4. Introduction

Deciding which of the various treatment options is most appropriate for the management of OSAHS depends on both the severity of the condition and the characteristics of an individual patient. The recommended treatment for moderate or severe OSHAS, continuous positive airway pressure (CPAP) is not always easy for patients to accept and the eventual decision to use it rests firmly with the patient.

Treatment options can be broadly divided into:

• behavioural interventions
• non-surgical options
• surgical options.

4.1.1 WHOM TO TREAT?

Current evidence from randomised controlled trials (RCTs) indicates that improvements with treatment can be found in symptomatic patients with AHI ³15 or a 4% oxygen saturation dip rate at the level of >10/hour.^{73, 74, 75} There is some evidence of benefit from the treatment of symptomatic individuals with AHI of 5-14, however further studies are required to confirm this (see section 4.3.1).^{33, 34} These benefits are in daytime sleepiness, simulated driving performance, quality of life, blood pressure and mood. Treatment should thus be focussed on these benefits and not given in the hope of diminishing vascular risk in patients without daytime sleepiness or in asymptomatic individuals, for whom RCTs show no benefit.⁷⁶

4.1.2 COMMUNICATION WITH PATIENTS

Following diagnosis, treatment options with their respective advantages and disadvantages should be discussed with the patient and their partner, fully involving them in the decision making process. This should be done using a multidisciplinary approach with written and video material where available. As patients have to commit to lifelong treatment they and their partners need to fully understand the implications of treatment. This is likely to aid compliance and treatment outcomes.⁷⁷

4.2 Behavioural interventions

Overweight patients should be advised to lose weight as weight reduction improves OSAHS symptoms and other excess weight related disorders. Unfortunately sustained weight reduction is rarely achieved whatever resources are applied.^{78, 79, 80} In a limited proportion of cases, weight reduction of 10-15% has been associated with improvement in desaturation index and other markers of OSAHS. There is a poor correlation between the amount of weight lost and clinical response.^{79, 80} The role of gastro-intestinal surgery in the morbidly obese patient is under debate but may be considered in appropriate patients.⁸⁰ OSAHS can recur even after surgically induced weight loss.⁸¹ Evidence level 1++,2+,3

Patients who smoke should be advised to stop for general health reasons. While there is epidemiological evidence linking smoking to OSAHS,⁴ there is no evidence that stopping smoking improves apnoeic symptoms, indeed any weight gain after quitting smoking may worsen OSAHS.

Alcohol should not be taken in the evenings and sedatives and sleeping tablets avoided as all of these decrease airway dilator function and worsen OSAHS.

Weight loss should be encouraged in all patients with obesity contributing to their OSAHS. Attempts at weight loss should not delay the initiation of further treatment. Weight loss should also be encouraged as an adjunct to CPAP or intra-oral devices as it may allow discontinuation of therapy.

Patients who smoke should be advised to stop Alcohol and sedatives or sleeping tablets should be avoided Non-sleepy snorers should be discouraged from sleeping on their backs.

These measures may suffice in simple snorers or in those with very mild OSAHS and few symptoms but most patients with OSAHS need additional treatment.

A Cochrane review of lifestyle modifications for OSAHS identified no RCT evidence supporting their use and concluded that any decision to institute such interventions should not delay the institution of therapies of proven effectiveness, such as CPAP.⁸² Evidence level 2+

4.3 Non-surgical interventions

4.3.1 CONTINUOUS POSITIVE AIRWAY PRESSURE

Continuous positive airway pressure (CPAP) functions as a pneumatic splint to maintain upper airway patency throughout all phases of sleep breathing. It operates by means of a flow generator which delivers pressure through air tubing to a nasal or face mask worn overnight. Most patients require lifelong treatment and therefore long term access to a CPAP machine.

Efficacy

CPAP has been established as the treatment of OSAHS with the firmest evidence base.^{73, 83} Randomised controlled trials show that CPAP improves subjective and objective sleepiness^{84, 85, 86, 87} cognitive function, vigilance, mood^{73, 88} and quality of life measures.^{74, 84, 89} Objective improvements are found in symptomatic patients with AHI >15 or > ten 4% desaturations/hr.⁷⁴ Evidence level 1++,1+

Meta-analysis⁷³ of several RCTs showed evidence of differential outcome with OSAHS severity, with significant improvements in subjective Epworth sleepiness score in two studies of patients with severe OSAHS^{89, 90} (mean –6 points, 95% CI –8 to –3, Effect Size 1.5 SDs) and non-significant changes in four studies of patients with mild OSAHS (-1, CI -1 to +0.3).^{34, 85, 86, 87} Objective mean sleep latency, from the multiple sleep latency test (see section 3.3), showed similar significant improvements in sleep latency in a single trial of patients with severe OSAHS90 (+2 mins, CI 1 to 3, ES 0.5 SDs) but not amongst trials of patients with mild OSAHS (0 mins, CI -1 to +1).^{85, 87} A sham-controlled RCT of CPAP in asymptomatic individuals with raised AHI (but not diagnosed OSAHS) found no benefit for sleepiness from active CPAP.⁷⁶ In an RCT on patients with moderate to severe OSAHS using maintenance of wakefulness as the outcome measure, the change following treatment was from 23.4 to 30.4 mins (Effect Size 0.7 SDs).⁷⁴

