Pressure modification for improving usage of continuous positive airway pressure machines in adults with obstructive sleep apnoea



Obstructive sleep apnoea (OSA) is the repetitive closure of the upper airway during sleep. This results in disturbed sleep and excessive daytime sleepiness. It is a risk factor for long‐term cardiovascular morbidity. Continuous positive airway pressure (CPAP) machines can be applied during sleep. They deliver air pressure by a nasal or oronasal mask to prevent the airway from closing, reducing sleep disturbance and improving sleep quality. Some people find them difficult to tolerate because of high pressure levels and other symptoms such as a dry mouth. Switching to machines that vary the level of air pressure required to reduce sleep disturbance could increase comfort and promote more regular use. Humidification devices humidify the air that is delivered to the upper airway through the CPAP circuit. Humidification may reduce dryness of the throat and mouth and thus improve CPAP tolerability. This updated Cochrane Review looks at modifying the delivery of positive pressure and humidification on machine usage and other clinical outcomes in OSA.


To determine the effects of positive pressure modification or humidification on increasing CPAP machine usage in adults with OSA.

Search methods

We searched Cochrane Airways Specialised Register and clinical trials registries on 15 October 2018.

Selection criteria

Randomised parallel group or cross‐over trials in adults with OSA. We included studies that compared automatically adjusting CPAP (auto‐CPAP), bilevel positive airway pressure (bi‐PAP), CPAP with expiratory pressure relief (CPAPexp), heated humidification plus fixed CPAP, automatically adjusting CPAP with expiratory pressure relief, Bi‐PAP with expiratory pressure relief, auto bi‐PAP and CPAPexp with wakefulness detection with fixed pressure setting.

Data collection and analysis

We used standard methods expected by Cochrane. We assessed the certainty of evidence using GRADE for the outcomes of machine usage, symptoms (measured by the Epworth Sleepiness Scale (ESS)), Apnoea Hypopnoea Index (AHI), quality of life measured by Functional Outcomes of Sleep Questionnaire (FOSQ), blood pressure, withdrawals and adverse events (e.g. nasal blockage or mask intolerance). The main comparison of interest in the review is auto‐CPAP versus fixed CPAP.

Main results

We included 64 studies (3922 participants, 75% male). The main comparison of auto‐CPAP with fixed CPAP is based on 36 studies with 2135 participants from Europe, USA, Hong Kong and Australia. The majority of studies recruited participants who were recently diagnosed with OSA and had not used CPAP previously. They had excessive sleepiness (ESS: 13), severe sleep disturbance (AHI ranged from 22 to 59), and average body mass index (BMI) of 35 kg/m2. Interventions were delivered at home and the duration of most studies was 12 weeks or less. We judged that studies at high or unclear risk of bias likely influenced the effect of auto‐CPAP on machine usage, symptoms, quality of life and tolerability, but not for other outcomes.

Primary outcome

Compared with average usage of about five hours per night with fixed CPAP, people probably use auto‐CPAP for 13 minutes longer per night at about six weeks (mean difference (MD) 0.21 hours/night, 95% confidence interval (CI) 0.11 to 0.31; 31 studies, 1452 participants; moderate‐certainty evidence). We do not have enough data to determine whether auto‐CPAP increases the number of people who use machines for more than four hours per night compared with fixed CPAP (odds ratio (OR) 1.16, 95% CI 0.75 to 1.81; 2 studies, 346 participants; low‐certainty evidence).

Secondary outcomes

Auto‐CPAP probably reduces daytime sleepiness compared with fixed CPAP at about six weeks by a small amount (MD ‐0.44 ESS units, 95% CI ‐0.72 to ‐0.16; 25 studies, 1285 participants; moderate‐certainty evidence). AHI is slightly higher with auto‐CPAP than with fixed CPAP (MD 0.48 events per hour, 95% CI 0.16 to 0.80; 26 studies, 1256 participants; high‐certainty evidence), although it fell with both machine types from baseline values in the studies. Ten per cent of people in auto‐CPAP and 11% in the fixed CPAP arms withdrew from the studies (OR 0.90, 95% CI 0.64 to 1.27; moderate‐certainty evidence). Auto‐CPAP and fixed CPAP may have similar effects on quality of life, as measured by the FOSQ but more evidence is needed to be confident in this result (MD 0.12, 95% CI ‐0.21 to 0.46; 3 studies, 352 participants; low‐certainty evidence). Two studies (353 participants) provided data on clinic‐measured blood pressure. Auto‐CPAP may be slightly less effective at reducing diastolic blood pressure compared to fixed CPAP (MD 2.92 mmHg, 95% CI 1.06 to 4.77 mmHg; low‐certainty evidence). The two modalities of CPAP probably do not differ in their effects on systolic blood pressure (MD 1.87 mmHg, 95% CI ‐1.08 to 4.82; moderate‐certainty evidence). Nine studies (574 participants) provided information on adverse events such as nasal blockage, dry mouth, tolerance of treatment pressure and mask leak. They used different scales to capture these outcomes and due to variation in the direction and size of effect between the studies, the comparative effects on tolerability outcomes are uncertain (very low‐certainty evidence).

