Obesity Hypoventilation Syndrome: A Review of Epidemiology, Pathophysiology, and Perioperative Considerations

With the help of a research librarian, we searched Medline (January 1948 – May 2011), Medline in-process & Other Nonindexed Citations (May 2011), EMBASE (January 1980 – May 2011), Cochrane Database of Systematic Reviews (January 2005 – April 2011) and the Cochrane Central Register of Controlled Trials (May 2011). Searches were conducted using the following components: “Obesity Hypoventilation Syndrome and related terms,”“Anesthesia and related terms,”“Screening or Preoperative Assessment and related terms,” and “Therapy or Treatment or Management and related terms.” The following key terms were used for the literature search: “obesity hypoventilation syndrome,”“OHS,”“Pickwick,”“hypoventilation,”“obesity,”“overweight,”“anesthesia,”“anesthesiology,”“preoperative care,”“screening,”“therapeutics,”“disease management,”“treatment outcome,” and “therapy.” The results of the search were limited to adult human studies published in the English language. To ensure that all potentially relevant articles were included, the reference lists of relevant reviews and included articles were searched manually for further studies.

Studies were selected independently by three reviewers (EC, DL, JW) who screened the titles and abstracts to identify studies reporting prevalence and treatment of patients with OHS. OHS was defined as daytime hypercapnia and hypoxemia (PaCO²≥45 mmHg and PaO²<70 mmHg at sea level) in obese patients (BMI ≥30 kg/m²) with sleep-disordered breathing in the absence of any other cause of hypoventilation. Any disagreements were resolved by consensus or by consulting the senior author (FC). All study designs including randomized control trials and observational studies were included.

Data extraction was performed independently by two reviewers (DL, EC). Disagreements were resolved by consensus or by consulting the senior author (FC). The following data were extracted from each study: first author, publication year, study design, sample size, treatment method, and treatment duration. The following parameters were collected from the studies: age, sex, neck circumference, waist/hip ratio, arterial oxygen tension, arterial carbon dioxide tension, pH, serum bicarbonate (HCO³⁻), hemoglobin, and hematocrit. Lung function such as percent predicted forced expiratory volume in the first second (FEV¹% pred), percent predicted forced vital capacity (FVC % pred), FEV¹/FVC, carbon dioxide sensitivity, and sleep parameters such as apnea-hypopnea index (AHI), percentage of total sleep time with oxygen saturation (SpO²) <90%, average awake SpO², and minimum SpO²were tabulated. Mean values of the collected parameters were calculated for patients with OHS and eucapnic obese individuals. Statistical significance of each parameter between the two groups was tested with the Student t  test. Pooled SD was calculated as previously described.6 

The search strategy identified 583 articles (fig. 1). After screening the abstracts, 21 studies were included and another 26 studies were manually retrieved from the reference lists. There were 30 prospective studies, 12 retrospective studies, 4 randomized controlled studies, and 1 case-control study. Of the 47 studies, 5 prospective studies and 4 randomized controlled trials investigated pharmacologic treatment of OHS. The remainder examined ventilation therapy. In total there were 1,077 patients in whom OHS was diagnosed by fulfilling three criteria: (1) daytime hypercapnia and hypoxemia (PaCO²>45 mmHg and PaO²<70 mmHg at sea level) in the absence of any other cause of hypoventilation; (2) obesity (BMI ≥30 kg/m²); and (3) sleep-disordered breathing.7 

The prevalence of OHS in the general population is unknown because it has not been studied. Because approximately 1.5% of the general United States population has severe obesity and OSA, and 10–20% of the severely obese patients with OSA have OHS,8,9the prevalence of OHS among the general adult population in the United States is estimated to be 0.15–0.3%.3Five studies (four prospective and one retrospective) including a total of 1,326 patients, evaluated patients referred to sleep centers with clinical symptoms of OSA10,–,13(table 1). The prevalence of OHS ranged from 11% to 20%, with an overall prevalence of 16%. Three retrospective studies including a total of 1,927 patients evaluated patients with a known diagnosis of OSA.14,–,16The prevalence of OHS ranged from 9% to 14%, with an overall prevalence of 11%. Finally, in three studies (one prospective and two retrospective) of a total of 388 patients awaiting bariatric surgery, the prevalence of OHS was found to be 7–22% with an overall prevalence of 8%.17,–,19 

In summary, the prevalence of OHS is 11% in patients with known OSA and 8% in bariatric surgical patients.

