Data Quality

• Little comparative or generalizable patient outcome data on diagnostic and treatment methods for OSA in adults are available.
• Evidence in the medical literature shows that diagnosis and treatment of OSA in adults vary greatly between facilities. Interpretation of data was limited due to small study size and lack of controls or randomization of many studies.
• Few cost-effectiveness analyses on any of the diagnostic methods or OSA treatment were identified in a search of the medical literature.

Diagnosis

• OSA is a complex condition which may be exacerbated or relieved by a number of physical and physiological factors. As a result, particular attention must be paid to informing patients on what is and is not known about the effectiveness of diagnosis and treatment strategies for OSA.
• Professional groups recommend that patients with suspected OSA undergo a complete physical and otolaryngologic examination including medical history and determination of body mass index (BMI).
• Overnight polysomnography (PSG) is considered the gold standard of sleep apnea diagnosis. There is controversy, however, on whether the costs of overnight PSG provide substantially better diagnostic information compared to alternatives such as in-home sleep studies.

Oversight of Sleep Laboratories

• Although the American Academy of Sleep Medicine (AASM), formerly the American Sleep Disorders Association (ASDA), accredits sleep laboratories and centers, there is no centralized oversight of sleep facilities. A clinic, hospital, or other facility may establish a sleep disorders program with no review or oversight by a professional group. Thus, determining the quality and standard of care of a particular program is difficult for professionals and even more difficult for patients.

Treatment Selection

• Guidelines for diagnosis and treatment of OSA in adults have been developed by several professional organizations; however, there is no clear consensus on patient selection criteria for sleep studies and treatment options.
• Selection of treatment(s) for individual OSA patients should be based upon balanced consideration of disease severity and site(s) of obstruction, subjective symptoms, risks of morbidity and mortality, and patient choice.
• Treatment effectiveness is variable and dependent on patient needs. It is believed that treatment must be evaluated over time for good patient outcome.

Medical and Behavioral Treatments

• Several medical and behavioral treatments for OSA in adults are available including change in sleeping position, drug therapies, and use of dental devices. This assessment focuses on two approaches - weight reduction and nCPAP.
• Data in the literature indicate that weight loss in obese adults may be associated with improvements in the objective and subjective signs of OSA; however, many patients return to obese body weight over time.
• For patients who are inappropriate candidates for behavioral treatments, nCPAP is the mainstay of therapy. Compliance rates are variable due to discomfort and side effects of nCPAP. Compliance is highest among patients who experience improvement in levels of sleepiness, alertness, and other daytime symptoms.

Surgical Treatments

• Surgery may be appropriate for patients who cannot comply with or are not appropriate candidates for conservative therapies or nCPAP alone. The type of surgery performed should be based upon the specific pathophysiology of a patient's condition.
• Careful and thorough preoperative examination by radiography, imaging, and direct visualization is needed to identify the airway obstruction site(s) and to select the appropriate surgery.
• Uvulopalatopharyngoplasty (UPPP) is most efficacious for patients with oropharyngeal obstruction. The failure rate increases when it is performed alone in the presence of retrolingual obstruction. Response rates to UPPP decrease over time.
• When patients have multiple sites of obstruction or craniofacial skeletal abnormalities, UPPP combined with mandibular osteotomy with genioglossus advancement and hyoid myotomy (GAHM), or mandibular osteotomy with maxillomandibular osteotomy and advancement (MMO), or MMO alone, are more effective than UPPP alone for decreasing the respiratory disturbance index (RDI), excessive daytime somnolence (EDS), and snoring.
• The efficacy of midline glossectomy (MLG) following failed UPPP is relatively low and is variably affected by body weight and OSA severity; the long-term outcome after MLG is unknown.
• While data indicates laser-assisted uvulopalatoplasty (LAUP) can improve snoring and EDS, no evidence based guidelines suggest LAUP is effective or safe as a treatment for OSA.
• Tracheostomy is an effective treatment for OSA. Use of this procedure is limited to patients for whom other treatment options have failed due to associated morbidity and negative cosmetic effects.

Risk Factors and Occurrence of OSA

Data from the Wisconsin Sleep Study are often used to estimate the prevalence of OSA. This study of 602 men and women between the ages of 30 and 60 estimated 2% of middle-aged women and 4% of middle-aged men in the United States have OSA if a respiratory disturbance index of 5 or greater and excessive daytime somnolence (EDS) are used as minimal criteria.⁹ These figures, as applied to the Minnesota population, indicate that an estimated 40,000 to 80,000 Minnesotans between the ages of 30 and 60 may have OSA. Research indicates OSA is higher among adults age 65 and older. One self-administered survey showed 4% of women and 13% of men age 65 and older had episodes of sleep apnea.¹⁰ These figures, as applied to the Minnesota population, indicate that an estimated 31,000 men and 13,600 women age 65 and over may have OSA.

