C-reactive protein to albumin ratio in critical care settings

Abstract

Purpose

Compare the performance of the C-reactive protein to albumin ratio (CAR) in predicting in-hospital death between patients with and without COVID-19.

Materials and methods

This prospective cohort study included critically ill patients with and without COVID-19. CAR was obtained at admission and on Day 3, along with fluid balance and other predictive scores. The primary outcome was all-cause in-hospital death, and the Youden index was used to determine CAR cut-off points.

Results

We included the data of 176 patients; 85 had COVID-19. All-cause in-hospital death was 31.3%. The CAR cut-off points for patients without COVID-19 were 28.5 mg/g and 31.6 mg/g at admission and on Day 3, respectively, and 13.5 mg/g and 7.4 mg/g for COVID-19 patients. Day 3 CAR was a better predictor of all-cause in-hospital death than CAR at admission in both groups. Non-survivors in the COVID and non-COVID groups had higher positive fluid balance rates than survivors.

Conclusions

The CAR predicted in-hospital death in both groups, with different cut-off points. CAR on Day 3 had better diagnostic performance than that at admission. Cumulative fluid balance may influence CAR kinetics. An optimal CAR cut-off point for critically ill patients, optimal measuring time, and impact of fluids therapy still need to be established.