Daytime hypercapnia is the distinguishing feature of OHS that separates it from simple obesity and OSA. It is entirely due to hypoventilation given that a short course of noninvasive positive airway pressure therapy (less than 2 weeks) improves hypercapnia without any significant changes in body weight, carbon dioxide production, or the volume of dead space.20There are three leading hypotheses for the pathogenesis of chronic daytime hypoventilation in OHS: impaired respiratory mechanics because of obesity, leptin resistance leading to central hypoventilation, and impaired compensatory response to acute hypercapnia in OSA (fig. 2).3,21 

Obesity imposes a significant load on the respiratory system and could result in hypoventilation secondary to fatigue and the relative weakness of the respiratory muscles.22,–,24Several studies that compare patients with OHS with individuals who are eucapnic and obese have noted a significantly higher BMI in the OHS group.10,14,16,25However, because less than a third of the morbidly obese individuals develop hypercapnia, other mechanisms may result in hypoventilation.10,16,26 

Leptin is a protein produced specifically by the adipose tissue that regulates appetite and energy expenditure.27,–,29It crosses the blood-brain barrier and exerts its effect through binding to leptin receptors in various areas of the brain.28In obese patients, a higher level of leptin is found causing an increase in ventilation to compensate for the increased carbon dioxide production associated with excess body mass.27,30,31Patients with OHS exhibit an even higher serum leptin level than eucapnic individuals matched for BMI, and their serum leptin level drops after positive airway pressure (PAP) therapy.32,33These observations suggest leptin resistance may contribute to the hypoventilation in OHS.

Obstructive apneas, hypopneas, and long periods of hypoventilation during sleep result in transient episodes of acute hypercapnia. Compensatory mechanisms, including hyperventilation during brief periods of arousal between the obstructive events and renal bicarbonate retention, are required to maintain carbon dioxide homeostasis.34Chronic hypercapnia in OHS may occur if these compensatory responses are impaired. In eucapnic subjects with OSA, periods of apnea are separated by periods of hyperventilation such that the accumulated carbon dioxide load is eliminated.35However, when apneas become three times longer than the breathing interval, carbon dioxide accumulates (fig. 3).35 

Patients with OHS, in comparison with those with eucapnia, have a reduced duration of ventilation between periods of apnea.36This is possibly related to a gradual adaptation of chemoreceptors secondary to mild elevation of serum HCO³⁻that can occur even during acute hypercapnia. In eucapnic individuals, arterial carbon dioxide tension is restored to normal during wakefulness and the excess HCO³⁻is excreted. However, a transition from acute to chronic hypercapnia may result if the small amount of retained HCO³⁻is not excreted by the kidneys, leading to a reduction of ventilatory carbon dioxide responsiveness. In a computer model, when both carbon dioxide response and the rate of renal HCO³⁻excretion was abnormally low, a rise in awake arterial carbon dioxide tension and HCO³⁻developed over multiple days.37 

Compared with obese patients with eucapnia, patients with OHS demonstrate four main clinical features: more severe upper airway obstruction, impaired respiratory mechanics, blunted central respiratory drive, and increased incidence of pulmonary hypertension. Table 2compares various reported demographic and physiologic parameters between patients with OHS and obese patients with eucapnia.5,10,11,14,16,25,26,38,–,46The clinical features of OHS are summarized in figure 4.

Patients with OHS display increased upper airway resistance in both the sitting and supine position in comparison with obese individuals with eucapnia.47This may contribute to the increased work of breathing found in OHS patients.48 

Simple obesity impairs respiratory mechanics leading to reduced lung volumes, decreased chest wall compliance, increased respiratory resistance, and increased work of breathing.49,50These parameters are further impaired in OHS patients (table 2). Spirometric values from patients with OHS typically reveal a restrictive pattern with a reduction in FEV¹and FVC but normal FEV¹/FVC. Functional residual capacity, total lung capacity, and expiratory reserve volume are also reduced in OHS compared with eucapnic obesity.