In addition to the healthy elderly and males, incidence of OSA appears to be higher in adults of any age who have coronary artery disease or hypertension. Obesity is a significant risk factor for OSA.⁹ However, the majority of patients with OSA are not morbidly obese.¹¹ Other risk factors for OSA include hypothyroidism, adenotonsillar hypertrophy, and nasal obstruction.¹²

Health Effects

Periods of apnea or hypopnea result in frequent arousal during sleep, abnormally negative intrathoracic pressures, disrupted sleep architecture, and loud snoring. These contribute to EDS, which is associated with deficits in neurocognitive and psychosocial functioning, depression, impotence, morning headaches and decreased work productivity.^3,13-15 Daytime sleepiness has been associated with an increased risk for fatigue-related accidents, including motor vehicle crashes.¹⁶ Population studies have shown strong associations between sleep-disordered breathing, including sleep apnea, and an increase in motor vehicle accidents.^17,18 No studies analyzing the morbidity, mortality, and economic implications of these accidents were found in the medical literature.

Although many conditions have been associated with OSA, no causal links between untreated OSA and these conditions have been established. Conditions associated with untreated OSA include systemic and pulmonary hypertension, myocardial infarction, cerebrovascular disease, cardiac arrhythmias, and sudden death. Patients with OSA appear to have diminished long-term survival with a higher likelihood of cardiovascular disease-related morbidity and mortality. Early diagnosis and treatment of OSA may reduce the risk of cardiovascular disease and other morbidities in these patients.^{1,5,7,14,19-23}

The National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health (NIH) has initiated a multicenter cohort study, the National Sleep Heart Health Study, to determine whether cardiovascular disease and other conditions are a consequence of sleep-disordered breathing, including OSA. The study has enrolled approximately 6500 subjects from 6 investigative centers throughout the United States. The University of Minnesota is one of the field centers. Data from the study will assist in determining whether sleep-disordered breathing is directly associated with an increased risk of coronary heart disease events, stroke, increased blood pressure, and all-cause mortality. Follow-up studies on the entire cohort or subsets of the cohort will help to determine whether sleep disordered breathing is associated with a variety of conditions, including recurrent heart disease, stroke, daytime sleepiness, blood pressure, and cardiac arrhythmias. Final results from this study are not expected for several years. Interim findings will be published in the next 2-3 years.^24,25

Modified Mueller Maneuver

The Modified Mueller Maneuver (MMM) is used to identify a specific site of pharyngeal obstruction. With the patient sitting or lying in the supine position, the upper and lower pharynx are visualized by a fiberoptic nasopharyngoscope at end expiration. The patient then attempts to inspire while the mouth and nose are occluded. Obstruction or collapse that occurs during negative inspiration, while not a direct correlate for the site of obstruction during sleep, is believed to indicate the most likely site(s) of obstruction during sleep. The degree of collapse is graded separately at the retropalatal area, the lateral pharyngeal walls, and the tongue base as follows: 0 for no collapse; 1+ for a 25% reduction in cross-sectional area; 2+ for a 50% reduction in area; 3+ for a 75% reduction in area; and 4+ for complete obstruction.^2,7,15,21,39

Accurate identification of the exact site(s) of airway obstruction in OSA patients is complicated by the fact that obstruction occurs during sleep and is difficult to assess by procedures in which the patient is awake.⁷ A limitation of the MMM for diagnosis and postoperative assessment of OSA is that it is a subjective procedure during which the observer cannot readily be blinded since the surgical changes would be evident. In addition, only the airway lumen can be observed, not the surrounding soft-tissue structures. The data provided by this procedure has not yet been proven to accurately predict surgical response.^2,7,39,40

Radiograph

Cephalometric radiographs are used to screen patients for craniofacial skeletal abnormalities prior to treatment. Treatment efficacy is also assessed by a comparison of pre- and postoperative standard cephalometric radiographs. Outlines of the posterior pharyngeal wall, soft palate, base of the tongue, hard palate, maxillary incisors, mandible, and hyoid are traced, digitized by a scanner, and the images are analyzed using a computer program. Measurements are made of the posterior airway space, minimal posterior airway space, posterior uvula airspace, tip of the incisor-to-tongue base distance, mandibular plane-to-hyoid distance, and cross-sectional area of the third cervical vertebrae. The preoperative values are compared with scaled postoperative measurements.^15,39

