Introduction/Background

Obstructive sleep apnea (OSA) is a chronic condition that is characterized by recurrent episodes of partial or complete collapse of the upper airway during sleep. The reduction or cessation of airflow during these obstructive episodes may result in significant decreases in oxyhemoglobin saturation and hypercarbia and eventual arousal from sleep. Patients with sleep apnea may have a variety of nocturnal symptoms, such as loud disruptive snoring, choking, and gasping, and they may have observed pauses in breathing. Because sleep is fragmented, daytime symptoms include excessive daytime sleepiness, mood disorders, and neurocognitive impairment, which lead to an increased likelihood of accidental injury or death. Additionally, it is well accepted that the abnormalities in gas exchange that result from OSA are associated with adverse cardiovascular, endocrine, and cerebrovascular consequences.

Public awareness of OSA and its health consequences is increasing, and concern among health care providers is growing that patients with sleep apnea may be at risk of adverse perioperative outcomes, including death. General population studies suggest that 5% of middle-aged women and 9% of middle-aged men have OSA, and data suggest that the prevalence of OSA is even higher in the elderly population. Unfortunately, the prevalence of OSA in adult patients undergoing outpatient surgery is still unknown. Furthermore, it has been estimated that up to 90% of those with the disease carry no formal diagnosis. With 15 million patients undergoing outpatient surgeries in free-standing ambulatory surgical centers each year, statistically, more than 1 million of them may have disordered breathing.

The presence of OSA in the surgical patient is thought to lead to potential problems with mask ventilation, tracheal intubation, extubation, and the ability to provide adequate analgesia without respiratory compromise. When the diagnosis of OSA is known, there is an opportunity to arrange for additional resources to deal with anticipated potential airway complications and the need for possible prolonged postoperative monitoring. However, the patient who has signs and symptoms of OSA but does not have a formal diagnosis poses a particular problem for the ambulatory anesthesiologist who must decide whether to proceed with surgery or delay the case until the patient undergoes a formal evaluation. Additionally, the anesthesiologist must decide whether the patient is a candidate for a free-standing ambulatory surgical center.

The gold standard test used to determine the presence of OSA is polysomnography (PSG). PSG is a relatively expensive, time-consuming, and labor-intensive test that cannot be performed on the day of the surgical procedure. The patient who undergoes PSG is brought to a sleep laboratory in the evening, monitors are applied, and simultaneous recordings of several physiologic signals are acquired over an 8-hour period while the patient sleeps. Most sleep laboratories define an abnormal breathing episode of obstructive apnea as the complete cessation of airflow for a minimum of 10 seconds during sleep while the patient makes persistent efforts to breathe. Although the definition of hypopnea is less uniform, the most common description is a decrease in airflow of greater than 30% associated with a decrease in oxyhemoglobin saturation of 4% or more. The apnea–hypopnea index (AHI) is the total number of all recorded episodes of apneas and hypopneas per hour of total sleep time, and if sleep-disordered breathing is detected, it is reported as mild, moderate, or severe, based on the AHI. It is important to note that the criteria for diagnosis and the presentation of OSA differ between the adult and pediatric populations, and what is discussed in this review applies only to the management of adults.

Options

At present, there is no consensus to define the specific additional risk, if any, that the presence of OSA poses to the ambulatory surgical patient. Because the risk of potential OSA during outpatient surgery is poorly defined, postponement of a surgical procedure to define the patient’s risk may seem unreasonable to the patient and the surgeon. There are both financial and social pressures to proceed because the patient may have made arrangements for time away from work, as well as provisions for family members to help during the recovery period. Additionally, even though the procedure may have been scheduled as an elective outpatient procedure, the nature of the surgery may still be considered relatively urgent, as in the case of a breast biopsy to rule out cancer. Delay of this type of procedure can have tremendous psychological consequences for the patient and may result in delay of treatment. Although no large-scale, randomized trials have compared perioperative adverse outcomes of patients with OSA with those of healthy patients, several observational studies have examined this question. Therefore current perioperative care is based on clinical judgment and an understanding of the pathophysiologic mechanism and consequences of OSA.