Studies on OHS respiratory mechanics reveal an excessive load imposed on the respiratory system. Chest wall compliance was reduced 2.5-fold in patients with OHS versus  those with eucapnic obesity.51In addition, pulmonary resistance is increased in OHS, likely secondary to the reduction in functional residual capacity. These alterations in respiratory mechanics double the work of breathing in OHS patients compared with patients with eucapnic obesity.48,51The work of breathing is further increased when these patients adopt the supine position from sitting due to the cephalad shift of abdominal contents.48,50 

Obese individuals need to generate higher levels of minute ventilation to maintain eucapnia due to their higher basal oxygen consumption, carbon dioxide production, and work of breathing.30,31,49Obese individuals have a substantially increased central respiratory drive compared with normal-weight patients to compensate for the increased ventilatory requirements.50In contrast, patients with OHS have a blunted central respiratory drive to both hypercapnia and hypoxia. They do not hyperventilate to the same degree as obese individuals with eucapnia when forced to rebreathe carbon dioxide43,52,53or breathe a hypoxic gas mixture.53The blunting of central respiratory drive may result from leptin resistance and sleep-disordered breathing.29,34 

The incidence of pulmonary hypertension, as defined by a mean pulmonary arterial pressure ≥20 mmHg, is higher in patients with OHS than in obese patients with eucapnia, ranging from 30% to 88%.2,4,45The etiology of pulmonary hypertension is likely secondary to chronic alveolar hypoxia and hypercapnia. In some OHS patients, pulmonary hypertension may result from left-heart failure because left ventricular hypertrophy is a common finding due to associated cardiomyopathy in severe obesity,54and pulmonary arterial occlusion pressure has been reported to be increased in OHS.45 

Obesity and OSA are associated with a spectrum of comorbidities such as coronary artery disease, heart failure, stroke and metabolic syndrome, which result in increased morbidity and mortality.55,–,59Furthermore, patients with OSA are at increased risk of developing postoperative complications including arrhythmias and hypoxemia.60,–,62An increased risk of intensive care unit transfer and increased length of hospital stay were also observed among patients with OSA who underwent noncardiac surgery.61 

Several studies showed that patients with OHS may experience higher morbidity and mortality than patients who are similarly obese and have OSA. Compared with obese individuals with eucapnia, patients with OHS were more likely to develop heart failure (odds ratio (OR) 9, 95% CI 2.3–35), angina pectoris (OR 9, 95% CI 1.4–57.1) and cor pulmonale (OR 9, 95% CI 1.4–57.1).4They also received higher rates of long-term care at discharge (19%vs . 2%, P = 0.01), and invasive mechanical ventilation (6%vs . 0%, P = 0.01).5 

The mortality rate in patients with untreated OHS is high. A retrospective study described a mortality rate of 46% during an average 50-month follow-up period in OHS patients without therapy.63In addition, patients with OHS exhibit a higher mortality rate than obese patients with eucapnia. A group of patients with OHS and obese patients with eucapnia were followed after hospital discharge for 18 months. Those with OHS had a mortality rate of 23% compared with 9% in the eucapnic obesity group.5Most patients in the OHS group were discharged without any therapy. In a 1992 study evaluating open bariatric surgery, patients with either OHS or OSA suffered a surgical mortality rate of 4%, significantly higher than that reported in patients without OHS or OSA (0.2%, P < 0.01).64The major causes of death include pulmonary embolus and peritonitis from leaks. Perioperative safety of bariatric surgery has improved since.65,66However, in high-risk patients (OHS, previous history of venous thromboembolism, BMI ≥ 50 kg/m², male sex, hypertension, and age ≥ 45 yr) undergoing gastric bypass, mortality ranges between 2–8%.67,–,69 

In summary, patients with OHS experience higher morbidity and mortality than those who are obese with eucapnia. Previous history of venous thromboembolism, morbid obesity, male sex, hypertension, increasing age, and noncompliance with PAP treatment may further increase mortality risk. Surgical mortality rate in high-risk OHS patients undergoing bariatric surgery is between 2–8%.

Therapeutic interventions for OHS therapy include four main components: PAP therapy, supplemental oxygen, weight reduction surgery, and pharmacologic respiratory stimulants.

The two main forms of PAP therapy currently being used are continuous positive airway pressure (CPAP) and bilevel PAP.

Short-term benefits of PAP include an improvement in gas exchange and sleep-disordered breathing (tables 3and 4). There were five studies that evaluated the effects of a short course (⩽ 3 weeks) of PAP on PaCO²and PaO².63,70,–,73All five studies reported a significant decrease in PaCO²and four studies reported a significant increase in PaO². One possible explanation for the single study that showed a nonsignificant change in PaO²could be related to the short duration of therapy (five nights).70There were four studies that studied the benefits of short-term PAP on sleep-disordered breathing.70,–,72,74All four studies reported a significant improvement in AHI and oxygen saturation during sleep.