Weight Control

The role of body weight in development of OSA and the effectiveness of weight loss in improving OSA symptoms is complex and controversial. Obesity is a risk factor for OSA. However, daytime sleepiness with associated decreases in activity and other metabolic factors associated with OSA may promote weight gain.^42,43 OSA may progress in patients who demonstrate weight gain. Increased adipose tissue deposits in the lateral pharyngeal fat pads and the increased muscle mass associated with weight gain are believed to cause airway obstruction and OSA. The exact mechanisms whereby weight loss results in improvements in OSA are unknown. Potential mechanisms by which weight loss may improve OSA include changes in upper body structure and decreases in upper airway collapsibility.^1,11,44-46

Clinical studies that examined the effects of controlled diet and weight loss on OSA consisted of two prospective, non-randomized controlled trials, two uncontrolled case series, and one retrospective non-randomized controlled trial. Follow-up times ranged from 5 months to 7 years in study populations consisting of 13 to 127 patients with mild-to-moderate or moderate-to-severe OSA. In these studies, AHI/RDI, LSAT, and/or hypertension improved following weight loss, as did sleep architecture, EDS, and the Pcrit (nasal pressure below which the airway occludes). In one study, however, the patients continued using nCPAP, which may have confounded the effects of weight loss on OSA improvement. In one case series, weight loss had no additional effects on improvement in AHI or snoring achieved by nasal spray use and sleeping in the lateral position; however, in patients who lost >3 kg, snoring improved significantly. One long-term retrospective analysis compared the efficacy of tracheostomy and weight loss in patients with moderate-to-severe OSA. Results compared at 7 years showed 18.3% of tracheostomized patients had greater or unchanged EDS. Among weight loss patients, 81.9% had greater or unchanged EDS. Total and vascular disease-related mortality rates were significantly lower among the tracheostomized patients compared with the weight loss patients at the same follow-up (2.8% vs. 17.3%, and 1.4% vs. 11%, respectively).^11,19,20,47

In one long-term study on the effects of nCPAP or weight loss on OSA, multivariate analysis showed that a change in BMI was the only variable significantly and independently associated with reduced blood pressure in OSA patients with hypertension.²⁰ Weight levels must be maintained to be beneficial to the patient. Many patients are unable to maintain non-obese weight levels over the long term.⁴⁸

Nasal Continuous Positive Airway Pressure (nCPAP)

Use of nCPAP while sleeping, is the mainstay of treatment for patients who fail conservative therapies and is highly effective. While sleeping, the patient wears a plastic mask or nasal prongs connected to a positive pressure device with provisions for adjunct therapy with supplemental oxygen or humidification of inspired air. Blowers generate a bias flow to deliver positive pressure. Flow rates ranging from 20 to 60 L/min generate pressures of 2 to 20 cm H₂O pressure. With monitoring by PSG, the pressure level is titrated to a level that restores airway patency, improves arterial oxygenation, and ensures uninterrupted sleep. Patients must be followed regularly to assess treatment efficacy and to verify compliance.

The effects of nCPAP are only maintained while the device is being used. A number of patients find nCPAP uncomfortable and restrictive. As a result, long-term compliance is variable.⁴⁵ Subjective reports suggest that long-term compliance rates range from 50% to 85%, while covert monitoring has revealed compliance rates of less than 50% for all night usage. The most consistent factor associated with improved compliance is improvement in levels of sleepiness, alertness, and other daytime symptoms recognized by patients or their families. ^45,49-51

The efficacy of nCPAP was investigated in 3 randomized controlled trials, 2 non-randomized controlled trials, and 1 quality of life case-series analysis reviewed for this assessment. Follow-up times ranged from 3 to 6 months to a mean of 512 days, whith populations comprised of 27 to 126 patients with mild-to-severe OSA. Response rates, defined as improvements in AHI, AI, LSAT, and/or ODI₄, ranged from 62% to 100%. EDS, snoring, and sleep architecture were generally improved when nCPAP was used. In one study, AI and hypertension improved in 29% of hypertensive patients with OSA during a follow-up of > 1 year; BMI was the only variable that significantly and independently predicted improvement. In another study comparing survival in nCPAP and UPPP-treated patients, 2.4% of nCPAP patients died compared with 4% of UPPP-treated patients during 6 years of follow-up; however, the probability of 5-year survival was not significantly different between the nCPAP-treated patients (0.95 ± 0.03) and the UPPP patients (0.94 ± 0.02). Survival was not significantly affected by postoperative AI severity or arterial SaO₂ levels. In a randomized controlled trial, significantly more nCPAP treated patients had a normal ODI₄ at 1 year (100%) compared with conservatively treated patients, i.e., those advised to lose weight (5%). Relapse rates were not evaluated in these studies.^{8,14,20,38,50} The quality of life case-series analysis found that after 8 weeks of nCPAP therapy in 29 patients with PSG documented sleep apnea, quality of life measures were improved. Based on administration of a standardized pretreatment and post-treatment survey, vitality improved by 75%, social functioning by 90%, and mental health by 96%.⁵²

