Sleep Apnea Measure Predicts Complications After Major Surgery

A metric for obstructive sleep apnea (OSA) that accounts for oxygen desaturation along with frequency of obstruction correlated with hard outcomes among patients undergoing major noncardiothoracic surgery, a study showed.

The rate of 30-day postoperative mortality and cardiovascular complications increased significantly from 1.6% in patients with a low sleep apnea-specific hypoxic burden (SASHB; less than 32% min/h) to 5.8% among those with a high SASHB (at least 80% min/h) at diagnosis, Sébastien Bailly, PharmD, PhD, of University Grenoble Alpes in France, and colleagues reported in JAMA Network Open.

Risk climbed with SASHB, with odds ratios of 1.76 for those with 32-80% min/h and 2.79 at over 80% min/h as compared with less than 32% min/h.

This measure, defined by the area under the desaturation curve associated with sleep-related obstructive respiratory events, describes both the frequency of upper airway obstructions as well as the duration and depth of oxygen desaturations during sleep.

The typical measure of OSA severity — apnea-hypopnea index — doesn’t adequately capture risk, Bailly’s group argued, with prior studies showing a stronger association of SASHB with cardiovascular adverse events.

The finding that SASHB is independently associated with risk in postoperative settings is consistent with data from nonsurgical settings, the researchers said.

The anatomical and physiological factors that lead to OSA also “increase vulnerability to airway obstruction under general anesthetics, sedatives, and postoperative analgesics that relax the upper airway muscles and impair ventilatory response,” potentially predisposing OSA patients to postoperative mortality complications, the group noted.

“Further research is needed to determine whether interventions guided by SASHB scores can modify postoperative risk in patients with OSA,” they wrote.

The study didn’t compare outcomes between patients prospectively managed according to one risk tool versus another. However, SASHB did look promising to improve estimation of risk of postoperative complications, according to the researchers.

A risk score based on age, emergency admission before surgery, and SASHB had an area under receiver operating characteristic curve of 0.73 (95% CI 0.68-0.77). A simplified, and more accessible, version of SASHB automatically derived from the single oximetry signal extracted from diagnostic sleep studies performed similarly (0.75, 95% CI 0.67-0.78).

An elastic net regression analysis comparing performance of those two versions against the apnea-hypopnea index and other conventional measures of OSA severity showed that the SASHB had stronger associations with the primary outcome of 30-day mortality or cardiovascular complications.

The negative predictive value of the full risk score was high at 96% or more, as postoperative complications were rare in the cohort studied, Bailly and colleagues noted. “This finding suggests that the score performs best as a rule-out tool, which might be clinically useful for identifying individuals who may safely undergo standard postoperative monitoring, thereby helping to prioritize resources for higher-risk patients.”

The study included 2,286 adult OSA patients (median age 58, 64.4% men) who were treated at healthcare clinics participating in the Pays de la Loire Sleep Cohort and underwent major noncardiothoracic surgery between OSA diagnosis and December 2024, a median of 4.5 years after OSA diagnosis.

The participants were followed for outcomes using a French health administrative database. Altogether, 80 patients (3.5%) had a primary endpoint within 30 days of surgery, encompassing stroke, atrial fibrillation, heart failure, myocardial infarction, venous thromboembolism, and all-cause mortality.

Findings were similar across demographics, history of cardiovascular diseases, type of sleep study, delay between OSA diagnosis and surgery, and use of positive airway pressure (PAP) therapy.

Limitations included reliance on administrative health databases for outcomes, potential for residual confounding from unmeasured factors (which included opioid or oral morphine equivalents usage and use of PAP in the recovery room and postoperative period), and inability to determine causality.

“Further studies are required to define a specific actionable SASHB threshold for perioperative PAP protocols and enhanced monitoring,” the researchers noted.