Long-term benefits of PAP include an improvement in gas exchange, lung volumes, and central respiratory drive to carbon dioxide. Nine studies examined the relationship between long-term PAP (≥4 weeks) and gas exchange (table 3).75,–,83All but one study showed a significant improvement in PaCO²and PaO².

There were four studies that investigated the effects of long-term PAP on FEV¹and FVC (table 5).71,76,77,84Three of the four studies found a significant improvement in pulmonary function. Two of the positive studies did not report a significant change in BMI. This improvement of restrictive ventilatory defect is assumed to be due to a decrease in the premature closure of dependent airways during expiration and to an opening of microatelectasis. Lin only reported a trend toward improved FEV¹with 4 weeks of CPAP.71However, the course of treatment was much shorter than that examined by the other three studies (24–48 weeks).

Five studies evaluated the effects of PAP on central respiratory drive, as measured by carbon dioxide sensitivity, calculated as the change in minute ventilation per unit change in end-tidal carbon dioxide (table 4).41,70,71,77,80Three studies demonstrated a significant increase in carbon dioxide sensitivity, whereas the other two studies reported a trend toward an increase in carbon dioxide sensitivity.

PAP may also reduce mortality in OHS. Two retrospective studies have demonstrated a mortality rate of 13–19% in patients with OHS on PAP throughout a mean period of 4 yr.76,79Through indirect comparison, this mortality rate is lower than the 23% mortality rate reported in patients with untreated OHS at 18 months of follow-up (fig. 5).5 

In summary, short-term (⩽3 weeks) PAP therapy improves gas exchange and sleep-disordered breathing. In addition, long-term (≥4 weeks) PAP therapy improves lung volumes and central respiratory drive to carbon dioxide and lowers mortality. Because of its noninvasiveness and effectiveness, PAP is considered the first-line therapy for OHS.

CPAP failure, defined by a residual AHI ≥5 or a mean nocturnal SpO²<90%, has been documented in some patients with OHS. These nonresponders improved after treatment with bilevel PAP.81 

A recent prospective randomized study compared the long-term efficacy of bilevel PAP versus  CPAP after excluding nine patients with OHS in whom CPAP titration failed.78Two groups of 18 patients with OHS who underwent successful CPAP titration study were randomized to either bilevel PAP or CPAP for 3 months. Both groups experienced a similar degree of improvement in PaCO²and daytime sleepiness. Overall, bilevel PAP was not considerably superior to CPAP if CPAP titration was successful. However, if CPAP titration is unsuccessful, the bilevel PAP should be strongly considered and treatment should be individualized to each patient.3Bilevel PAP should be instituted if the patient is intolerant of higher CPAP pressure (more than 15 cm H²O) or if hypoxemia persists despite adequate resolution of obstructive respiratory events.85 

The various degrees of response to CPAP suggests that OHS is a heterogeneous disease. Patients with OHS who were prescribed bilevel PAP due to failed CPAP trial had significantly reduced FEV¹and FVC versus  CPAP responders.72In contrast, AHI was significantly higher in the CPAP group. It seems that in some patients, severe OSA is a major contributor to OHS and these patients could be successfully treated with long-term CPAP therapy. In other patients with severe restrictive defect secondary to morbid obesity, long-term bilevel PAP may be required.

Approximately 40% of patients with OHS continue to desaturate to SpO²<90% during sleep while on adequate CPAP settings, thereby requiring supplemental oxygen.74Oxygen therapy may act as a double-edged sword, as the administration of high-concentration supplemental oxygen without any form of positive airway pressure therapy may worsen hypercapnia by reducing minute ventilation.86,87In a recent study, some patients with OHS whose condition is stable experienced a significant increase in transcutaneous carbon dioxide tension when administered 100% oxygen compared with room air.88This was associated with a significant decrease in minute ventilation. Therefore, clinicians should administer the lowest concentration of oxygen to OHS patients to avoid worsening of hypercapnia while maintaining optimized oxygenation, particularly in patients with OHS experiencing an exacerbation or recovering from sedatives/narcotics or general anesthesia.89 

