Introduction CPAP compliance is a challenge in the management of obstructive sleep apnoea (OSA). Pre-operative screening with a sleep questionnaire for OSA followed by sleep studies is common in bariatric services. The King’s bariatric service performs overnight pulse oximetry in all patients considered for surgery. Therefore, additional cases may be detected in patients with lower clinical suspicion. We hypothesised that CPAP compliance in bariatric patients with OSA confirmed by routine screening would be lower than that seen in patients referred to the sleep clinic for suspected OSA.

Method Case records of all bariatric patients screened for OSA over a 16-month period were reviewed, and CPAP compliance data retrieved for those with confirmed OSA who commenced CPAP therapy. Retrospective case-control analysis was made against 50 randomly selected patients from the sleep clinic that commenced CPAP for confirmed OSA within the same study period. CPAP compliance at 4 weeks was compared between the two groups.

Results 409 patients were screened through the bariatric pathway during the study period. 49 (12.0%) patients were diagnosed with OSA and were commenced on CPAP. Baseline characteristics of the bariatric and sleep clinic groups are shown in Table 1. There was no significant difference between gender, baseline overnight desaturation indices and baseline Epworth Sleepiness Scale scores between the two groups.

21 (42.9%) patients in the bariatric group used their CPAP for ≥75% of nights within the 28-day period compared with 25 (50.0%) patients in the sleep clinic group; p = 0.48. There was a trend to significance for proportion of patients who used CPAP for mean ≥4 hours per night (20 (40.9%)) patients in the bariatric group vs 29 (58.0%) patients in the sleep clinic group; p = 0.087).

Conclusion Our study suggests that the number of patients who are compliant with CPAP at mean use of ≥4 hours per night may be lower in those with confirmed OSA found on routine screening within a bariatric pathway than in patients referred to a sleep clinic, but our finding was limited by the small sample size. Future study to investigate this trend and its underlying causes could improve the success of intervention in the bariatric patients.