Background Assessment of chronic hypercapnia is recommended in certain patients with obstructive sleep apnoea (OSA). However, in many cases, although the overnight oximetry is suggestive for hypoventilation, patients do not have significant daytime hypercapnia. The aim of this quality improvement project was to identify the prevalence of chronic hypercapnia in patients with OSA and to identify factors in the oximetry signal which could predict pCO2 levels. Ultimately, the aim is to reduce the referral to treatment time (RTT) by rationalising blood gas requests.

Methods We analysed clinical data, sleep study results and earlobe blood gas values in 105 patients with OSA (55±15 yeas, 66 males, BMI 41±8 kg/m2, AHI 56±31/h) attending the early morning blood gas clinic following their sleep study suggestive for underlying sleep-related hypoventilation.

Results Thirty-seven percent of the patients have already had a diagnostic study at the referral, hence the average waiting time for a diagnostic sleep test was 102±104 days. From referral, the average time for the blood gas clinic was 209±136 days, and the average RTT was 260±113 days suggesting that waiting for the blood gas test had a significant impact on the RTT. Thirty-five patients (33%) had pCO2>6kPa, but only 5 had pCO2 ≥7kPa (5%). Time spent with oxygen saturation below 90% (T90) was significantly related to pCO2 (p<0.05, r=0.43), but there was no relationship with AHI, ODI or the use of sedative medications. The results were unchanged if patients with COPD (n=26) were analysed separately. T90>40% was sensitive to detect both pCO2 >6kPa (89%) and pCO2>7kPa (100%).

Discussion Clinically significant chronic hypercapnia is relatively uncommon in patients with OSA despite the oximetry data are suggestive for this. We suggest performing blood gases if the sleep test shows T90%>40% in patients with obstructive sleep apnoea and a clinical suspicion for hypoventilation syndrome.