Randomised controlled trials of treatment for patients with polysomnographically mild OSAHS do not support reduced objective sleepiness after CPAP.^{73, 76} Within the population sampled in these studies are conflicting individual findings. Sampling biases vary across these RCTs with excessive sleepiness constituting an inclusion criterion in one study⁸⁶ and an exclusion criterion in another.⁸⁷ A single study of patients with mild OSAHS but considerable daytime sleepiness did find significant improvements in Epworth score with CPAP, in contrast to the others in less sleepy samples.^{73, 89}

Autotitrating CPAP machines adjust the CPAP pressure delivered overnight to prevent apnoeic episodes.^{91, 92} These devices can be used to establish the pressure required for later home use. They can be useful in a troubleshooting role to investigate problems with CPAP use, or when there may be pressure requirement changes following, for example, weight change. Theoretically, these machines should be more comfortable for long term use as the pressure will match changes in required pressure and in general be lower. This may produce less mask leakage and flow through the nose leading to fewer nasal side effects.^{93, 94} They are more expensive than conventional CPAP machines and, as yet, there is no convincing evidence that they produce better outcomes than conventional fixed pressure machines.

CPAP is the first choice therapy for patients with moderate or severe OSAHS that is sufficiently symptomatic to require intervention.

Side effects and compliance

Major side effects of CPAP use (eg significant epistaxis, paranasal sinusitis) are rare, but minor side effects (rhinitis, nasal bridge sores, discomfort, claustrophobia, abdominal bloating, noise) are common. Intensive efforts can achieve CPAP uptake of up to 95%³⁷ and an average nightly use of three to five hours.^{95, 96} Nasal symptoms are usually due to mouth leaks causing high flows of cool air through the nose. Attempts should be made to reduce these using chin straps or full face masks. In a few patients nasal corticosteroids can be useful. A heated humidifier may help to improve comfort and compliance.⁹³

The requirements of an adequate CPAP service have not been clearly established, and a variety of locally tailored systems may be satisfactory.⁹⁷ An essential element appears to be a dedicated CPAP nurse or technician responsible for mask fit, CPAP initiation and patient support. Few studies of CPAP have specified type or sequence of mask fitting, pressure set-up (titration) or pattern of humidification use, although all these elements are considered relevant to effectiveness and tolerability.⁹³ Published series are largely based on traditional single night CPAP titration studies. There is some evidence that split night studies with a diagnostic study for the first part of the night and a CPAP titration study for the remainder can be effective in many patients.⁹⁸

The early pattern of use during the first few days and weeks (hours used per night) predicts long term use.^{99, 100} A threshold of less than two hours of CPAP use per night appears to distinguish those who are prepared to comply with therapy in the long term from those who withdraw. The likelihood of stopping treatment is far greater in those who at three months use CPAP for less than two hours per night.³⁷ There is some correlation between severity of symptoms and CPAP use, with those patients having more commanding symptoms more likely to tolerate the treatment Also, the lowest effective CPAP pressure contributes to better compliance.⁹⁶ Higher CPAP pressures are associated with a greater reduction in daytime sleepiness.³⁷ Evidence level 2+,2++

Persistent low CPAP use (less than two hours per night) over six months, following efforts to improve patient comfort, should lead to a review of treatment.

CPAP therapy should not be abandoned without: the attention of a trained CPAP nurse / technician a titration study / use of autotitrating CPAP to troubleshoot problems the use of heated humidification.

4.3.2 ECONOMIC ANALYSIS OF CPAP USAGE IN OSAHS

Patients with undiagnosed OSAHS are heavy users of the healthcare system. Expenditure on undiagnosed patients is approximately twice that of age and gender matched controls. This difference extends back over 10 years prior to the diagnosis of OSAHS being made.¹⁰¹

Treatment with CPAP reduces these costs with evidence of decreased hospitalisation due to cardiovascular and pulmonary disease.¹⁰² Hospitalisation and other costs associated with road traffic accidents are also reduced in those using CPAP therapy.^{103, 104, 105} Overall mean hospitalisation days per year decreased with CPAP use.¹⁰⁶

