The discriminative power of STOP-Bang as a screening tool for suspected obstructive sleep apnea in clinically referred patients: considering gender differences

• Published in: Sleep & breathing Vol. 23; no. 1; pp. 65 - 75
• Main Authors: Mou, Jin, Pflugeisen, Bethann M., Crick, Brian A., Amoroso, Paul J., Harmon, Kirk T., Tarnoczy, Stephen F., Shirley Ho, S., Mebust, Kimberly A.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01-03-2019; Springer Nature B.V
• Subjects: Adult; Aged; Apnea; Body Mass Index; Dentistry; Female; Gender differences; Humans; Internal Medicine; Male; Mass Screening - standards; Medicine; Medicine & Public Health; Middle Aged; Neck; Neurology; Otorhinolaryngology; Pediatrics; Pneumology/Respiratory System; Polysomnography; Severity of Illness Index; Sex differences; Sex Factors; Sleep; Sleep apnea; Sleep Apnea, Obstructive - diagnosis; Sleep Breathing Physiology and Disorders • Original Article; Sleep disorders; Surveys and Questionnaires - standards; Sleep disorder; Screening; Obstructive sleep apnea; Gender disparity; STOP-Bang
• ISSN: 1520-9512; 1522-1709
• DOI: 10.1007/s11325-018-1658-y

Purpose
Obstructive sleep apnea (OSA) is the most commonly seen clinical sleep disorder. STOP-Bang, a widely used screening tool, yields a composite score based on eight dichotomized items including male gender. This study was designed to validate STOP-Bang among clinically referred patients and tested alternative scoring designs on tool performance, with a focus on gender differences in OSA.
Method
STOP-Bang was administered to 403 female and 532 male subjects, followed by comprehensive sleep evaluation that included measurement of apnea-hypopnea indexes. Gender differences in STOP-Bang scores, OSA diagnosis, and severities were explored, and gender-specific alternative score cutoffs evaluated. Optimal operating points (OOP) were tested for female body mass index (BMI) and male neck circumference to inform STOP-Bang threshold refinement. Receiver operating characteristic curves were used to compare conventional and modified STOP-Bang.
Results
STOP-Bang performance by gender showed extremely low specificity in males at the recommended cutoff of ≥3. Better utility was presented at a cutoff of 4 or 5 among clinically referred patients irrespective of gender differences. Screening performance was improved by modifying BMI and/or neck circumference thresholds using gender-triaged OOP estimation. Three gender-based model revisions outperformed conventional STOP-Bang.
Conclusion
Our study suggests that gender-specific consideration needs to be incorporated into the application of STOP-Bang in a clinically referred patient population with a higher risk of OSA. Alternative scoring systems may improve predictive performance of STOP-Bang.