Using the Karolinska Sleep Questionnaire to identify obstructive sleep apnea syndrome in a sleep clinic population.

INTRODUCTION: In Scandinavia, portable monitoring has virtually replaced standard polysomnography for diagnosis of obstructive sleep apnea syndrome (OSAS). Because waiting times for specialized OSAS care remain long, an accurate screening tool to exclude low-risk patients from diagnostic testing would be valuable.
OBJECTIVES: To examine the diagnostic accuracy of the Karolinska Sleep Questionnaire (KSQ) for OSAS.
METHODS: Consecutive patients, 30–66 years old, attending a large sleep clinic in Sweden for OSAS evaluation completed the KSQ and underwent in-home portable monitoring and medical history evaluation. OSAS was defined as apnea-hypopnea index ≥5 with symptoms of disease. We calculated sensitivity and specificity of apnea/snoring and sleepiness indices of the KSQ. Retrospectively, we combined six KSQ items (snoring, breathing cessations, disturbed sleep, etc.) and four clinical variables (age, sex, body mass index, smoking status) predictive of OSAS into a new instrument, which we also evaluated. Instrument score ranged between 0 and 21; a higher score indicated more severe symptoms.
RESULTS: Of 103 patients, 62 were diagnosed with OSAS. Sensitivity and specificity of the indices were 0.56 and 0.68 (apnea/snoring), and 0.37 and 0.71 (sleepiness). The new instrument performed optimally at a score of 9. Sensitivity was 0.76 (95% confidence interval 0.63–0.86) and specificity 0.88 (0.74–0.96). Between 19.4% and 50.5% of patients were unaware of having apnea/snoring symptoms.
CONCLUSIONS: Diagnostic accuracy of the apnea/snoring and sleepiness indices for OSAS was poor but could be improved by combining clinical and KSQ items. The usefulness of the apnea/snoring index and the combined instrument was questionable because of extensive symptom unawareness.