There is a need for more studies on the cost effectiveness of CPAP and other treatments for OSAHS. One study carried out in 1994 has estimated that CPAP resulted in an average gain of 5.4 Quality Adjusted Life Years (QALYs) at a cost of Canadian $3400 to $9800 / QALY gained, equivalent to £1500-£4400.¹⁰⁷ A British study suggests a cost of £3200 / QALY gained over a five year period.¹⁰⁸ These do not include any benefits that would be expected to accrue from a decrease in BP or road traffic accident decrease with CPAP treatment. These studies indicate a cost-effectiveness ratio in line with other routinely funded procedures within the NHS, but more studies are needed to confirm this. (See Annex 2)

4.3.3 BI-LEVEL POSITIVE AIRWAY PRESSURE

These machines allow independent adjustment of inspiratory and expiratory pressures rather than having a fixed pressure as with CPAP. A well-designed randomised study found no advantage for bi-level support over CPAP in straightforward OSAHS.¹⁰⁹ This mode of non-invasive ventilation may be more appropriate for patients with ventilatory failure.¹¹⁰ Evidence level 1+,4

Bi-level ventilation should not be used routinely in OSAHS but should be reserved for patients with ventilatory failure.

4.3.4 INTRA-ORAL DEVICES

Intra-oral devices (IODs) are a range of appliances designed to alter upper airway patency. Several techniques have been employed, but mandibular advancement has gained most acceptance. Most mechanisms have a similar action in producing anterior displacement of the mandible, thus increasing upper airway diameter in some appliances. The degree of mandibular protrusion can be specified and subsequently adjusted.

Effectiveness

Recent randomised controlled trials have evaluated the effectiveness of intra-oral devices (IODs) against no treatment^{111, 112} or placebo IOD^{113, 114, 115, 116, 117} with all trials incorporating crossover designs. IODs have also been assessed against CPAP^{83, 118, 119, 120, 121, 122} and against uvulopalatopharyngoplasty.¹²³

The patients recruited into these trials comprised snorers without desaturation^{114, 124} patients with mild to moderate, symptomatic OSAHS^{112, 113, 115, 116} or mixed snorers and apnoeic patients.¹¹¹ These trials have examined outcomes including symptom ratings,^{113, 114, 116} daytime sleepiness^{113, 114, 116} and polysomnography^{111, 113, 115, 116} but many trials have excluded participants who were intolerant of IODs.^{111, 113, 115}

The effect sizes of changes in apnoea/hypopnoea index with IOD compared to no treatment or a placebo device in patients with mild to moderate OSAHS have varied from small (0.3 SD)¹¹² to very large (1.6 SD),¹¹⁵ as have differences in snoring frequency (0.3 SD¹¹¹ to 1.1 SD¹¹⁵). In these studies mean AHI following therapy using real IODs ranged from 12¹¹⁶ to 23¹¹³ per hour, suggesting that IODs failed to normalise sleep disordered breathing variables in these subjects. Evidence level 1-,1+

Symptom and sleepiness ratings from this group of trials also favour IOD over no IOD or a placebo IOD. In snorers lacking significant desaturation a real IOD produced large effect sizes (>1 SD) in partners’ ratings of loudness and frequency of snoring over a placebo device but smaller benefits for patient ratings of Epworth sleepiness (0.2 SD) and waking unrefreshed (0.5 SD).¹¹⁴ In patients with mild to moderate OSAHS, the same IOD produced better mean scores on these ratings than the placebo IOD, but these remained non-significant (Epworth difference of 1 point (0.2 SD), snoring loudness 0.7 SD).¹¹³ In another placebo-controlled trial of IOD in OSAHS patients, sleep latency on the MSLT was significantly improved by one minute on real IOD (effect size 0.3 SD) and Epworth score significantly decreased by two points (0.4 SD).¹¹⁶ Evidence level 1+

Randomised controlled trials comparing IODs to CPAP treatment are all crossover studies conducted in patients with symptomatic OSAHS and have assessed outcomes of polysomnography,^{118, 119, 120, 121} sleepiness^{119, 120, 121} and health status and cognitive function.^{120, 121} Most have experienced some subject attrition due to refusal to crossover and some have analysed only patients tolerant of one or both treatments.^{117, 118, 120}

For polysomnographic outcomes, the mean values for AHI, oxygen saturation and sleep fragmentation favour CPAP in all studies. Not all studies have reported statistical tests of IOD versus CPAP.^{118, 119} Of those that have, two demonstrated a significantly greater reduction in AHI whilst using CPAP than with IODs.^{120, 121} The intertreatment effect sizes for AHI range from moderate (0.4 SDs)^{120, 121} to large (0.9 SDs).¹¹⁸ Mean AHI following CPAP therapy ranged from 3 - 8 per hour, and following IOD therapy ranged from 8 - 15 per hour.^{118, 119, 120, 121} The highest AHI values for both treatments were from an intention-to-treat study in which treatment was used as long as tolerable during polysomnography.¹²⁰ Evidence level 1++,1+