Bariatric surgery is now widely accepted as a mainstay treatment in the management of obesity, especially for morbidly obese patients in whom more conservative approaches have failed or who have developed comorbidities. Bariatric surgery improves gas exchange and pulmonary function in OHS. At 1 yr after surgery, PaO², PaCO², FEV¹, and FVC all improved significantly.64,90To better understand the effect of surgical weight loss on OSA, Greenburg et al.  performed a meta-analysis of 12 studies including a total of 342 patients in whom polysomnography pre- and postmaximum weight loss were available.91They found that bariatric surgery led to significant weight loss with a mean reduction in BMI from 55.3 kg/m²to 37.7 kg/m². This robust weight loss was accompanied by a 71% reduction in the AHI from baseline values of 55 events/h (95% CI 49–60) to 16 events per hour (95% CI 13–19). Only 38% achieved cure defined as AHI less than 5. In contrast, 62% of patients had residual disease with the mean residual AHI of 16 events/h. Many of these patients had persistent OSA of moderate severity (AHI ≥ 15 events/h). Thus, although improvements should be anticipated, OSA will not resolve in all patients after surgically achieved weight loss but it should be noted that, despite weight loss, most patients were still obese at the time of the second sleep assessment. Although there is a drastic reduction in OSA severity, some patients still have moderate OSA after maximum weight loss and could therefore benefit from CPAP therapy.91Similarly, 14% of OHS patients still require PAP therapy after weight loss.90Therefore, OHS patients should undergo reevaluation postbariatric surgery before discontinuing PAP therapy.

Bariatric surgery is associated with significant risk. The overall perioperative mortality ranges between 0.5–1.5%.66,92The presence of OSA and extreme preoperative weight are independent risk factors associated with perioperative death and adverse events including venous thromboembolism, surgical reintervention, and prolonged hospital stay.65,66 

Medications that increase respiratory drive have been investigated for the treatment of OHS. Limited evidence was available for two respiratory stimulants: medroxyprogesterone acetate and acetazolamide.

Medroxyprogesterone acetate stimulates respiration at the hypothalamic level.93Its role in OHS is uncertain. An early study reported an increase in PaO²and a decrease in PaCO²in OHS patients treated with medroxyprogesterone acetate.94However, a later study did not demonstrate the same benefits.20Because medroxyprogesterone acetate increases the risk of venous thromboembolism,95administration to OHS patients whose mobility is limited may be unsafe.

Acetazolamide is a carbonic anhydrase inhibitor that increases minute ventilation by inducing metabolic acidosis through increased excretion of bicarbonate by the kidneys. Acetazolamide has been shown to improve AHI, increase PaO², and reduce PaCO²in patients with OSA.96,97More recently, in mechanically ventilated patients with OHS, acetazolamide reduced plasma HCO³⁻and increased hypercapnic drive response.98Given the very limited data on pharmacotherapy and the fact that it is not used widely, we do not recommend it as a mainstay therapy in patients with OHS.

Although there is increased awareness of OSA among anesthesiologists, OHS is often undiagnosed and may greatly increase perioperative risk. A high level of suspicion can lead to early recognition and treatment. Routine screening for hypercapnia in obese patients with OSA may help to identify OHS and allow for referral to sleep medicine for appropriate PAP therapy, modifications in the surgical approach, anesthetic technique, and postoperative monitoring.

Several findings are supportive of OHS, yet the definitive test for alveolar hypoventilation is an arterial blood gas performed on room air during wakefulness. Three clinical predictors of OHS have been suggested: serum HCO³, AHI, and lowest oxygen saturation during sleep.10Increased serum bicarbonate level due to metabolic compensation of respiratory acidosis is common in patients with OHS and points toward the chronic nature of hypercapnia. In a cohort of obese patients with OSA referred to the sleep laboratory for suspicion of OSA, a serum HCO³⁻threshold of 27 mEq/L demonstrated a 92% sensitivity in predicting hypercapnia on arterial blood gas.10To complement the highly sensitive serum HCO³⁻, a highly specific (95%) AHI threshold of 100 was identified. A two-step screening process was proposed, with serum HCO³⁻as the initial test to exclude patients without OHS and then AHI as the second test to improve specificity. In addition, hypoxemia (SpO²<90%, corresponding to PaO²< 60 mmHg)99during wakefulness should lead clinicians to suspect OHS in patients with OSA. In a recent meta-analysis, OSA patients with higher BMI, higher AHI, and more restrictive chest wall mechanics were more likely to develop OHS.100In these patients with OHS, the mean BMI, AHI, FEV¹% pred,and FVC % pred were 39 kg/m², 64 events/h, 71%, and 85%, respectively.

