Subjective Evaluation of Sleep Apnea Is Not Sufficient in Stroke Rehabilitation

Abstract

Objective: Sleep apnea (SA) is common in stroke patients and is associated with poor functional outcome. Therefore, we studied the clinical value of the Berlin Questionnaire (BQ), a subjective screening method to detect stroke patients with SA in a neurorehabilitation care unit, and compared it with respirographic sleep studies. Another aim was to explore the association of SA with functional status in the subacute phase after stroke. Methods: Sixty-eight stroke patients were subjected to respirographic sleep studies to determine their apnea-hypopnea index (AHI). Subjective evaluation was performed by means of the BQ. Functional outcome was assessed by the Barthel Index (BI). Sensitivity and specificity of the BQ for detecting SA in stroke patients were calculated. Results: Respirographic sleep studies showed a high prevalence of moderate and severe sleep apnea in 56% of stroke patients. Sensitivity and specificity of the BQ for SA (AHI > 15) were 0.69 and 0.15, respectively. The area under the receiver operating characteristics (ROC) curve for detection of SA was 0.58. The AHI correlated significantly with the BI (r = -0.57, P ≤ .001). Conclusion: As the BQ demonstrated low diagnostic utility in stroke rehabilitation, screening for SA should not be based solely on clinical interviews. SA is linked to poor functional status in the subacute phase post stroke.