P055 Identifying Sleepiness Profiles in Obstructive Sleep Apnoea

Abstract Aim Sleep clinics frequently measure the patient’s subjective sleepiness, but these measurements often conflict with more objective measures of sleepiness. This suggests that subjective and objective sleepiness have distinct properties and may have different causes in OSA. Using cluster analysis, we aimed to classify OSA patients into subgroups based on measures of objective and subjective sleepiness and assess each subgroup’s characteristics. Methods The study divided all patients diagnosed with OSA between 2015 and 2018 into four clinical presentation phenotypes: no sleepiness (ESS-/PVT-), subjectively sleepy only (ESS+/PVT-), objectively sleepy only (ESS-/PVT+), and subjectively and objectively sleepy (ESS+/PVT+). Results K means clustering of 1038 patients produced 4 groups of 463, 371, 126, and 78 patients for ESS-/PVT-, ESS+/PVT-, ESS-/PVT+, and ESS+/PVT+, respectively. Each group had no differences in AHI or BMI. The ESS-/PVT- cluster was associated with shorter sleep duration and poor sleep efficiency. The ESS+/PVT-cluster had a longer sleep duration and were younger than the rest of the groups. The ESS-/PVT+ cluster also had shorter sleep duration and poorer sleep efficiency, similar to ESS-/PVT-, but had a greater comorbidity than all the other clusters. The ESS+/PVT+ cluster exhibited a higher female prevalence and was linked to more severe desaturations compared to the other clusters. Discussion/Conclusion Sleep duration, comorbidity, sex, and hypoxic burden seem to be the biggest predictors of sleepiness in patients. Further work is required to explore the influence of OSA phenotypic characteristics on patient sleepiness.