Is overnight monitoring required for adult patients undergoing surgery for obstructive sleep apnea?


  • The authors have no funding, financial relationships, or conflicts of interest to disclose.


Multilevel surgery for obstructive sleep apnea (OSA) carries a theoretical risk of postoperative respiratory complications. In the recent past, OSA surgery was carried out with planned postoperative intensive care monitoring; however, there is a contemporary trend toward less acute postoperative care environments, or even to outpatient surgery, for many of these patients. This trend, however, was confounded by the 2006 OSA perioperative management guidelines published by the American Society of Anesthesiology suggesting that the incidence of respiratory complications after OSA surgery was relatively high (>10%), and that all patients undergoing surgery for OSA should have continuous overnight oxygen saturation monitoring after surgery.1 It remains controversial what level of monitoring after OSA surgery is the most appropriate, both in terms of patient management and sound resource allocation.


Spiegel et al. performed a retrospective review of 117 patients undergoing multilevel OSA surgery. Only three patients (2.6%) developed any respiratory problems (all had oxygen desaturations). The desaturations were comparable to levels seen on their preoperative sleep studies. All respiratory complications took place within 4 hours of surgery.2

Kieff and Busaba conducted a case-control study of 86 patients who underwent multilevel surgery for OSA. Twenty-three of these patients met the study criteria for discharge on the day of the surgery. The criteria included: oxygen saturation ≥94% on room air while asleep, no history of cardiopulmonary disease, adequate oral analgesia, adequate oral intake, hemostasis, and normal vital signs. Postoperative respiratory complications were not observed in any of the patients discharged on the day of surgery. This study concluded that same day discharge is safe for selected patients undergoing multilevel surgery.3

Terris et al. retrospectively analyzed records of 109 patients who underwent multilevel surgery for OSA. Only 4% of patients were noted to have adverse respiratory events, those being postoperative oxygen desaturations to below 90%. The study further demonstrated that respiratory complications all emerged within 2 to 4 hours after surgery. This study concluded that routine postoperative monitoring for all adult patients undergoing surgery for OSA is unnecessary, and the decision regarding level of postoperative monitoring may be determined during the period of time the patient is in the recovery room.4

Kezirian et al. performed a large multisite cohort study involving 3,130 patients undergoing inpatient surgery for OSA. The respiratory complication rate was 1.1%, which included an array of event types. Of note, this study demonstrated that the majority of postoperative complications requiring intervention emerged within 2 to 4 hours after surgery.5

A recent prospective cohort study performed at our institution examined the early postoperative course of 121 inpatients following multilevel surgery for OSA. This study found that the incidence of respiratory events (all being oxygen desaturations) in the early postoperative course was low (3.4%), and that all such patients were all identified within 4 hours of surgery. The incidence of adverse respiratory events was too small in this study to be able to predict risk factors thereof.6


The rate of adverse respiratory events after surgery for OSA is low, and the literature does not support the contention that all patients undergoing OSA surgery require overnight oximetry monitoring. Certain patient groups (such as those with higher preoperative apneic indices and/or cardiovascular comorbidities, and those undergoing tongue base surgery) may be at higher risk for respiratory complications, and should be considered for overnight oximetry monitoring. Most respiratory complications seem to occur within 4 hours after surgery; therefore, after being monitored for several hours after OSA surgery without respiratory events, and meeting other discharge criteria as appropriate, lower risk patients are suggested as being safe for same-day discharge home at the surgeon's discretion. Higher-level evidence (1 and 2) would be preferred to support a recommendation to change current practice. However, although it is difficult to obtain such evidence because of randomization concerns, the available evidence, although not randomized, is strongly consistent. Multicenter cohort or randomized trials are needed to better define risk factors for respiratory adverse events after OSA surgery.


Two studies in this paper were retrospective case series (level 4), one was an individual case-control study (level 3b), and two were prospective cohort studies (level 2b).