Carbon monoxide poisoning remains one of the most common and potentially fatal toxicological emergencies worldwide, accounting for an estimated 8–34% of all poisoning cases globally. While normobaric oxygen therapy has long served as the standard initial treatment, the decision to escalate to hyperbaric oxygen therapy—a pressurized, 100% oxygen intervention—has historically lacked universally standardized criteria. A new retrospective cohort study published in June 2026 sheds light on what actually drives clinicians to choose HBOT over conventional oxygen, and the findings challenge some long-held assumptions about laboratory thresholds.
The Study: 272 Patients, Five Years of Data
Researchers at Niğde Ömer Halisdemir Training and Research Hospital in Turkey analyzed 272 adult patients diagnosed with carbon monoxide poisoning between January 2020 and January 2025. Of these, 103 patients (37.9%) received hyperbaric oxygen therapy, while 169 (62.1%) were treated with normobaric oxygen alone. The study’s primary objective was to identify which clinical, biochemical, electrocardiographic, and imaging factors most strongly influenced the decision to administer HBOT.
The patient population reflected the seasonal nature of carbon monoxide exposure: 61.4% of admissions occurred during winter months, with stove-related poisoning accounting for 87.1% of cases. Notably, more than half of all patients presented between midnight and 6:00 AM, underscoring the often accidental and nocturnal nature of these incidents.
What Actually Predicts HBOT Selection
The study’s multivariable logistic regression analysis revealed three independent predictors of HBOT administration. First and foremost was neurological status, as measured by the Glasgow Coma Scale. Lower GCS scores showed a strong inverse association with HBOT selection (odds ratio 0.66, 95% CI 0.57–0.76), meaning that each one-point decrease in GCS significantly increased the likelihood of receiving hyperbaric treatment. This finding reinforces the central role of cerebral hypoxia in carbon monoxide toxicity and validates the clinical practice of prioritizing neurological assessment over isolated laboratory values.
Second, ischemic electrocardiographic findings emerged as a powerful predictor (odds ratio 7.31, 95% CI 1.56–34.30). Patients with ST-segment elevation, ST-segment depression, or T-wave inversion were more than seven times as likely to receive HBOT compared to those with normal ECGs. This aligns with carbon monoxide’s well-documented affinity for myocardial tissue and its propensity to cause cardiac ischemia even in previously healthy individuals.
Third, elevated lactate levels showed a modest but statistically significant association (odds ratio 0.81, 95% CI 0.65–0.99), suggesting that markers of tissue hypoperfusion and anaerobic metabolism factor into clinical decision-making, albeit less prominently than neurological and cardiac indicators.
The Surprising Limitations of Carboxyhemoglobin
Perhaps the most striking finding was what did not predict HBOT selection. Despite being the hallmark laboratory marker of carbon monoxide exposure, carboxyhemoglobin levels above 20% were not independently associated with the decision to administer hyperbaric oxygen after multivariable adjustment. While COHb levels were indeed higher in the HBOT group (25.26% versus 22.83%), this difference alone did not drive treatment decisions.
Receiver operating characteristic analysis confirmed this limitation. COHb demonstrated only modest discriminative ability for predicting HBOT selection, with an area under the curve of 0.599—barely better than chance. The optimal cutoff of 23% yielded 61.4% sensitivity and 55.0% specificity, hardly the precision one might expect from a biomarker so central to the diagnosis of carbon monoxide poisoning.
This finding carries significant clinical implications. For decades, specific COHb thresholds have been cited in emergency medicine lore as indications for HBOT. Yet this large cohort study suggests that clinicians in practice are already looking beyond these numbers, prioritizing the physiological consequences of poisoning—neurological impairment and cardiac injury—over the raw exposure metric.
Clinical Outcomes and Severity Markers
The stark differences between treatment groups underscored the severity of illness in patients selected for HBOT. In-hospital mortality was three times higher in the HBOT group (21.4% versus 7.1%), and intubation rates were more than triple (33.0% versus 10.1%). Patients receiving hyperbaric oxygen also had significantly longer hospital stays and higher rates of comorbidities, including chronic obstructive pulmonary disease.
Cardiac biomarkers told a similar story. Troponin levels were nearly three times higher in HBOT recipients (29.38 versus 9.92), and creatine kinase-MB levels were more than double (4.57 versus 2.01). Arterial pH was significantly lower, reflecting the metabolic acidosis that accompanies severe tissue hypoxia. These findings collectively paint a picture of HBOT as a therapy reserved for the sickest patients—those with multi-system involvement and hemodynamic compromise.
FDA Context: Cleared Indication, Real-World Complexity
The U.S. Food and Drug Administration classifies carbon monoxide poisoning as an established indication for hyperbaric oxygen therapy, one of thirteen conditions for which HBOT devices have received 510(k) clearance. This regulatory status places CO poisoning in the company of decompression sickness, gas gangrene, and certain non-healing wounds as an evidence-backed use case.
However, the FDA’s August 2025 safety communication serves as a critical counterpoint to any uncritical enthusiasm. That letter to healthcare providers emphasized the serious risks associated with HBOT devices, including fires that have resulted in injuries and deaths. The agency stressed the importance of following manufacturer instructions, ensuring proper grounding equipment, maintaining fire prevention protocols, and providing adequate staff training. Hyperbaric oxygen may be a proven therapy for carbon monoxide poisoning, but it is not a risk-free intervention.
The FDA has also consistently warned against the proliferation of unproven HBOT uses—everything from autism and Alzheimer’s disease to cancer and Lyme disease. While carbon monoxide poisoning falls squarely within the cleared indications, the agency’s broader caution about device safety and proper facility accreditation remains relevant to all HBOT applications.
Implications for Emergency Practice
For emergency physicians and toxicologists, this study offers a data-driven framework for HBOT decision-making that aligns with what many clinicians likely already practice intuitively. The findings support a multidimensional assessment that weighs neurological status, cardiac involvement, and markers of tissue hypoperfusion rather than relying on any single laboratory value.
The poor predictive performance of carboxyhemoglobin levels, in particular, suggests that rigid COHb thresholds may be less useful than previously assumed. A patient with a COHb of 15% but altered mental status and ischemic ECG changes may warrant HBOT more urgently than an asymptomatic patient with a COHb of 30%. This clinical nuance is difficult to capture in rigid protocols but essential for optimal patient care.
The Path Forward
As hyperbaric oxygen therapy continues to expand into wellness and longevity markets—applications the FDA explicitly does not clear—it is worth remembering the therapy’s roots in acute, life-threatening conditions like carbon monoxide poisoning. The 2026 Turkish cohort study reinforces that in these emergency contexts, HBOT remains a critical tool for the sickest patients, selected not by algorithm but by careful clinical judgment that prioritizes organ-specific injury over abstract numbers.
For patients and providers alike, the message is clear: when it comes to carbon monoxide poisoning, the brain and heart speak louder than the blood test.
Sources
- Yüceer Ö. Neurological and Cardiac Determinants of Hyperbaric Oxygen Therapy in Carbon Monoxide Poisoning: A Retrospective Cohort Study. Int J Gen Med. 2026;19:604040.
- FDA — Follow Instructions for Safe Use of Hyperbaric Oxygen Therapy Devices (August 2025)
- FDA — Hyperbaric Oxygen Therapy: Get the Facts
- Undersea and Hyperbaric Medical Society (UHMS) — HBO Indications