Background There is an increased risk of sleep apnea in patients using opioids for chronic pain. We hypothesized that a simple model comprizing of: (1) STOP-Bang questionnaire and resting daytime oxyhemoglobin saturation (SpO₂); and (2) overnight oximetry will identify those at risk of moderate-to-severe sleep apnea in patients with chronic pain.

Method Adults on opioids for chronic pain were recruited from pain clinics. Participants completed the STOP-Bang questionnaire, resting daytime SpO₂, and in-laboratory polysomnography. Overnight oximetry was performed at home to derive the Oxygen Desaturation Index. A STOP-Bang score ≥3 or resting daytime SpO₂ ≤95% were used as thresholds for the first step, and for those identified at risk, overnight oximetry was used for further screening. The Oxygen Desaturation Index from overnight oximetry was validated against the Apnea-Hypopnea Index (≥15 events/hour) from polysomnography.

Results Of 199 participants (52.5±12.8 years, 58% women), 159 (79.9%) had a STOP-Bang score ≥3 or resting SpO₂ ≤95% and entered the second step (overnight oximetry). Using an Oxygen Desaturation Index ≥5 events/hour, the model had a sensitivity of 86.4% and specificity of 52% for identifying moderate-to-severe sleep apnea. The number of participants who would require diagnostic sleep studies was decreased by 38% from Step 1 to Step 2 of the model.

Conclusion A simple model using STOP-Bang questionnaire and resting daytime SpO₂, followed by overnight oximetry, can identify those at high risk of moderate-to-severe sleep apnea in patients using opioids for chronic pain.

Trial registration number NCT02513836.