords: APACHE II; Elderly; Frailty; Intensive care unit; SOFA score.

MeSH terms

  • APACHE
  • Aged
  • Critical Illness / epidemiology*
  • Female
  • Frail Elderly / statistics & numerical data*
  • Geriatric Assessment / methods
  • Humans
  • Intensive Care Units / statistics & numerical data
  • Length of Stay / statistics & numerical data
  • Male
  • Mortality
  • Prognosis
  • Prospective Studies
  • Research Design
  • Statistics as Topic
  • Turkey / epidemiology