Discussion
OSA is a common disorder caused by disruption to breathing during sleep due to recurrent collapse of the pharyngeal airway leading to hypopnoea or apnoea events which has both short term and long term health effects (17). Studies of patients with OSA have demonstrated an increased risk of post-operative complications including respiratory and cardiac events such as myocardial infarction or cardiac arrhythmias and patients are more likely to require intensive care input (2, 18, 19). A sensitivity to narcotics resulting in respiratory depression and desaturations is also a recognised complication among this cohort of patients which usually results in increased post-operative monitoring in at least a high-dependency setting (20).
However, with progress in day surgery, the Association of Anaesthetists and the British Association of Day Surgery have published a consensus document recognising that not all OSA patients need overnight monitoring and selected patients can be safely discharged the same day (1). These guidelines do advise certain considerations such as avoiding postoperative opioid medications, use of regional anaesthesia if possible, optimising comorbid conditions, the post-operative use of a CPAP device if patients were already utilising one and a postoperative review prior to discharging the same day. Some of these considerations are difficult to achieve in upper airway surgeries and the risk of airway complications or respiratory events are higher than in other surgeries.
Our review indicates that despite the majority of the patients falling into the moderate or severe OSA category and having upper airway surgery, almost half of the patients were discharged on the same day with minimal respiratory events either in the immediate post-operative period or during follow-up in the form of readmissions. Oxygen desaturations were the most commonly reported respiratory event and these were often managed with oxygen supplementation with no further complications observed during the remainder of the inpatient stay. This would suggest that this group of patients can be safely monitored in an area with continuous monitoring and increased nurse to patient ratio, but not necessarily needing intensive care or high dependency input. Major airway complications such as laryngospasms were almost always picked up immediately post-extubation and this would prompt post-operative care in a more appropriate setting. There was no mortality reported among the 1836 patients in this review.
Major respiratory events following nasal and palatopharyngeal surgery for OSA are rare. Concurrent tongue base surgery however, can be associated with more serious respiratory events and in these patients, overnight observation would be prudent.
One large study evaluated a North American database for morbidity and mortality following uvulopalatopharyngoplasty (UPPP) (n = 1096) and multilevel sleep surgery (n = 1578) for OSA (3). The multilevel sleep surgery included patients who had UPPP in addition to other procedures (including tonsillectomy and adenoidectomy, tongue and mouth surgery, epiglottidectomy, glossectomy, limited pharyngectomy, hyoid myotomy and suspension, excision of lingual tonsil, neurostimulators (intracranial) procedures, reconstruction of lower jaw, other unspecified procedure). They reported a total of four (0.15%) deaths within 30 days of surgery: one death in the UPPP only group (0.09%), and three deaths following UPPP with concomitant procedures (0.19%). There were no reported details of the cause and timing of deaths. It is therefore not clear if the risk of death is an issue in performing such cases as a daycase surgery.
The findings from the systematic review was limited by a lack of well-designed prospective studies with pre-defined discharge criteria and a comprehensive assessment of complications such as continuous pulse oximetry monitoring among all patients. Given patients were not all systematically followed-up on discharge, it is unclear if there were any significant respiratory events out of hospital or if patients presented to other hospitals with complications. There was limited data on the post-operative use of CPAP which is important given issues around nasal packing and CPAP compliance. There was also significant heterogeneity in the methodology of the studies and thus no inferential statistics could be performed. None of the included studies were conducted in the United Kingdom and this may limit the applicability of the results to a UK population.
The included studies have shown that it is feasible to perform upper airway surgeries in carefully selected patients with OSA as day cases. Table 4 is a summary of the characteristics of daycase patients from studies that have specified them. Patients with mild to moderate OSA and no cardiopulmonary comorbidities were performed as day cases if there were no concurrent tongue surgery. However, most had a post-operative review prior to final discharge decision. Those that had episodes of desaturations (<94% on room air) in recovery, had inadequate oral intake or needed strong opioids for pain relief were admitted for further observation.
There is a need for further well-designed prospective studies with clear criteria for daycase patients and those needing an overnight admission. Such studies should capture information on perioperative CPAP use, have a comprehensive assessment of postoperative complications including readmissions to other hospitals.