Objective: To evaluate the sensitivity, specificity, and area under receiver operating characteristics (AUROC) of methemoglobin levels to predict worsening sequential organ failure assessment (SOFA) scores at 72 hours in sepsis and septic shock in the intensive care setting.

Materials and Methods: The authors conducted a prospective pilot study in a single university hospital. The present study subjects were adults aged more than 19 years, sepsis and septic shock patients admitted to intensive care units, with an arterial line in place. All patients’ arterial blood samples were collected and sent to the central laboratory to analyze for methemoglobin levels at the enrolment time, then 12 and 24 hours later. Patient’s characteristics, SOFA scores and other related parameters at enrolment, 12, 24, and 72 hours later were extracted through electronic medical records. Higher SOFA score at 72 hours, comparing to the enrolment value, was considered as a reference standard.

Results: Among 30 patients, seven had worsening SOFA scores. Initial methemoglobin level had an AUROC of 0.5404 (95% confidence interval 0.2603 to 0.8204). Using Youden’s index, the determined cut-off value was 0.75%. The sensitivity of this cut-off value was 71.4% (95% confidence interval 29 to 96.3) and the specificity was 52% (95% confidence interval 30.6 to 73.2).

Conclusion: Initial Methemoglobin level had an insufficient AUROC of 0.5404 to predict worsening organ failure in critically ill patients with sepsis. As the present study is a pilot study, a larger scale study may be required.

Keywords: Sepsis; Methemoglobin; Organ Failure; SOFA; Septic shock

DOI: 10.35755/jmedassocthai.2022.05.13312

Received 25 October 2021 | Revised 19 April 2022 | Accepted 25 April 2022