Complications associated with nCPAP use include: local skin irritation and nasal and throat dryness (50%), nasal congestion or rhinorrhea (25%), and eye irritation (25%). Rare complications include pneumocephalus, bacterial meningitis, conjunctivitis, massive epistaxis, and atrial arrhythmia.⁴⁹ In the studies reviewed, the following complications were reported: rhinorrhea (54%), dry throat and nose (15%), mask use discontinuation due to discomfort (15%), disturbed by machine noise (7.7%), nasal symptoms (29%), and sense of suffocation (29%).^8,38,50

Part-time or inconsistent use of the device does not adequately control OSA. The mask pressure required to suppress apneas and hypopneas and their associated O₂ desaturations and arousals varies with body position and sleep stages and does not remain constant throughout the night. Patients who fail or cannot comply with medical treatment may be candidates for surgical treatment. ^{1,5,8,12-15,38,50}

Uvulopalatopharyngoplasty (UPPP)

Introduced in 1980, UPPP corrects retropalatal obstruction and increases the size of the oropharyngeal airway. This surgical procedure involves excision of the tonsils, if present; excision of the uvula and posterior portion of the soft palate, and excessive pharyngeal tissue; and the removal and reshaping of the posterior and anterior tonsillar pillars in the lateral pharyngeal walls.^1,2,21 UPPP may be combined with other pharyngeal surgeries, or with nasal septoplasty or tonsillectomy if there are multiple sites of obstruction. Other pharyngeal and tongue base reduction surgeries improve patency in other areas of the airway. This and other pharyngeal surgeries are performed under general anesthesia and require hospitalization.

Careful preoperative examination is needed to identify the airway obstruction site(s) and to select the appropriate surgery. Studies of UPPP have been unable to determine reliable preoperative criteria to predict the success of the procedure, and the exact mechanisms for treatment success or failure following UPPP are unknown.⁵³ Studies evaluating the efficacy and safety of UPPP for mild-to-severe OSA consisted of two prospective and one retrospective case series, one nonrandomized controlled trial, and one randomized controlled trial, with follow-up times ranging from 3 months to 8 years and populations of from 10 to 149 patients. The response rates, with respect to improvements in RDI, AHI, AI, and/or LSAT, ranged from 18% to 81.4% and varied according to the criteria for response. Response rates were lower at later follow-ups compared with earlier postoperative follow-ups. In one study, the response rate decreased from 64% at 6 months to 48% at 4 to 8 years.^{1,5,12,14,19,22,23,37,54}

Of note was the significant rate of relapse over time after UPPP. In up to 37% of patients who initially responded, treatment failed at 2 to 8 years after surgery. In the randomized controlled trial, the median ODI₄ was abnormal; however, a significantly higher percentage of UPPP-treated patients (39%) had a normal ODI₄ compared with conservatively treated patients treated with conservative treatment (7%).³⁸

Data from most, but not all, studies showed significant improvement in EDS, snoring, and sleep architecture. Similarly, in some studies, a lower preoperative BMI and/or lower OSA severity were significantly predictive of a successful outcome whereas, in other studies, these variables had no effect on surgical response. In the non-randomized controlled study that examined long-term survival in UPPP and nCPAP patients, death rates during 6 years of follow-up were higher in UPPP patients who were not followed with PSG, but were not significantly different between UPPP- and nCPAP-treated patients. There were more cardiovascular deaths in the UPPP group, but, due to the small numbers, it was not possible to point to OSA as the cause.^{5,14,21,23,37,38}