Baseline Epworth scores before treatment averaged 11, 14 and 13 in the three studies evaluating this.^{119, 120, 121} Post-treatment Epworth sleepiness ratings on IOD and CPAP were not statistically different in two studies^{119, 121} and significantly better by 5 points (0.8 SDs) with CPAP in the third which had most severe baseline subjective sleepiness.¹²⁰ Objective sleep latency from a maintenance of wakefulness test and four objective cognitive performance scores from this third study were not different between treatments, but all three SF-36 health status summary scores (effect sizes 0.2 - 0.5 SDs) were significantly better on CPAP. Satisfaction with treatment was significantly higher for IOD in two studies^{118, 119} and higher for CPAP in another.¹²⁰ For long term treatment, IOD was preferred by 17 of 25 patients,¹¹⁸ 13 of 20,¹¹⁹ 19 of 48¹²⁰ and 17 of 21.¹²¹ Thus 66 of 114 patients in these trials preferred IOD treatment, despite worse sleep-disordered breathing on this treatment. Evidence level 1++,1+

A meta-analysis of patients’ treatment preference (CPAP and IODs) in three crossover studies in mild to moderate OSAHS showed a significant patient preference for IODs (OR 9.5, 95% CI 4 to 21), despite lesser nocturnal efficacy for breathing pauses (-7 per hr, 95% CI -10 to –5).⁸³ This was not confirmed in a later study.⁵⁵ Patients’ preference for IODs is important, but it is not known if this means that they feel symptomatically better when using IODs or whether they find the concept of an unobtrusive intra-oral device preferable to using an obtrusive CPAP device. Evidence level 1+

One group has compared the effectiveness of IOD against UPPP in a parallel-group, longitudinal follow-up study, with the latest report at four years post-randomisation.¹²³ In this, 72 of 95 patients with mild to moderate OSAHS have returned for polysomnography, which showed large effect sizes (>1.0 SDs) significantly favouring IOD over UPPP for improvements in AHI and desaturation index, but no significant difference in snoring duration between treatments. Evidence level 1+

It is unclear whether CPAP or IODs have the greater cost effectiveness and the answer may vary with OSAHS severity. The basic cost of many IODs is less than the cost of a CPAP machine. Some adjustable IODs are more expensive than CPAP, especially when the cost of multiple dental visits to adjust the IOD are included. (See Annex 2) The role of IODs as a first line management strategy for subjects with mild OSAHS has been demonstrated. However, their effectiveness in managing more severe OSAHS is limited by the present lack of identifiable prognostic indicators for their success.

Intra-oral devices are an appropriate therapy for snorers and for patients with mild OSAHS with normal daytime alertness.

Intra-oral devices are an appropriate alternative therapy for patients who are unable to tolerate CPAP.

Side effects and compliance

Hypersalivation and teeth/gum discomfort are common early side effects but usually decline if patients are able to persevere with IOD use.¹²⁵ Recurrent dislodgement of the device during sleep and temporomandibular pain are longer term problems. Side effects abate with appliance withdrawal. Early perseveration of use inside the first three months predicts long term success. These devices often require adjustment and may need to be replaced.¹²⁶ Evidence level 3

The reasons for IOD failure, even when worn correctly, are complex and mainly relate to different combinations of skeletal, soft tissue and functional factors that produce upper airway obstruction in each patient. Failure to control the OSAHS will lead to poor symptom control and therefore follow up of these patients is important, particularly as regards a return to driving.^{119, 127} Evidence level 1++,3

The use of intra oral devices should be monitored following initiation of therapy to allow device adjustment and assessment of OSAHS control and symptoms.

4.4 Pharmacological treatments

The evidence base to support pharmacological treatment as an effective therapeutic option is small. The main systematic review of pharmacotherapy concluded that no medication demonstrated a consistent response.^{128, 129} Evidence level 1++,1+

Drugs which suppress rapid eye movement (eg protryptyline, acetozolamide and progesterone) do not show clinical benefit in treating OSAHS in controlled trials. Trials of theophylline show an inconsistent effect with a tendency to disrupt sleep in patients with OSAHS, but improve Cheyne- Stokes respiration. Hypnotics, such as benzodiazepines, may worsen OSAHS.¹²⁸ Evidence level 1++,1+

There is some evidence to suggest that the addition of alerting drugs, such as modafanil, may have a small beneficial effect on sleepiness in some patients who remain sleepy despite good CPAP compliance. However, they may decrease CPAP use and longer term studies of their value and risks are needed.¹³⁰ There is no evidence to suggest that they could be used as an alternative to CPAP and they are not a substitute for careful attention to improving CPAP comfort and efficacy. Evidence level 1++,1+

Pharmacological therapy should not be used as first line therapy for OSAHS.