In summary, patients presenting with a high BMI and AHI should alert the physician to screen for OHS. The serum HCO³⁻level is an easy test to screen for hypercapnia before elective surgery. If the serum HCO³⁻is increased or there is presence of hypoxemia by room air SpO²during wakefulness, a confirmatory test with a measurement of arterial blood gases is recommended. Once hypercapnia is confirmed, referral to sleep medicine and further testing, such as pulmonary function testing, chest imaging, thyroid-stimulating hormone, and clinical assessment of neuromuscular strength, is recommended to rule out other important causes of hypoventilation.

An algorithm for screening and perioperative management of OHS is provided in figure 6. For patients at high risk of OHS undergoing major surgery, additional testing for sleep-disordered breathing and pulmonary hypertension should be sought. Perioperative OHS precautions should also be considered.

The three main challenges in OHS are OSA, obesity, and hypoventilation (hypercapnia and hypoxemia). The preoperative assessment of the patient with suspected OHS should begin with a history and physical examination directed to identify comorbidities in OSA and obesity. Patients with untreated OSA and obesity are at greater risk of developing coronary artery disease, congestive heart failure, and diabetes mellitus.56,–,58The frequency and severity of comorbidities increase proportionally with the weight of the patient. Due to the changes in cardiac hemodynamics associated with obesity, severely obese patients may develop a cardiomyopathy characterized by both diastolic and systolic dysfunction.101 

A focused cardiopulmonary examination should be directed at discovering signs of congestive heart failure (rales, S3, jugular venous distension) and pulmonary hypertension (loud P2, right ventricular heave, congestive hepatomegaly). A detailed examination of the airway and sites for venous access should be performed.

The diagnosis of OSA is established by polysomnography. Because most sleep clinics typically have long wait lists for polysomnography, multiple screening tools were developed to evaluate patients at risk for OSA. The STOP-Bang questionnaire was used in preoperative patients.102,103It is a scoring model combining the STOP (s  noring, t  iredness, o  bserved apneas, and increased blood p  ressure) questionnaire and Bang (B  MI ≥35, a  ge more than 50 yr, n  eck circumference more than 40 cm, and male g  ender). A positive screen (three or more questions answered yes) indicates high risk for OSA. A systematic review has suggested using the STOP-Bang questionnaire in the surgical population due to its high methodologic quality and easy-to-use features.104 

Patients identified as at high risk for OSA who present for major elective surgery should be referred to the sleep clinic to establish the diagnosis and to titrate PAP therapy. Therapy should be started during the few days or weeks before surgery. Even with a period as short as 5 days, gas exchange and sleep-disordered breathing can improve significantly with either CPAP or bi-level PAP in OHS.70Ideally, an early consultation would allow the sleep physician adequate time to devise a management plan. If this is not achieved, these “high-risk” patients should proceed to surgery and be managed cautiously as if they were known to have OSA.105,106 

The Lee revised cardiac risk index represents a valuable tool to predict cardiac risk for elective major noncardiac surgery in the general population.107However, other risk factors specifically related to OHS, such as pulmonary hypertension and history of venous thromboembolism, should be considered when evaluating perioperative risk. Most of the data on evaluating surgical risk in severely obese patients are derived from bariatric surgical studies. A mortality risk score for patients undergoing gastric bypass include hypertension, BMI ≥ 50 kg/m², male sex, age ≥ 45 yr, and known risk factors for pulmonary embolism (OHS, previous thromboembolism, preoperative vena cava filter, pulmonary hypertension).67,–,69The obesity surgery mortality risk score stratifies mortality risk into low (zero or one comorbidity), intermediate (two to three comorbidities) and high (four to five comorbidities). Mortality rates were 0.2%, 1.2%, and 2.4% for low, intermediate, and high risk class, respectively.68The most common causes of death were pulmonary embolism (30%), cardiac causes (27%) and gastrointestinal leak (21%).

A 12-lead electrocardiogram should be obtained in patients suspected to have OHS. Signs of right ventricular hypertrophy on electrocardiogram including right-axis deviation and right bundle-branch block suggest pulmonary hypertension. In contrast, a left bundle-branch block on electrocardiogram suggests occult coronary artery disease. Preoperative chest x-ray should also be considered. A chest x-ray showing cardiomegaly or abnormal pulmonary vascularity suggests undiagnosed heart failure and pulmonary hypertension. Furthermore, it could act as a baseline study when evaluating for causes of postoperative respiratory difficulties.

