A novel algorithm for selecting patients who require treatment for obstructive sleep apnea (OSA) before undergoing bariatric surgery proved safe in a prospective cohort study of 1,103 patients.
Screening for OSA is recommended before bariatric surgery. OSA has been associated in several meta-analyses with increased risk for postoperative complications – not limited to bariatric surgery – and some studies have suggested that this increased risk may be limited to severe OSA, said Frédéric Series, MD, of Université Laval, Quebec City, at the virtual annual meeting of the Associated Sleep Societies.
The preoperative screening algorithm, which utilizes the results of nocturnal home oximetry and morning capillary gas measurements, effectively stratified patients for the risk of postoperative adverse events and “safely selected patients who don’t need [continuous positive airway pressure] before bariatric surgery,” he said. “The risk of postoperative adverse events following bariatric surgery was not increased in untreated OSA patients with low or moderate risk of severe OSA and hypoventilation.”
The study also demonstrated, he said, that patients with severe OSA with or without hypoventilation, even when correctly treated, remain at higher risk for complications.
The algorithm utilizes an oxygen desaturation index (ODI) corresponding to 3% drops in SaO2 and the percent of the total recording time with an SaO2 below 90%, as well as capillary gas measurements (PCO2). Treatment was initiated for those with severe OSA (ODI ≥ 25/hr, < 10% of recording time with a SaO2 below 90%) or OSA with hypoventilation (PCO2 ≥ 45).
“When the ODI was less than 25 per hour, and when the total recording time spent below 90% SaO2 was less than 10%, with PCO2 < 45 mmHg, we expected no need for CPAP treatment,” Dr. Series said. For analysis, the investigators considered part of the untreated group – those with an ODI < 10/hr (no or mild OSA) – as a control group.
Treated patients underwent CPAP/BiPAP for a mean duration of 1.5 months. Good treatment compliance was mandatory for surgery, and treatment was continued immediately after extubation, in the recovery room, in nearly all patients, Dr. Series reported.
The analysis covered 1,103 patients: 447 controls (40.8%), 358 untreated (32.7%), 289 treated for OSA (26.4%) and 9 (0.8%) treated for OSA + hypoventilation. Patients with OSA, particularly those with severe OSA and those with hypoventilation, were older and heavier and significantly more likely to have hypertension and diabetes than controls.
There were no differences between the four groups in 10-day reoperation or 30-day readmission occurrence, and postoperative complications were “particularly infrequent in the control and OSA-untreated groups, with no differences between these two groups,” Dr. Series said.
Cardiac arrhythmia (mainly atrial fibrillation) occurred more frequently in the OSA-treated group (2.4%) and the OSA/hypoventilation patients (11%) than in the other groups (0.5%-0.6%).
Respiratory failure occurred in about one-third of patients with hypoventilation, and admission to the ICU was “dramatically higher” in patients with hypoventilation (67%), because of respiratory failure, arrhythmia, or other unstable medical conditions, Dr. Series said.
There were no differences between the groups in the duration of surgery or the amount of anesthetic used, but the length of stay in the recovery room was significantly longer in the OSA-treated and hypoventilation groups. The length of hospital stay was also longer in these groups. Sleeve gastrectomy was the most frequent bariatric surgical procedure across all groups, including 100% of patients with hypoventilation, he noted.
Asked to comment on the study, Octavian C. Ioachimescu, MD, PhD, of Emory University in Atlanta and the Atlanta Veterans Affairs Medical Center in Decatur, said the algorithm “clearly deserves further validation in other clinical-based cohorts and longer-term outcome assessment.”
Dr. Series reported that he has no relevant disclosures. Dr. Ioachimescu also said he has no relevant disclosures.