The evidence base for other interventions is smaller, and does not provide sufficient information to determine whether there are important differences between pressure modification strategies and fixed CPAP on machine usage outcomes, symptoms and quality of life. As with the evidence for the auto‐CPAP, adverse events are measured disparately.

Authors’ conclusions

In adults with moderate to severe sleep apnoea starting positive airway pressure therapy, auto‐CPAP probably increases machine usage by about 13 minutes per night. The effects on daytime sleepiness scores with auto‐CPAP are not clinically meaningful. AHI values are slightly lower with fixed CPAP. Use of validated quality of life instruments in the studies to date has been limited, although where they have been used the effect sizes have not exceeded proposed clinically important differences. The adoption of a standardised approach to measuring tolerability would help decision‐makers to balance benefits with harms from the different treatment options available. The evidence available for other pressure modification strategies does not provide a reliable basis on which to draw firm conclusions. Future studies should look at the effects of pressure modification devices and humidification in people who have already used CPAP but are unable to persist with treatment.

Plain language summary

How does changing pressure in continuous positive airway pressure machines increase their usage by adults with sleep apnoea?

What is the aim of this review?

This review looks at different ways of helping people who have obstructive sleep apnoea (OSA) to use continuous positive airway pressure (CPAP) machines. OSA refers to the temporary but frequent closing and opening of the throat during sleep. Because adults with OSA do not get the sleep that they need, they can feel tired in the day and this impacts on their quality of life. They are at risk of falling asleep while carrying out their daily activities and they are at risk of having heart disease or a stroke in the long term.

Key messages

Adults who use CPAP devices that automatically vary treatment pressure probably increase use of their machines by about 13 minutes per night but it is unlikely that they will feel less tired compared with people who use fixed pressure CPAP machines. Fixed pressure CPAP is slightly better at reducing the number of episodes where the airway closes at night. We need more information to be able to assess whether there are important differences between these machines on quality of life and tolerability.

What was studied in this review?

CPAP machines are worn over the nose and mouth as people sleep at night. They blow air at a fixed pressure through the nose and mouth to keep the airway open. This makes refreshing sleep easier to achieve, but some people find the machines difficult to use regularly. They can find the level of pressure too high or they have a dry mouth when they wake up. If the pressure needed to keep the airway open is lower, the machines could be easier to use more often.

We looked for studies that compared different ways of varying the way that pressure is delivered. We have focused on studies that compared CPAP machines that automatically vary treatment pressure as the person sleeps (e.g. auto‐CPAP) with machines that deliver pressure at the same level throughout the night (fixed pressure CPAP).

What are the main results of the review?

We found 64 studies in 3922 people. Thirty‐six studies in 2135 people were relevant to the comparison of auto‐CPAP and fixed pressure CPAP. The studies were from Europe, USA, Hong Kong and Australia. Seventy‐five per cent of people recruited to the studies were men who were recently diagnosed with sleep apnoea and had no experience of using CPAP.

Compared with fixed pressure CPAP, people starting treatment with auto‐CPAP will probably use their machine by about 13 minutes more per night at around six weeks (moderate‐certainty evidence), although when we looked at the number of people who use the machines for more than four hours per night, we did not have enough information to know whether there was a difference between the different machines (low‐certainty evidence).

Auto‐CPAP probably reduces daytime symptoms by a small amount compared with fixed pressure CPAP (moderate‐certainty evidence). A similar number of people withdrew from the studies: 11% in fixed pressure CPAP and 10% with auto‐CPAP (moderate‐certainty evidence). Both machines reduced the number of times that the upper airway closed during sleep, although fixed pressure CPAP was slightly better (high‐certainty evidence). Three studies used a scale that we chose as the most relevant one to measure quality of life in sleep apnoea (the Functional Outcomes of Sleep Questionnaire). In people using the machines, the average difference between the devices on this scale was small, but there was not enough evidence to be confident about this result (low‐certainty evidence). Auto‐CPAP may be less successful in controlling blood pressure than fixed pressure CPAP, but more studies are needed to confirm this result. We are uncertain as to how people found using their devices because information on tolerability (blocked nose, dry mouth, mask leak or feeling that the pressure level was too high) was measured differently across the studies (very low‐certainty).