Pathophysiology/Mechanism of Action

The occurrence of pharyngeal collapse during sleep suggests that sleep onset is associated with functional alterations in airflow in the upper airway that reduce patency and increase resistance to airflow. The point of obstruction can occur anywhere in the upper airway, from the soft palate and nasopharynx to the base of the tongue and epiglottis, and frequently occurs at different sites during the various stages of sleep. Bachar and colleagues demonstrated sites and patterns of obstruction with the use of sleep endoscopy in 55 surgical patients. They found that the most common site of obstruction was uvulopalatine and also noted that many patients (72%) had multiple sites of obstruction. Regardless of where the obstruction occurs, two subsequent effects are thought to follow. First, with repetitive episodes of hypoxia and hypercapnia and the reoxygenation that occurs during arousal, oxidative stress ensues and systemic inflammation follows. Reactive oxygen species are formed and cause injury to the surrounding tissue. Although these molecules trigger pathways that are adaptive to hypoxia, they have also been found to have an association with harmful inflammatory and immune responses. Among the changes are activation of endothelial cells, leukocytes, and platelets. Sympathetic activity is increased, which, after repetitive cycles of hypoxia and hypercarbia, results in upregulation of both alpha- and beta-receptors. This may have a role in the pathogenesis of coronary and cerebrovascular disorders.

One of the most commonly recognized cardiac sequelae of OSA is right-sided heart dysfunction. The increased sympathetic activity associated with the hypoxia and hypercarbia leads to an increase in pulmonary vascular resistance. The endothelial wall thickens, and pulmonary hypertension can ensue. The right ventricle hypertrophies to meet the demand and, if unremedied, can eventually dilate and enlarge. However, although historically most attention has been directed toward the status of the right side of the heart during a preoperative assessment in the patient suspected of having OSA, there is a far greater association with systemic hypertension and, more specifically, uncontrolled hypertension. Of patients with documented OSA, 60% to 70% have a concomitant diagnosis of systemic hypertension, whereas only about 20% of those with OSA have progression of the disease resulting in pulmonary hypertension severe enough to cause right ventricular dysfunction.

OSA has been implicated in the pathogenesis of various other comorbidities, including coronary artery disease, congestive heart failure, cardiac arrhythmias, sudden death, stroke, and impaired glucose metabolism.

Evidence

To date, there is a paucity of outcome data generated from surgical patients with diagnosed or undiagnosed OSA and even less that addresses outcomes in the ambulatory surgical population. Recent studies suggest that 24-hour observation in a monitored environment confers a minimal, if any, advantage in risk reduction for ambulatory surgical patients with uncomplicated OSA.

Most available data arise from otolaryngologic studies, specifically patients undergoing uvulopalatopharyngoplasty (UPPP). Several studies have addressed the question of whether patients with OSA undergoing upper airway procedures should be monitored in an intensive care unit (ICU) postoperatively, but the data are retrospective and inconclusive. Mickelson and Hakim retrospectively analyzed 347 consecutive patients who underwent UPPP. Of the 14 patients who had complications, five involved the airway, and the episodes occurred in the immediate perioperative period. Additionally, no correlation was seen between the rate of complication and the severity of OSA. Of the five patients with airway complications, three required reintubation. One patient had bronchospasm immediately after extubation, one patient was thought to have been prematurely extubated in the operating room and experienced subsequent respiratory arrest, and one patient was reported to have respiratory distress in the recovery room of unknown etiology. Respiratory complications developed in two of the five patients after admission to the ward; however, neither required reintubation. The authors concluded that ICU care postoperatively was not required for most patients undergoing UPPP and that the rate of complication was substantially higher in patients who had undergone simultaneous otolaryngologic procedures in addition to UPPP. Hathaway and Johnson examined the outcomes of 110 patients scheduled for outpatient UPPP. Twenty of the 110 patients required admission (18%); however, no patient required transfer to an ICU. Although three patients were admitted for postoperative oxygen desaturation, this did not correlate with the severity of AHI. Additionally, the majority of admissions were for control of pain and nausea. The authors emphasized that appropriate patient selection is essential in minimizing the risk of perioperative complications in patients undergoing UPPP, and in their study, any patient with severe cardiopulmonary comorbidities was eliminated as a candidate for UPPP. Terris and colleagues found similar results when they performed a retrospective analysis of 109 patients with OSA who were scheduled for 125 upper airway procedures. The rate of airway complications was 0.9% (1 of 109), and the one patient who experienced airway obstruction did so in the immediate postoperative period. Again, the authors concluded that ICU monitoring for all patients undergoing UPPP was unnecessary and that the decision for discharge to the floor or home could be made based on the patient’s status in the recovery room within 2 hours of the surgical procedure. In another retrospective analysis of OSA patients undergoing airway procedures, Spiegel and Tejas found that, if airway complications were to occur, they could be identified within 2 to 3 hours postoperatively and also concluded that same-day discharge was an option for some patients. Although it appears that select patients with OSA can be safely discharged to home after UPPP, it seems prudent that this be done in a facility with provisions for transfer to an overnight ward for observation.