Indications for further cardiovascular testing should be based on patient cardiovascular risk factors and the invasiveness of surgery according to current American Heart Association guidelines.107,108The assessment of functional capacity is of particular importance in obese individuals because cardiorespiratory fitness levels and postoperative complication rate are inversely related to BMI.109,110Functional exercise testing is the preferred evaluation modality but its use may be limited in the severely obese who cannot exercise due to their weight or orthopedic issues. If these patients are undergoing major surgery and present with multiple cardiac risk factors, pharmacologic stress testing and transthoracic echocardiogram may be considered if management will be changed.108 

Studies evaluating postoperative pulmonary complications have generally found no increased risk attributable to obesity.111In a recent retrospective study of a National Inpatient Surgical sample of 110,000 OSA patients, patients with OSA were found to have a higher risk of pulmonary complications than patients without OSA.112Routine pulmonary function tests may not translate into an effective risk prediction for postoperative pulmonary complications in noncardiothoracic surgery.113However, if coexisting chronic obstructive pulmonary disease is suspected in the patient with OHS, spirometry may be considered for diagnosis and subsequent optimization. Arterial blood gas measurements should be obtained to confirm the presence and severity of daytime hypercapnia in obese patients with hypoxemia during wakefulness or an increased serum bicarbonate level.

OSA is a risk factor for both difficult mask ventilation and tracheal intubation.114In addition, patients with severe OSA (AHI ≥ 40 events/h) showed a significantly higher prevalence of difficult intubation than patients with lower AHI.115 

Obesity results in a threefold increase in difficulty with mask ventilation.116Whether obesity increases the difficulty of tracheal intubation is more controversial. A retrospective study of 18,500 surgical patients reported that obesity is a risk factor for difficult intubation.117However, other studies have not found an association between BMI and intubation difficulties.118,119More recently, Kheterpal et al.  identified five risk factors (limited mandibular protrusion, thick/obese neck anatomy, OSA, snoring, and BMI more than 30 kg/m²) as independent predictors of difficult mask ventilation and intubation during anesthesia induction.120This suggests that OHS patients with limited mandibular protrusion are in the highest risk group for airway complications.

During induction of anesthesia, patients with OHS should be placed in the ramp position with elevation of the torso and head. This has been shown to improve the ease of ventilation and glottic view from the neutral position.121Preoxygenation for more than 3 min with a tightly fitted mask can increase apnea tolerance time. Strategies to reduce atelectasis during preoxygenation include optimal patient positioning and the application of PAP. When the patient is in a 25° head-up tilt position, PaO²was higher and the time to desaturate to 92% during apnea was increased when compared with preoxygenation while supine.122The application of CPAP and positive end-expiratory pressure during preoxygenation also achieved a higher oxygen tension and longer time to desaturation.123A variety of airway adjuncts and skilled anesthesiology assistance should be made available in advance. Awake intubation should be considered when mandible advancement, neck extension, or mouth opening is limited in OHS patients.

Patients with OHS are sensitive to the respiratory depressant effects of anesthetic agents due to the propensity of airway collapse, sleep deprivation, and blunting of physiologic response to hypercapnia and hypoxemia. A semiupright or lateral position is recommended at the end of surgery for better oxygenation and airway maintenance.124Rapid emergence from anesthesia is preferred because tracheal extubation should be performed only after the patient is fully conscious. A systematic analysis of the literature comparing postoperative recovery after propofol, isoflurane, desflurane, and sevoflurane-based anesthesia in adults demonstrated that early recovery was faster in the desflurane and sevoflurane groups.125Another strategy to accelerate emergence is to decrease volatile anesthetic requirement and minimize washout time from fat/muscle by using other short-acting anesthetic adjuvants, such as remifentanil, or a combined general-regional anesthetic.106 

Opioid-induced ventilatory impairment (OIVI) is a term describing opioid-induced central respiratory depression, decreased level of consciousness, and upper airway obstruction. The end-result of OIVI is a decrease in alveolar ventilation. The incidence of OIVI after major surgery varies with the different routes of opioid administration. The incidence of decreased respiratory rate was 0.8, 1.2, and 1.1% for intramuscular, intravenous patient-controlled analgesia and epidural analgesia, respectively.126The incidence of oxygen desaturation was 37, 11.5, and 15.1% for intramuscular, intravenous patient-controlled analgesia and epidural analgesia, respectively.126Patients with OHS could be at significant risk for OIVI due to their susceptibility to upper airway obstruction, depressed central respiratory drive, and impaired pulmonary mechanics. An opioid-sparing analgesic regimen, including local anesthetic infused nerve block catheters and nonopioid adjuncts (acetaminophen, nonsteroidal antiinflammatory drugs), should be considered in these patients.