The results of UPPP on OSA are variable. Most studies report a response in approximately 50% of patients. In a meta-analysis of 37 published studies, the overall response to UPPP was 40.7% when the criteria for success was defined as > 50% reduction in AI or RDI and a postoperative AI of < 10 or an RDI of < 20. The subjective improvement following UPPP does not always correlate with objective indicators of success determined by PSG, such as a decreased RDI or AI. OSA patients require regular follow-up with objective tests after treatment since silent apneas may persist despite symptomatic improvement. It has been shown that long-term survival after UPPP is poorer in patients who are not followed with PSG compared with those who are monitored. ^{1,6,7,14,21,23,37,40,41,45,54}

UPPP appears to be the most successful in patients with mild OSA with airway obstruction limited to the soft palate rather than the base of the tongue.⁵³ Patients with severe OSA generally respond poorly after single surgical procedures such as UPPP. The high failure rate for the surgery may be attributable to the presence of upper airway obstruction at multiple sites in the pharynx. The failure rate increases when UPPP is performed alone in the presence of retrolingual obstruction. Treatment with UPPP is more likely to fail in patients with severe OSA, retrolingual obstruction, and/or who are obese. UPPP may fail in obese patients due to the compression of the airway by oropharyngeal fat deposits. In non-obese patients, UPPP failures are often related to the presence of obstruction at the level of the hypopharynx or to the existence of a skeletal abnormality that is not corrected by the surgery. ^{2,21,23,37,38,39,40,41,54}

While UPPP may eliminate snoring and other OSA symptoms, the underlying airflow limitations may remain. There has been some concern that, if symptomatic improvement occurs in the absence of a sufficient reduction in objectively measured outcomes such as the RDI, the patient may continue to experience severe apneas that will remain untreated. This could potentially lead to a higher rate of cardiovascular complications.^{14,38,41,45,55} UPPP can increase the amount of air leaking around the mouth during nCPAP therapy. As a result, success of subsequent nCPAP therapy may be compromised.⁵⁶

This procedure carries a number of safety concerns. In the studies reviewed for this assessment, 20% of patients undergoing UPPP required postoperative tracheostomy and 40% required reoperation for infection. Other complications included: velopharyngeal insufficiency (7.3% to 15%); myocardial infarction (6%); transient ischemic attack (6%); bleeding (5%); deep vein thrombosis (2.4%); cricopharyngeus spasm (11%); and mouth floor hematoma (11%).^5,38

Pharyngeal Surgeries

UPPP alone does not address hypopharyngeal airway obstruction in all instances. When patients have multiple sites of obstruction, collapse of the oropharyngeal and hypopharyngeal airways, or craniofacial skeletal abnormalities, UPPP combined additional pharyngeal surgeries, is more effective for decreasing the RDI, EDS, and snoring. Hypopharyngeal surgery may be used alone for patients whose airway obstruction is limited to the hypopharynx or as a second-line approach in patients who have obstruction of both the oropharyngeal and hyopharyngeal airways and who fail UPPP.^7,21,41,54

In a retrospective study that examined the frequency of complications in upper airway surgery which required intensive care monitoring after 125 surgical procedures for OSA, the following complications were reported: airway complications (0.8%), O₂ desaturation to < 90%, bleeding (3.2%), severe hypertension (4.8%), mild hypertension (5.6%), wound complications (7.2%), urinary retention (4%), premature ventricular contractions (0.8%), and chest pain (0.8%).⁵⁷ Surgeries such as inferior sagittal mandibular osteotomy with genioglossus advancement and hyoid myotomy with suspension or advancement (GAHM) and mandibular or maxillary osteotomy and advancement (MMO) are used to correct retrolingual or hypopharyngeal obstruction by moving the base of the tongue forward and increasing the size of the posterior airway space. Midline glossectomy (MLG) removes redundant tissue at the base of the tongue.^{1,2,15,21,40,41,54}

Genioglossus Advancement and Hyoid Myotomy (GAHM)

In GAHM, the genial tubercle, which serves as the anterior attachment of the tongue, and the hyoid bone are advanced following a limited mandibular osteotomy. The hyoid is fixed to the anterior margin of the mandible or, in a more recent modification, fixed to the thyroid cartilage. A partial GAHM consists of the same procedure, but the hyoid is not suspended or advanced. Dental occlusion remains the same after either procedure.

