Introduction Clinical commissioning standards have been developed to streamline clinical pathways. It is now common practice for obese patients with suspected sleep disordered breathing to undergo nocturnal oximetry monitoring prior to the clinic consultation. Although this test is useful for diagnosis and risk stratification of patients, there are limited data reporting the use of oximetry to predict hypercapnia. We hypothesised that overnight oxygen saturations could be used to predict hypercapnia.

Method 186 oximetry studies from patients with a body mass index (BMI) > 30 kg.m^-2 and an FEV₁/FVC >0.7 were analysed, including the percentage of total analysis time spent with an oxygen saturation (S_pO₂) below 90% (T < 90%), 80% (T < 80%) and 70% (T < 70%) as well as 4% and 3% oxygen desaturation index (ODI). Correlations and linear regression analyses were performed to determine the variables that predicted a daytime arterial partial pressure of carbon dioxide (P_aCO₂) > 6.0 kPa. Binary logistic regression and receiver-operator characteristic analyses assessed the utility of these parameters in predicting hypercapnia.

Results Compared to the eucapnic group the hypercapnic patients had a higher 4% ODI (42.6 ± 35.5 events/hour vs. 24.5 ± 19.5 events/hour, p = 0.003), lower mean SpO₂ (89.0 ± 7.4% vs. 94.1 ± 3.2% p = ns) and higher T < 90% (36.3 ± 32.1% vs. 13.5 ± 20.4%, p < 0.001).

Significant, albeit weak, correlations between PaCO₂ and 4% ODI, 3% ODI, T < 90%, T < 80%, T < 70% were observed (Table 1). Only T < 90% was predictive of hypercapnia. Using the total analysis time with an S_pO₂ < 90%, a cut off level of ≥7.2% had a sensitivity of 80% and a specificity of 60% in predicting a PaCO₂ >6 kPa, area under the curve was 0.76.

Conclusion The proportion of time spent with an S_pO₂<90% predicted hypercapnia in obese patients. This has the potential to risk stratify patients, optimising both the timing and type of treatment delivered, which in turn will enhance the delivery of care. Specifically, this would facilitate clinical decision making in directing patients towards investigation for receiving non-invasive ventilation rather than continuous positive airway pressure therapy if hypercapnia were predicted from the proportion of the time with an S_pO₂ < 90%.