A Cochrane review [Abstract] 1 included 30 studies with a total of 2 047 subjects. Studies were categorised by intervention type: supportive interventions during follow-up, educational interventions and behavioural therapy. Across all three intervention classes, most studies incorporated elements of more than one intervention. Interventions were categorised by the prevailing type of intervention, which was expected would have the greatest impact on the study outcome. All three types of interventions led to increased machine usage in CPAP-naive participants with moderate to severe obstructive sleep apnoea (OSA) syndrome.
Compared with usual care, supportive ongoing interventions increased machine usage by about 50 minutes per night (0.82 hours, 95% CI 0.36 to 1.27; 13 studies, n = 803), increased the number of participants who used their machines for longer than 4 hours per night from 59 to 75 per 100 (OR 2.06, 95% CI 1.22 to 3.47; 4 studies, n=268) and reduced the likelihood of study withdrawal (OR 0.65, 95% CI 0.44 to 0.97; 12 studies, n = 903). With the exception of study withdrawal, considerable variation was evident between the results of individual studies across these outcomes. Evidence of an effect on symptoms and quality of life was statistically imprecise (ESS score -0.60 points, 95% CI -1.81 to 0.62; 8 studies, n=501; Functional Outcomes of Sleep Questionnaire 0.98 units, 95% CI -0.84 to 2.79; 2 studies, n=70).
Educational interventions increased machine usage by about 35 minutes per night (0.60 hours, 95% CI 0.27 to 0.93; 7 studies, n=508), increased the number of participants who used their machines for longer than 4 hours per night from 57 to 70 per 100 (OR 1.80, 95% CI 1.09 to 2.95; 3 studies, n=285) and reduced the likelihood of withdrawal from the study (OR 0.67, 95% CI 0.45 to 0.98, 8 studies, n=683). Participants experienced a small improvement in symptoms, the size of which may not be clinically significant (ESS score -1.17 points, 95% CI -2.07 to -0.26; 5 studies, n=336).
Behavioural therapy led to substantial improvement in average machine usage of 1.44 hours per night (95% CI 0.43 to 2.45; 6 studies, n=584) and increased the number of participants who used their machines for longer than 4 hours per night from 28 to 47 per 100 (OR 2.23, 95% CI 1.45 to 3.45; 3 studies, n=358) but with high levels of statistical heterogeneity. The estimated lower rate of withdrawal with behavioural interventions was imprecise and did not reach statistical significance (OR 0.85, 95% CI 0.57 to 1.25; 5 studies, n=609).
Comment: The quality of evidence is downgraded by study limitations (unclear allocation concealment and lack of blinding) and by inconsistency (variability in results).
Primary/Secondary Keywords