OSA in the Perioperative Setting

The prevalence of obstructive sleep apnea (OSA) in western countries is estimated at about 5%, though recent data suggest that 26% of the adult population may be at risk (Hiestand et al. Chest. 2006;130[3]:780). Despite increased recognition of this disorder and availability for testing, the overwhelming majority of cases (estimated at 80%) has not been diagnosed. OSA, characterized by the repetitive partial (hypopneas) or complete (apneas) collapse of the upper airway during sleep, has been associated with a variety of adverse health consequences. Aside from the chronic health effects, there is growing interest in the potential for acute worsening of OSA in the perioperative setting, resulting in poor patient outcomes. Case series from the 1990s suggested that patients with OSA undergoing general anesthesia could suffer serious adverse effects, including death, in the perioperative period. These reports were followed by well-controlled retrospective studies confirming that bad outcomes – including difficult mask ventilation, difficult or failed tracheal intubation, hypoxemia, postoperative airway obstruction, cardiac arrhythmias, and MI – were more likely in patients with OSA undergoing a variety of surgeries.

Individuals with OSA appear to be at increased risk for preoperative, intraoperative, and postoperative adverse events. Poor visualization of a crowded oropharynx may make airway control difficult, a problem further complicated by the effects of induction agents. The lingering effects of general anesthesia, neuromuscular relaxants, narcotics, and sedatives may all enhance pharyngeal muscle relaxation and depress the arousal response, resulting in more frequent and longer apneas postoperatively. Supine positioning, often required following surgery, and the potential for rapid eye movement (REM) rebound sleep after the first postoperative night can both lead to worsening of sleep apnea. Surgery and general anesthesia can unfavorably impact pulmonary function, impairing gas exchange and leading to hypoxemia that may be exacerbated during sleep.

Data suggest that the majority of postoperative complications occur in the first 72 h following surgery, mostly in the first 24 h (Gupta et al. Mayo Clin Proc. 2001;76[9]:897). The effects of anesthetics, narcotics, and surgery are likely responsible for the high complication rate seen early after surgery. Recent work suggests that the apnea-hypopnea index progressively increases over the 3 nights following surgery (Chung et al. Sleep. 2010;33:A126). This may help to explain late postoperative complications seen in patients with OSA. REM rebound sleep has also been suggested to play a role in OSA worsening later in the postoperative course. During REM rebound sleep, REM-associated hypoxemic episodes can increase threefold on the second and third nights compared with the night before surgery (Knill et al. Anesthesiology. 1990;73[1]:52).

Like the general population, most patients with OSA undergoing surgery have not been diagnosed prior to surgery. In a study of 2,877 elective surgery patients, 24% were found to be at risk for having OSA, and 81% of these had not been diagnosed (Finkel et al. Sleep Med. 2009;10[7]:753). This raises the question as to whether all patients undergoing elective surgery should be screened for OSA. A recent metaanalysis of clinical prediction tools for OSA found that, while the Berlin questionnaire and the Sleep Disorders Questionnaire were the most accurate questionnaires, and morphometry and combined clinical cephalometry were the most accurate clinical tools, all instruments showed poor reproducibility and significant false-negative rates, resulting in no single ideal screening test (Ramachandran and Josephs. Anesthesiology. 2009;110[4]:928). It can be argued that instead of identifying all patients with undiagnosed OSA preoperatively, identifying only those at risk for complications is more important. Two studies, one utilizing overnight oximetry preoperatively (Hwang et al. Chest. 2008; 133[5]:1128), and the other the sleep apnea clinical score coupled with witnessed respiratory events in recovery (Gali et al. Anesthesiology. 2009;110[4]: 869), suggest this may be the case.

The need for perioperative management guidelines for patients with OSA has been recognized by both the American Academy of Sleep Medicine and the American Society of Anesthesiology (ASA) in reviews on this topic. However, both acknowledge that there is currently little evidence to guide recommendations. With this in mind, the ASA has offered expert consensus guidelines for the perioperative care of patients with OSA, though these require validation (Anesthesiology. 2006;104[5]:1081). Interventions could be performed preoperatively, intraoperatively, or postoperatively.