This Plain Language Summary is current to October 2018

Authors’ conclusions

Implications for practice

In adults with moderate to severe sleep apnoea who are starting positive airway pressure therapy, average machine usage is probably increased by about 13 minutes per night with automatically adjusting continuous positive airway pressure (auto‐CPAP) compared with fixed CPAP (moderate‐certainty evidence), although we do not have enough evidence to determine whether either device leads to more people using machines for four hours or more (low certainty). The reduction in daytime sleepiness associated with the average increased use of auto‐CPAP is small (moderate‐certainty evidence), and fixed CPAP is slightly more effective at reducing the Apnoea Hypopnoea Index (AHI) (high‐certainty evidence). Most of the studies to date have followed participants for shorter than 12 weeks. It is unclear whether modifying pressure in participants with mild and less pronounced sleep apnoea symptoms is beneficial. About 25% of people recruited to the studies are female. Generally, research to date has not addressed what interventions are effective in people who find CPAP difficult to tolerate.

Use of specific quality of life instruments in the studies to date has been limited, although where they have been used the effect sizes have not exceeded clinically important differences. Evidence for blood pressure outcomes is insufficient to determine the anti‐hypertensive effects between the different machine types. Tolerabillty outcomes were measured too inconsistently to further our understanding of how varying treatment pressure with auto‐CPAP changes the experience of applying positive airway pressure.

The evidence base for the remaining interventions included in the review is smaller, with a similar focus on machine usage and symptoms over quality of life. The effect on machine usage with expiratory pressure CPAP (CPAPexp) is uncertain and there is little or no difference in symptoms when compared with fixed CPAP. Although there is some evidence of higher machine usage following the addition of humidification, the certainty of evidence is low, and the effects on symptoms and tolerability are uncertain. Effects of humidification may also be affected by local ambient humidity.

Implications for research

Study investigators and research funders initiating further research in this area should address the following issues related to the design, analysis and reporting of studies.

  1. To date, only a small proportion of eligible trials have recruited people who have been unable to persist with CPAP. Further trials in this population are needed to inform decisions between alternative management strategies where acceptability of CPAP is lower than in the populations recruited to studies in this review.
  2. Assessment of the effects of different pressure machines in studies that recruit a greater proportion of women and participants with lower AHIs than those in the existing studies would broaden the applicability of the evidence base.
  3. Adoption of a standardised approach to defining machine tolerability is needed. Variation in the definition and measurement of machine tolerability has made it challenging to evaluate the harms from using different pressure modes, and to determine their relationship with machine usage. Agreement on consistent approaches to measuring mask leak, pressure tolerance, and oral and nasal symptoms would help decision‐makers to balance benefits with harms from different treatment options available. Consensus on which blood pressure parameters to measure is needed in order to better understand the relationship between blood pressure and long‐term cardiovascular outcomes.
  4. We are aware of one large completed study (N = 800) comparing auto‐CPAP and fixed CPAP, which we expect to incorporate in future versions of this review when its results are available (NCT02749812). Two smaller ongoing studies are addressing similar outcomes to those already addressed (ACTRN12618000379213pNCT03428516). Future studies should include longer‐term follow‐up and measure quality of life.
  5. We have identified one ongoing study of CPAPexp (NCT01753999), and large multicentre studies in bilevel positive airway pressure (bi‐PAP), CPAPexp, humidification and automatically adjusting CPAP with expiratory pressure relief (auto‐CPAPexp) are warranted.
  6. Studies should better report design features (e.g. blinding and allocation concealment) and be more explicit about handling data from participants who do not continue to use machines during the studies.
  7. Given the range of treatment options that continue to emerge, head to head comparisons between different ways of modifying pressure would help to inform choice of treatment where alternatives to fixed CPAP are being considered. Such studies could also serve as a basis for a network meta‐analysis of pressure modification strategies, including a cost‐effectiveness assessment to inform local decision making.
Comments Off on Pressure modification for improving usage of continuous positive airway pressure machines in adults with obstructive sleep apnoea
  • The review abstracts published on this site are the property of John Wiley & Sons, Ltd., and of the Cochrane Review Groups that have produced the reviews.