Studies in the literature examining nonotorhinolaryngologic surgeries in patients with OSA are scant. However, studies that retrospectively analyze outcomes of inpatient surgical procedures have suggested that OSA is an independent risk factor for adverse outcomes. Gupta and colleagues studied 110 patients with OSA diagnosed either before or after total hip or knee replacement and matched the population with control subjects. OSA was associated with an increased incidence of “serious” adverse perioperative events requiring transfer to an ICU. Although the severity of OSA or AHI was not related to the incidence of complications, OSA patients who were compliant with continuous positive airway pressure (CPAP) preoperatively were noted to have a decreased incidence of complications when compared with patients with OSA who did not use CPAP.

Sabers and colleagues at the Mayo Clinic in Rochester, Minnesota, designed a retrospective study to determine whether the preoperative diagnosis of OSA was an independent risk factor for perioperative complications after outpatient surgery. A total of 234 patients who had been previously diagnosed with OSA by PSG were scheduled for ambulatory surgical procedures and were matched with control subjects. All types of surgery were included with the exception of otorhinolaryngologic procedures. The primary outcome measured was unplanned hospital admission or readmission; however, recorded data included episodes of bronchospasm, airway obstruction, and reintubation during the recovery period. Previously diagnosed OSA was not found to be an independent risk factor for unplanned admissions or for other adverse perioperative events.

We have examined the prevalence of OSA and propensity to OSA in our own outpatient surgical population at Johns Hopkins Hospital. A previously validated prediction model was used to determine the pretest probability for OSA in 3557 consecutive adult patient undergoing ambulatory surgery of all types except ophthalmologic procedures. Propensity to OSA was determined by logistic regression analysis. Relevant perioperative data such as anesthetic technique, difficulty with endotracheal intubation, need for supplemental oxygen, need for assisted ventilation, reintubation, unplanned admission, and death were recorded; 2.6% of the patients had a greater than 70% propensity for OSA but had not yet been given a diagnosis. Of these high-risk patients, only 28.2% (31 of 110) of male patients and 21.6% (11 of 51) of female patients had a previous self-reported diagnosis of possible OSA. The results of the study suggested that OSA is relatively common in an ambulatory surgical population and that the majority of patients with a propensity for OSA who undergo ambulatory surgery remain undiagnosed. There was a positive correlation of patients with a higher propensity to OSA (versus non-OSA) and increased difficulty of intubation, administration of intraoperative ephedrine, metoprolol, and labetolol, and need for prolonged supplemental oxygen. However, we found no relationship between unplanned admission or readmission, life-threatening events such as reintubation, cardiac arrhythmia, or death in patients with either a diagnosis or higher propensity for OSA. Therefore our data suggest that patients with OSA may require additional perioperative interventions; however, they can be treated safely in an ambulatory care center.

Acknowledging the weakness of the data available to guide the perioperative management of patients with uncomplicated OSA, it appears that these patients can be safely managed as outpatients. However, those patients with comorbid illnesses may need to be managed differently. Moreover, as the complexity and invasiveness of ambulatory surgical procedures increase with advances in technique and technology, the appropriateness of care of patients with OSA in an ambulatory surgical center may need further exploration.