# The Importance of Frailty in Determining Survival After Intensive Care

## TL;DR

Frailty scores help predict survival and recovery after intensive care as effectively as traditional medical scores.

## Contribution

The study shows that frailty scores can improve mortality prediction when combined with acute illness scores.

## Key findings

- Frailty scores (CFS) predict mortality as accurately as APACHE II and SAPS II systems.
- Adding CFS to acute scoring systems improves mortality prediction accuracy.
- Survivors' mental health one year post-discharge is similar to the general population.

## Abstract

Background: Estimating survival and long-term quality of life after intensive care has been a crucial bioethical endeavour in recent decades to improve end-of-life decision-making. Scientific studies have also shown that patient frailty influences survival, but only a few long-term data are available. Methods: We conducted a prospective observational study at the Department of Anaesthesiology and Intensive Care of Semmelweis University, Hungary, to investigate the association between physical status on admission, the chance of survival, and the long-term quality of life of the patient. We recorded the pre-admission frailty score (Clinical Frailty Scale), APACHE II, and SAPS II scores on admission. The first follow-up was 3 months after discharge when the quality of life of the patient was assessed using the EQ5-D questionnaire. During the second follow-up one year later, we recorded the EQ5-D, Mini-Mental Test, and the Beck Depression Inventory scales. Results: Our study demonstrated that the ROC analysis of predicted overall mortality based on CFS score is similar in accuracy to that of predicted mortality by APACHE II and SAPS II point systems. The multivariate logistic regression calculations show that the best performing of the three independent variables is the SAPS II estimator (78.5%), but the estimators of both acute condition scoring systems (APACHE and SAPS) can be improved (79.5% vs. 84%) when taking into account the CFS value. The prevalence of mood and mental disorders among patients who survived one year was not different from that of the general population. Conclusions: The physiological scoring systems examined are all suitable for estimating the risk of overall mortality. The CFS shows similar efficacy and appears to be additive in value, with scales describing the severity of acute illness, which are indicative of the chronic condition of the patient.