Improved postoperative monitoring is key in reducing the risk of OIVI. Patient-specific, anesthetic, and surgical factors determine the requirements for postoperative monitoring. OHS patients undergoing major surgery who require high doses of postoperative opioid should be monitored with continuous oximetry. Recurrent respiratory events in the postanesthesia care unit, including apnea for ≥10 s, bradypnea of less than 8 breaths/min, pain-sedation mismatch, or desaturations to less than 90%, can be used to identify patients at high risk of postoperative respiratory complications.127Recently, Macintyre et al.  proposed that sedation level is a more reliable sign of OIVI than respiratory rate128because multiple reports suggest that OIVI is not always accompanied by a decrease in respiratory rate.128,–,130Thus, sedation scoring systems should be used postoperatively to recognize OIVI so that appropriate interventions are triggered. In patients with OHS requiring high doses of postoperative opioids, sedation monitoring should be considered every 1 to 2 h for the first 24 h.131 

There is limited evidence demonstrating a reduction in postoperative complications with PAP in patients with OHS. However, a case series of 14 patients with OSA suggested that the use of CPAP continuously for 24 to 48 h postextubation may reduce the risk of postoperative complications.132In addition, PAP was found to decrease postextubation respiratory failure in severely obese patients admitted to the intensive care unit (absolute risk reduction of 16%).133Subgroup analysis of patients with hypercapnia showed reduced hospital mortality in the PAP group compared with the control group. Other potential benefits of perioperative CPAP include reduced hemodynamic fluctuations and arrhythmia related to hypoxemia.

A recent case report described a 59-yr-old patient with OHS who suffered multiple orthopedic injuries secondary to a mechanical fall (a simple fall not associated with any cardiac or neurologic event [e.g. , heart attacks or stroke]).134In the emergency department, she received opioid analgesics and subsequently developed severe hypoxemia refractory to naloxone. The initiation of bilevel PAP promptly restored adequate ventilation.

In summary, patients with OHS who were previously on PAP should resume therapy as soon as possible postoperatively. In patients suspected to have OHS experiencing postoperative ventilatory impairment, PAP should be considered as a rescue device. Based on the available literature, patients with OHS typically require an inspiratory PAP and the expiratory PAP of 16–18 cm H²O and 9–10 cm H²O, respectively, to achieve adequate resolution of upper airway obstruction and to improve ventilation. If previous PAP titration has not been performed or the data are unavailable to the anesthesiologists, bilevel PAP can be empirically set at these pressures in patients suspected of having OHS.

OHS is a disease entity that anesthesiologists need to have a thorough understanding of. The prevalence of OHS is estimated to be 0.15–0.3% in the general population3and 8% in patients undergoing bariatric surgery.17Patients with OHS have a syndrome distinct from mere obesity and OSA, as indicated by the severe upper airway obstruction, restrictive chest physiology, blunted central respiratory drive, and pulmonary hypertension. The mainstay of therapy for OHS is PAP therapy because it improves gas exchange, lung volumes, and sleep-disordered breathing, and reduces mortality.

Perioperative management begins with a high index of suspicion for OHS in the morbidly obese patient. Screening questionnaires such as the validated STOP-Bang questionnaire can identify patients at high risk of OSA. This screening tool can be further complemented by the presence of low SpO², increased PaCO², and serum HCO³⁻level to identify patients at high risk of OHS. Before major elective surgery, these patients should be referred to sleep medicine for polysomnography and PAP titration. An echocardiogram should be considered to assess right ventricular function and pulmonary hypertension. Perioperative precautions of OHS include prudent airway management, rapid emergence, monitoring for ventilatory impairment and early resumption of PAP therapy. Future research should focus on the perioperative strategies of screening, monitoring and treatment of OHS and associated complications.

The authors acknowledge the help of Marina Englesakis, B.A., M.L.I.S., Information Specialist, Toronto Western Hospital Health Sciences Library, University Health Network, Toronto, Ontario, Canada.