UPPP with GAHM is typically performed in patients with multiple sites of pharyngeal obstruction and who have moderate-to-severe OSA. The studies that evaluated the efficacy and safety of this procedure consisted of two prospective and two retrospective case series, with mean follow-up times ranging from approximately 4 to 9 months, and were comprised of from 9 to 239 patients. Response rates, with respect to improvements in RDI/AHI, AI and/or LSAT, ranged from 57.1% to 100% and varied with the criteria for response. In one study, 11% of patients demonstrated clinical worsening indicated by a fall in SaO₂. EDS, sleep architecture, and snoring were generally better after UPPP and GAHM. In one study, the response rate was lower for patients with more severe OSA. In another study, AI and BMI were not significantly related to surgical outcome; however, there was a paucity of data on variables that predicted surgical response. Long-term relapse, response, and survival rates were not investigated in these studies.^7,21,39,41

In the studies reviewed, the following complications were reported: bleeding (8.3% to 25%); odynophagia (8.3%); change in taste (8.3%); gingivolabial sulcus incision dehiscences (15.6%); gingivolabial sulcus wound infection (3.2%); transient paresthesia of the mandibular incisors (100%); screw extrusion and mandibular bony fragments (6.3%); and moderate ecchymosis and edema of neck and facial skin (3.2%).^40,54

Maxillomandibular Osteotomy and Advancement (MMO)

The mechanism by which MMO improves upper airway diameter is unknown.² The most common complication in MMO is transient anesthesia of the cheek and chin, which occurred in 100% of patients in one safety study. Eighty-seven percent of these cases were resolved within 6 to 12 months. No postoperative bleeding, infection, skeletal relapse, or relapse of OSA symptoms occurred.²¹

In MMO, the advancement of the mandible and hyoid bone results in a greater advancement of the pharyngeal muscles and base of the tongue and, thus, greater airway expansion. The maxilla and mandible are advanced simultaneously via Le Fort I maxillary and sagittal-split mandibular osteotomies. MMO is either performed after a failed UPPP or as the sole procedure for OSA patients with mandibular skeletal deformities associated with a narrowed posterior airway space and tongue base obstruction.^{1,2,7,21,41,54}

In one retrospective case series reviewed on MMO (mean follow-up, 9 months) in 91 patients with moderate-to-severe OSA who had failed UPPP and GAHM (n=24), who had MMO for a craniofacial skeletal abnormality (n=7), or who had failed UPPP (n=60), the response rates with respect to improvements in RDI and LSAT were 100%, 100%, and 96.7%, respectively. Sleep architecture was improved; however, EDS, snoring, and possible predictors of response were not evaluated. No long-term response, relapse, or survival rates were presented in either the study on MMO or those on MLG.^21,40,54

Midline Glossectomy and Epiglottidectomy (MLG)

MLG with tonsillectomy and epiglottidectomy is used alone or after UPPP failure to reduce the size of the tongue base. With the patient under general anesthesia given through a tracheostomy, laser is used to resect a segment of lingual tonsil and tongue base from approximately 1 cm anterior to the circumvallate papilla, extending posteriorly into the vallecula to the base of the epiglottis. A tissue section measuring 2.5 to 3.5 cm in width and 2 cm in depth is removed. Partial epiglottidectomy removes excess epiglottic tissue. An advantage of this procedure as opposed to GAHM is that it does not require osteotomy, fixation, or immobilization for bone healing. Obstructions of the nasopharynx are corrected by standard nasal septoplasty with inferior nasal turbinate resection.^1,7,40,54

MLG following failed UPPP has a low and variable success rate, with response rates ranging from 25% to 41.7% depending on the criteria used. While it appears that non-obese patients have a better outcome after MLG, the reason why remains unclear. Patients who respond may have specific hypopharyngeal anatomic features that are corrected by the procedure. The efficacy of this surgery requires further analysis in prospective studies.^40,54

In one prospective and one retrospective case series that examined the efficacy and safety of MLG and epiglottidectomy, each comprised of 12 patients with severe OSA and with follow-up ranging from a mean of 73 days to 15 months, from 25% to 41.7% of patients demonstrated an improved RDI. Sleep architecture improved in one study but was unchanged in the other. Similarly, the response rate was significantly better in less obese patients in one study, while preoperative BMI, RDI, and LSAT had no significant effect on surgical outcome in the other study. No long-term response, relapse, or survival rates were presented in the studies reviewed on MLG.^21,40,54

Laser-Assisted Uvulopalatoplasty (LAUP)

LAUP was developed as a treatment for snoring and may be performed in a physician's office under local anesthesia and light sedation. LAUP improves airway patency through removal of part of the uvula and soft palate bilaterally. Since it does not require general anesthesia or hospitalization, LAUP is less costly than UPPP and other pharyngeal surgeries. However, there are no evidence based guidelines which suggest LAUP is effective or safe as a treatment for OSA.

