# Reevaluating C-Reactive Protein for Perioperative Risk Stratification: The Overlooked Role of Sleep Apnea in Cardiac Surgery Outcomes

## TL;DR

This study shows that sleep apnea has a bigger impact on heart surgery outcomes than the commonly used CRP biomarker, suggesting better preoperative screening for sleep apnea could improve patient care.

## Contribution

The study reveals that sleep apnea severity is a stronger predictor of postoperative complications than CRP, challenging its use in risk stratification.

## Key findings

- OSA severity independently predicted prolonged intubation and new-onset AF.
- CRP had poor predictive power as a standalone biomarker for postoperative outcomes.
- A composite score combining AHI, BMI, and CRP strongly correlated with respiratory support needs.

## Abstract

Background/Objectives: C-reactive protein (CRP) is widely used as a marker of perioperative inflammation, but its predictive value for cardiac surgical outcomes remains uncertain. Obstructive sleep apnea (OSA), a prevalent and underrecognized comorbidity, may independently contribute to postoperative complications through non-inflammatory mechanisms. This study aimed to reevaluate the prognostic role of CRP and determine the clinical impact of OSA severity on postoperative recovery, focusing on new-onset atrial fibrillation (AF), prolonged intubation time, and postoperative CPAP/AIRVO use as indicators of respiratory burden. Methods: In this prospective cohort of 142 elective cardiac surgery patients, preoperative polysomnography and serial CRP measurements were obtained. Multivariable regression, mediation analysis, and propensity score matching (PSM) were performed to evaluate associations between OSA severity, CRP, and perioperative outcomes (AF, intubation time, CPAP/AIRVO use). Results: OSA severity independently predicted prolonged intubation (β = 1.74, p = 0.0019) and new-onset AF (β = 0.85, p = 0.004), even after excluding patients with preexisting arrhythmia. CRP showed poor discriminatory power as a standalone biomarker (AUC for IOT > 14 h = 0.445) and did not mediate OSA–outcome associations. However, CRP > 2.1 mg/dL doubled the odds of moderate-to-severe OSA (OR = 2.05, p = 0.041). A composite score integrating AHI, BMI, and postoperative CRP strongly correlated with postoperative respiratory support (p < 0.0001). Conclusions: OSA exerts a stronger and more consistent influence on perioperative outcomes than CRP, challenging reliance on CRP for risk stratification. Incorporating objective OSA screening and spirometry into preoperative assessment may enhance perioperative risk prediction and guide personalized management strategies.