In the preoperative setting, identifying known and suspected patients with OSA should help with planning airway management perioperatively. Some clinicians recommend awake intubation or use of fiberoptic intubation in patients with OSA due to the propensity for airway collapse with induction agents (Riley et al. Otolaryngol Head Neck Surg. 1997;117[6]:648). If induction prior to intubation is performed, reversal agents and fiberoptic equipment should be readily available.

Intraoperatively, potential advantages to the use of regional anesthesia in the patient with OSA include increased postoperative alertness, decreased requirement for opioids, and avoidance of tracheal intubation and airway instrumentation. However, general anesthesia is often required and, therefore, consideration should be given to using drugs that have minimal effect on respiration and/or are rapidly eliminated.

Most practitioners would agree that in patients with mild OSA undergoing minimally invasive procedures with little postoperative narcotic need, no specific additional monitoring is required postoperatively. Similarly, in patients with severe OSA undergoing major thoracic or abdominal surgical procedures with significant postoperative narcotic need, a higher level of monitoring is required. Unfortunately, the large gray zone between these extremes poses the greatest challenge with respect to postoperative decisionmaking in the vast majority of patients with OSA.

While not studied, some clinicians recommend extubation of patients with OSA only once fully awake. Complete recovery from neuromuscular blockade is required, and extubation should take place in the reverse Trendelenburg or semi-upright positions. Prolonged postanesthesia care unit (PACU) monitoring should be considered prior to discharge home or transfer to the inpatient unit. Patient-controlled analgesia with no basal rate and restricted dosing may limit narcotic dosing. Naloxone should be readily available in all cases. Sedatives should be avoided altogether in the postoperative setting if feasible. Positive pressure therapy (CPAP and bilevel pressure support) is the mainstay of treatment for OSA. However, there are no randomized controlled trials of the use of CPAP or bilevel pressure support in patients with OSA in the postoperative setting. Limited data suggest a possible benefit for those patients utilizing CPAP perioperatively (Gupta et al. Mayo Clin Proc. 2001;76[9]:897). Proper studies have yet to be performed to determine if self-adjusting CPAP devices would be beneficial for patients with OSA during the postoperative period. For some patients who adamantly refuse CPAP but are having difficulties, nasopharyngeal airways may offer some benefit.

The type and location of postoperative monitoring for patients with OSA is a difficult triage decision that will depend on a number of factors unique to each case. These factors generally include age, comorbidities, severity of the known sleep apnea, the nature of surgery, type of anesthesia, presence of PACU events, and need for postoperative narcotics. Unfortunately, there is little published data other than consensus guidelines available to help with this decision-making process. While many patients can be released after monitoring in the PACU, and some will clearly require ICU monitoring, how best to care for the remainder of the patients should be individualized, taking into account the factors mentioned.

In summary, patients with OSA subject to surgery appear to be at increased risk for perioperative complications. Perioperative caregivers need to be cognizant of this and the fact that most patients with OSA have not been diagnosed with the disorder. Preoperative screening for OSA requires further study to determine its impact on outcomes. Anesthesiologists need to be prepared for potential difficulty in managing the airway of these patients. Anesthetic, sedative, and analgesic drugs should be used with caution in patients with known or suspected OSA. Nasal CPAP before surgery and after extubation may improve outcomes in these patients, though further study is needed. Decisions regarding postoperative monitoring of patients with OSA should be tailored to the specifics of each case, though a conservative approach seems prudent for patient safety.

Dennis Auckley, MD, FCCP
Division of Pulmonary, Critical Care, and
Sleep Medicine; and
Norman Bolden, MD,
Department of Anesthesiology
MetroHealth Medical Center
Case Western Reserve University
Cleveland, OH