In 1997, the American Medical Association (AMA) published a Diagnostic andTherapeutic Technology Assessment (DATTA) on the efficacy and safety of LAUP for the treatment of simple snoring and OSA. A panel of 61 physicians in otolaryngology and head and neck surgery, neurology, and pulmonary medicine answered a survey regarding the effectiveness and safety of the surgery. Otolaryngologists on the panel rated the safety and effectiveness of LAUP for simple snoring as established. Overall, panel members defined the role of LAUP in OSA as primarily investigational.¹⁵

Clinical studies that investigated the efficacy and safety of LAUP consisted of one randomized controlled trial, one prospective case series, and one retrospective case series, with follow-up ranging from 77 days to > 4 months.^5-7 The studies included 34 to 56 patients with mild-to-moderate or mild-to-severe OSA. Response rates, which varied widely according to the criteria used, ranged from 33% to 65.8% and were lower for the stricter criteria. In the two studies that evaluated symptoms and sleep patterns, EDS, snoring, and sleep architecture were improved; however, daytime headaches and mental status were unchanged.^6,7

There is a paucity of data on specific variables that predict surgical outcome, as well as on the long-term durability of response after surgery. Only one of the studies examined variables that might affect outcome.⁵ In this study, LAUP responders had significantly lower AI than non-responders, but BMI, RDI, and LSAT were not significant predictors of surgical outcome. Additional long-term, randomized, controlled trials, with direct comparisons to other types of surgery and medical treatments are needed before definitive conclusions regarding the role of LAUP in OSA therapy can be reached.^6,7,15

Due to the problem of immediate and long-term postoperative OSA in some patients treated by LAUP, patients who undergo this procedure must be followed with PSG for extended periods of time. There is a need for further investigation of the long-term durability of response and the rate of relapse after this procedure.^5-7,58

In the studies reviewed, throat dryness and tightness occurred in all patients immediately following surgery. Bleeding occurred in 1.8% to 5.3% of patients. Other complications included oral candidiasis (5.3%), velopharyngeal insufficiency (2.6%), and vasovagal episode (1.8%).^5-7 Patients with marginal airways are at risk for postoperative airway complications that cannot always be predicted before surgery.⁵⁵

Tracheostomy

In this surgical procedure, a percutaneous opening is created in the trachea distal to the pharynx and larynx, bypassing the area of pharyngeal obstruction. The diameter of the stoma is maintained by stenting and insertion of a rigid or semirigid hollow tube that extends to the surface of the body. At bedtime, the external end of the tube is unplugged to allow breathing to take place through the tube. During wakefulness, the tube is kept plugged. This redirects airflow to the nose, mouth, pharynx, and larynx.¹ Tracheostomy is an effective treatment for OSA; however, due to its negative cosmetic effects and associated morbidity, the procedure is generally limited to emergency situations and when other treatment options are unacceptable.

Indications for Polysomnography and Related Procedures (1997)

The recommendations of the AASM for PSG, referring to full overnight PSG unless otherwise indicated, are briefly summarized as follows:³⁰

• PSG is routinely indicated for the diagnosis of sleep-related breathing disorders. A cardiorespiratory sleep study may be substituted for patients in a high-pretest-probability group, providing that overnight PSG is performed for patients in this group who test negative by the cardiorespiratory sleep study.
• Overnight PSG is indicated for nCPAP titration in patients with sleep-related breathing disorders.
• A preoperative clinical evaluation including PSG or a cardiorespiratory sleep study is routinely indicated to identify the presence of OSA in patients prior to LAUP.
• Follow-up PSG or a cardiorespiratory sleep study is routinely indicated for the assessment of treatment results following surgery in patients with moderate-to-severe OSA and in those whose symptoms recur after apparently successful surgery.
• Follow-up PSG is routinely indicated for nCPAP-treated patients who lose or gain substantial amounts of weight, and when the clinical response is insufficient or symptoms recur after an initially positive response.
• Follow-up PSG or a cardiorespiratory sleep study is not routinely indicated in patients treated with nCPAP with continued symptomatic improvement.
• A multiple sleep latency test is not routinely indicated for most patients with sleep-related breathing disorders. Subjective assessment of EDS should be obtained routinely.
• The routine use of unattended cardiorespiratory studies or unattended PSG cannot be supported until such studies have been validated.

