Skip to main content

Carboxyhemoglobin, a reliable diagnosis biomarker for hemolysis in intensive care unit: a retrospective study

Dear Editor,

Hemolytic anemia (HA) is a common condition in intensive care units (ICU) responsible for life-threatening organ failure in severe cases [1]. HA needs urgent treatment initiation, but its diagnosis remains challenging as none of the biological diagnostic parameters, including bilirubin, LDH and haptoglobin, are specific. During hemolysis, free hemoglobin released by red blood cells is catabolized by heme-oxygenase 1, leading to formation of iron, biliverdin and carbon monoxide [2]. Next, carbon monoxide binds to free hemoglobin to form carboxyhemoglobin. Carboxyhemoglobin is routinely measured in ICU and available within a few minutes by CO-oximetry, a point of care testing (GEM® Premier™ 4000, Werfen, Le Pré-Saint-Gervais, France) [3]. Our objective was to evaluate carboxyhemoglobin as a diagnostic tool for HA in adult patients admitted in ICU.

Methods and statistical analysis

We retrospectively analyzed data from patients hospitalized for HA and non-HA in our ICU during an 8-year and 1-year period, respectively. An adjudication committee consisting of three senior experts (from hematology, internal medicine and critical care departments) confirmed final HA and non-HA diagnosis with all available clinical and biological data. Carboxyhemoglobin was measured at ICU admission on arterial blood.

Differences between groups were compared using the Student's t test or Wilcoxon’s test. Correlations were computed using Pearson’s formula. Discrimination performances were assessed by using area under the receiver operating characteristic curve (ROC). Statistical analyses were performed using R software (v 2.12.0;


Overall, 187 patients were included, 94 patients with HA and 93 with non-HA (Table 1). Among patients with HA, 50 (54%) had thrombotic micro-angiopathy, 25 (26%) had auto-immune hemolytic anemia, and 19 (20%) had sickle cell disease.

Table 1 Characteristics of patients with and without hemolytic anemia

Carboxyhemoglobin levels were twofold higher in patients with HA in comparison with patients with non-HA (3.0 [2.3–3.9] vs 1.6 [1.0–2.1] %, p < 0.0001). Carboxyhemoglobin level at admission was an accurate diagnostic tool for HA as the area under the curve (AUC) was 0.93 (CI 95% [0.89–0.96]), higher than LDH (AUC = 0.80, CI 95% [0.73–0.86]), unconjugated bilirubin (AUC = 0.77, CI 95% [0.71–0.84]) and methemoglobin (AUC = 0.71, CI 95% [0.64–0.79]) (Fig. 1). A threshold of carboxyhemoglobin of 2.0% for detection of hemolysis, yielded a sensitivity of 85% (CI 95% [77–90]) and specificity of 86% (CI 95% [80–90]). Specificity of carboxyhemoglobin for hemolysis detection was ≥ 99% for levels ≥ 2.7%. Using a logistic regression, we adjusted the analysis of carboxyhemoglobin for age and SOFA and we found that carboxyhemoglobin remained strongly associated with hemolysis: crude OR 74 (CI 95% [19–281]), adjusted OR 53 (CI 95% [12–240], p < 0.001 (both Wald and LR test). Focusing on patients with HA, we found that carboxyhemoglobin levels inversely correlated with hemoglobin levels (r = 0.42, p < 0.0001).

Fig. 1
figure 1

Receiver operator characteristic curves for hemolysis biomarkers: carboxyhemoglobin (HbCO) (black square), AUC = 0.93 (CI 95% [0.89–0.96]); lactate dehydrogenase (gray diamond), AUC = 0.80 (CI 95% [0.73–0.86]), unconjugated bilirubin (light gray triangle), AUC = 0.77, (CI 95% [0.71–0.84]) and methemoglobin (reversed light gray triangle), AUC = 0.71 (CI 95% [0.64–0.79])


In anemic adult patients admitted in ICU, we found that carboxyhemoglobin was a reliable diagnostic biomarker of hemolysis. Diagnostic accuracy of HA was better using carboxyhemoglobin than LDH and unconjugated bilirubin, when an optimal threshold of 2.0% was used. A similar threshold of carboxyhemoglobin at 2.2% has recently been reported in a cohort of term newborns [4]. In critically ill patients with comorbidities and multiple organ failures, classical hemolysis biomarkers as LDH and unconjugated bilirubin may lack specificity [5]. Haptoglobin is another biomarker for hemolysis, but its level may change in several critical conditions including sepsis or red blood cell transfusion [6]. Unfortunately, in our study, haptoglobin was not available in patients with non-HA. We also found a significant relationship between plasma carboxyhemoglobin and hemoglobin levels, meaning that the higher the carboxyhemoglobin, the more severe the hemolytic anemia. Carboxyhemoglobin has to be analyzed after evaluation of confounding factors that potentially increase (heavy smoker, sepsis, carbon monoxide chronic sub-intoxication) or decrease (hyperoxia) its levels.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.


  1. Dhaliwal G, Cornett PA, Tierney LM Jr. Hemolytic anemia. Am Fam Phys. 2004;11:2599–606.

    Google Scholar 

  2. Maines MD. The heme oxygenase system: a regulator of second messenger gases. Annu Rev Pharmacol Toxicol. 1997;37:517–54.

    CAS  Article  Google Scholar 

  3. Beneteau-Burnat B, Pernet P, Pilon A, Latour D, Goujon S, Feuillu A, Vaubourdolle M. Evaluation of the GEM Premier 4000: a compact blood gas CO-Oximeter and electrolyte analyzer for point-of-care and laboratory testing. Clin Chem Lab Med. 2008;2:271–9.

    Google Scholar 

  4. Karabulut B, Arcagok BC. A neglected and promising predictor of severe hyperbilirubinemia due to hemolysis: carboxyhemoglobin. Fetal Pediatr Pathol. 2020;2:124–31.

    Article  Google Scholar 

  5. Cui J, Xiong J, Zhang Y, Peng T, Huang M, Lin Y, Guo Y, et al. Serum lactate dehydrogenase is predictive of persistent organ failure in acute pancreatitis. J Crit Care. 2017;41:161–5.

    CAS  Article  Google Scholar 

  6. Kormoczi GF, Saemann MD, Buchta C, Peck-Radosavljevic M, Mayr WR, Schwartz DW, Dunkler D, et al. Influence of clinical factors on the haemolysis marker haptoglobin. Eur J Clin Investig. 2006;3:202–9.

    Article  Google Scholar 

Download references


Not applicable.



Author information

Authors and Affiliations



All authors helped in study concept and design. K.H.P., G.H. and H.A.O acquired the data. G.H., K.H.P, H.A.O. drafted the manuscript. All authors critically revised the manuscript. G.H. and H.A.O statistically analyzed. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Hafid Ait-Oufella.

Ethics declarations

Ethics approval and consent to participate

Not applicable for retrospective monocenter study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hariri, G., Hodjat Panah, K., Beneteau-Burnat, B. et al. Carboxyhemoglobin, a reliable diagnosis biomarker for hemolysis in intensive care unit: a retrospective study. Crit Care 25, 7 (2021).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: