Authors: Wang W et al.
Anesthesia & Analgesia, 2026.
Strengthening the Evidence on Intraoperative Occult Hypoxemia.
This letter to the editor discusses methodological considerations related to a recent study examining intraoperative occult hypoxemia and its association with postoperative mortality. Occult hypoxemia refers to a condition in which arterial oxygen saturation measured by pulse oximetry appears normal while arterial blood gas measurements reveal clinically significant hypoxemia. This discrepancy has received increasing attention, particularly because pulse oximetry accuracy can vary with skin pigmentation.
The authors acknowledge the importance of the original study and agree that occult hypoxemia represents a meaningful intraoperative safety concern. Previous research has demonstrated that pulse oximetry may overestimate oxygen saturation in some patients with darker skin pigmentation, which can lead to delayed recognition of hypoxemia and worse clinical outcomes.
However, the authors highlight several methodological issues that may influence interpretation of the study’s findings.
First, they note potential limitations in how pulse oximetry and arterial oxygen saturation measurements were paired. In the referenced study, pulse oximetry values were paired with arterial blood gas measurements using a time window of five to ten minutes before each arterial blood sample. The authors suggest that this approach may misclassify episodes of acute oxygen desaturation. Transient physiologic instability might occur during this time interval, meaning the pulse oximetry value used for comparison may not accurately reflect the patient’s oxygenation at the moment the arterial sample was obtained.
Such temporal mismatches could lead to incorrect classification of some physiologic desaturation events as pulse oximeter measurement error.
Second, the letter discusses potential biases related to arterial blood gas sampling frequency. Patients undergoing more frequent arterial blood gas testing may be more likely to have occult hypoxemia detected. This introduces the possibility of sampling bias, particularly in critically ill patients or those undergoing complex procedures. Differences in monitoring intensity across patient groups could influence the observed relationship between occult hypoxemia and outcomes.
Third, the authors question the use of a binary definition of occult hypoxemia based on the difference between pulse oximetry saturation (SpO₂) and arterial oxygen saturation (SaO₂). They suggest that analyzing the magnitude of the SpO₂–SaO₂ discrepancy as a continuous variable might provide more clinically meaningful insights. Greater discrepancies could indicate more severe physiologic derangements or device limitations and might better correlate with adverse outcomes.
Finally, the authors note that pulse oximetry measurements can vary significantly even within the same patient. Studies have shown that individuals—particularly patients with darker skin pigmentation—may experience repeated episodes of discordant pulse oximetry readings over time. A single paired measurement may therefore underestimate the cumulative burden of occult hypoxemia during surgery.
The authors conclude that the study provides important evidence highlighting the risks of intraoperative occult hypoxemia but recommend additional analyses and methodological refinements. Future research incorporating more precise measurement timing, adjustments for monitoring intensity, and continuous measures of oxygen saturation discrepancies could help clarify the true clinical impact of occult hypoxemia.
Key Points
• Occult hypoxemia occurs when pulse oximetry readings appear normal while arterial blood gases reveal true hypoxemia.
• Pulse oximeter accuracy may vary across patients, particularly with darker skin pigmentation.
• Time windows used to match pulse oximetry readings with arterial blood gases may introduce misclassification.
• Differences in arterial blood gas sampling frequency may introduce monitoring bias.
• Evaluating SpO₂–SaO₂ differences as a continuous variable may better capture clinical risk.
• Repeated discordant readings may reflect cumulative measurement error not captured by single paired comparisons.
Thank you to Anesthesia & Analgesia for allowing us to summarize and share this article.