Surgery of the Upper Airway (UPPP, GAHM, MLG, and Tracheostomy)

Recommendations made by the association are as follows: the presence and severity of OSA must be determined by the evaluation of clinical symptoms and signs and by PSG before surgery is performed; this process identifies patients at risk for complications due to OSA and provides baseline measurements from which to determine surgical efficacy. The goal of OSA surgery is to alleviate clinical signs and symptoms and to normalize sleep quality, the AHI, and SaO₂ levels.¹²

The recommended therapy for moderate-to-severe OSA is nCPAP; this treatment may also be preferred by patients with mild OSA who are symptomatic. Surgery is indicated in patients whose OSA is related to a specific, surgically correctable anatomical abnormality, and may be indicated for patients who have failed or rejected more conservative treatments, who are medically stable, and who desire the surgery. However, in the absence of the identification of a specific anatomical defect, the optimal way to predict the patient's response to site-specific surgery has not been determined. Any one surgical procedure may not adequately eliminate the clinical signs and symptoms of OSA due to the complexity of upper airway collapse or narrowing during sleep.¹²

With respect to the individual surgical procedures, these recommendations were made:

• Tracheostomy is the only surgery that has been demonstrated to be consistently effective as the sole procedure for treating OSA; however, due to the cosmetic effects and morbidity associated with this surgery, it is rarely performed as a treatment for OSA, but is limited to situations in which other treatments are not acceptable, or for medical emergencies.¹²
• For all surgeries, the patient must be informed about surgical response rates and possible complications, as well as the availability and efficacy of treatment alternatives such as nCPAP and oral appliances.¹²
• UPPP, with or without tonsillectomy, may be appropriate for patients who demonstrate narrowing or collapse in the retropalatal region; however, surgical response rates are not necessarily predicted by a thorough preoperative evaluation, and the response rates are variable. UPPP should only be performed when less invasive treatments, such as nCPAP, have been considered.¹²
• With regard to retrolingual surgeries, specifically GAHM, surgical success depends upon the surgeon's skills and few centers have experience with this surgery. While GAHM appears to be the most promising, inadequate data are available on the effectiveness of MLG and lingualplasty.¹²
• MMO may successfully treat patients who fail UPPP, nasal surgery, MLG, or GAHM; however, this is generally not considered to be first-line therapy.¹²
• When postoperative healing is complete, OSA patients should be followed to assess the presence of residual disease, including an objective measurement of the presence and severity of OSA and of sleep disruptions. It is recommended that follow-up be performed with the patient off of nCPAP. In addition, due to the paucity of data on the durability of response following surgery, patients must be monitored for late relapses which can occur after initially responding to surgery.¹²

Laser-Assisted Uvulopalatoplasty (LAUP)

In a position paper developed by the AASM's Standards of Practice Committee and published in 1994, the AASM stated that the use of LAUP to treat OSA or snoring could not be condoned since there were was a paucity of controlled trials on the procedure in the peer-reviewed medical literature. The committee concluded that, although LAUP may have some efficacy for OSA, its role had not been defined. In fact, it was stated that the elimination of snoring by LAUP could mask the underlying presence of or delay the diagnosis of OSA. Thus, all patients who are being considered for LAUP must undergo a thorough preoperative clinical examination, including PSG, and be followed with PSG until disease remission is confirmed. The perioperative use of narcotics poses risks of airway complications for patients who have had LAUP.⁵⁵

Nasal Continuous Positive Airway Pressure (nCPAP)

The ATS published a statement in 1994 that delineated the indications and standards for use of nCPAP in OSA.⁴⁹ The major points of this statement are summarized as follows. nCPAP is effective for the treatment of adult patients with clinically important OSA. Treatment with nCPAP is indicated when there is documented evidence of OSA by PSG and clinical impairment. It is not routinely indicated for the treatment of simple snoring without the presence of OSA. The procedure is a safe and effective form of therapy for which complications are rare, and it effectively eliminates OSA, oxyhemoglobin desaturation, and respiratory event-related arousals from sleep. It is associated with reductions in EDS and improved cardiopulmonary function with possible, but unconfirmed, reductions in mortality. Relative contraindications to nCPAP use include bullous lung disease and recurrent sinus or ear infections; however, there are no absolute contraindications. While patients generally report that they are compliant with nCPAP, actual use of the device varies. Higher levels of patient compliance are associated with resolution of EDS, fatigue, and the restoration of mental alertness. The level of nasal airway pressure requires titration to a level that restores adequate ventilation, arterial oxygenation, and maintenance of sleep. Routine outpatient monitoring is recommended to ensure patient compliance, and to assess treatment efficacy. An initial follow-up at 1 month is recommended with additional follow-up at 3 to 6 month intervals. Due to the limitations of subjective reports of compliance, objective assessment is recommended. Nasal CPAP is not curative, and when it is not used, airway obstruction recurs.⁴⁹