Identifying phenotypes in OSA patients with an indication for CPAP treatment using clinical data and experienced symptom severity.

Background: Although the group of patients with obstructive sleep apnea (OSA) is very heterogeneous, OSA's severity is mainly expressed by an apnea-hypopnea index (AHI), which does not correlate well with the experienced symptom severity. As a first step to develop a more personalized approach for treatment, the purpose of the current study was to create, through cluster analysis, meaningful OSA phenotypes linked to the Patient Reported Apnea Questionnaire (PRAQ).

Methods: Through a survey, new OSA patients indicated for continuous positive airway pressure (CPAP) treatment completed the Epworth Sleepiness Scale (ESS) and the PRAQ to rate their experienced symptom severity. Clinical data, such as the AHI and comorbidity, were assessed from the patient file. Cluster analysis has been performed to derive OSA phenotypes.

Results: Based on the AHI, comorbidity and experienced symptom severity data of 151 patients, a two-step cluster analysis revealed five OSA phenotypes: "no comorbidity", "hypertension", "high symptom severity", "low symptom severity" and "unclassified". The five phenotypes mainly differ in the experienced level of fatigue, partner-observed snoring severity and symptoms related to performing regular daily activities.

Conclusion: Not only the AHI, but also comorbidity and subjective symptoms should be taken into consideration when diagnosing OSA, assessing its severity and in providing a more patient-oriented treatment, including deciding about CPAP treatment. Not the often-used ESS but the modified PRAQ scales provide relevant information to assess experienced symptom severity. In addition, for an improved prognostication, we propose an evaluation of the CPAP treatment effectiveness for the five reported OSA phenotypes.