Diagnostic accuracy of screening questionnaires for obstructive sleep apnea in psychiatric patients

Obstructive sleep apnea (OSA) is prevalent among psychiatric patients and can exacerbate psychiatric conditions. Screening for OSA in this population is challenging due to overlapping symptoms and adherence issues. This study evaluated the diagnostic accuracy of the STOP-Bang, BOAH, GOAL, NoSAS, and No-Apnea screening questionnaires in psychiatric patients.

In this observational cohort study, consecutive psychiatric patients admitted to a single tertiary care center between June 1, 2016 and December 31, 2022 were screened using the STOP-Bang questionnaire and additional data on parameters of the other questionnaires were collected. Polygraphy was performed using the Somnocheck micro CARDIO® system to record the Apnea-Hypopnea Index (AHI). Diagnostic accuracy was evaluated using receiver operating characteristic (ROC) curve, sensitivity, and specificity analysis.

Among 128 patients (median age 61.5 [IQR, 52.3-73.0] years; 77 [60.2 %] men), the No-Apnea questionnaire achieved the highest sensitivity for any OSA (AHI ≤5) at 96.00 % (95 % CI, 89.35 %-98.58 %; AUC, 0.68), followed by GOAL at 92.00 % (95 % CI, 83.78 %-96.27 %; AUC, 0.59) and STOP-Bang at 90.67 % (95 % CI, 81.85 %-95.26 %; AUC, 0.66). The specificities of all questionnaires were low, ranging from 15.09 % (STOP-Bang; 95 % CI, 8.35 %-25.90 %) to 50.94 % (BOAH; 95 % CI, 38.65 %-63.13 %).

The STOP-Bang, GOAL, and No-Apnea questionnaires effectively identify psychiatric patients at risk for OSA, though their low specificity underscores the necessity for confirmatory diagnostics, such as polygraphy. The No-Apnea score, which relies only on objective variables, may offer particular benefits in psychiatric settings.