STOP-BANG questionnaire performance in a Veterans Affairs unattended sleep study program.

Area of Science:

• Sleep Medicine
• Respiratory Medicine
• Clinical Diagnostics

Background:

• Limited data exists on the STOP-BANG questionnaire's utility in Veterans Affairs (VA) sleep clinic referrals.
• A validated STOP-BANG score cutpoint could potentially reduce the need for diagnostic sleep studies.

Purpose of the Study:

• To evaluate the STOP-BANG questionnaire's effectiveness in identifying sleep apnea in a veteran population.
• To assess the positive predictive value of STOP-BANG scores for significant sleep apnea (Respiratory Disturbance Index > 15/hour).
• To recalibrate the STOP-BANG score for this specific referral population.

Main Methods:

• Prospective administration of STOP-BANG questionnaires to veterans undergoing unattended sleep studies.
• Evaluation of sensitivity, specificity, and ROC AUC for STOP-BANG scores.
• Recalibration of the STOP-BANG score using logistic regression models.

Main Results:

• The mean STOP-BANG score was 5.7 ± 1.4 in 1,196 veterans; 67% had an RDI > 15/hour.
• Sensitivity decreased sharply at scores ≥ 6, while specificity increased linearly.
• Recalibrated models improved ROC AUC to 0.74; a maximal score of 8 had 85% positive predictive value.

Conclusions:

• The STOP-BANG questionnaire alone is insufficient for diagnosing significant sleep apnea.
• A maximal STOP-BANG score of 8 does not possess adequate positive predictive value to replace confirmatory sleep testing.