Modified STOP-BANG questionnaire to predict obesity hypoventilation syndrome in obese subjects with obstructive sleep apnea

Sleep Breath. 2016 May;20(2):495-500. doi: 10.1007/s11325-015-1213-z. Epub 2015 Jun 6.


Purpose: The STOP-BANG questionnaire (SBQ) has never been studied in the context of its ability to predict obesity hypoventilation syndrome (OHS). Our aim was to evaluate the predictive performance of the original and modified SBQs for OHS in obese subjects with obstructive sleep apnea (OSA).

Methods: Demographics, polysomnographic data, body mass index (BMI), Epworth Sleepiness Scale (ESS) scores, arterial blood gases, spirometric measurements, and SBQ scores were recorded. The modified SBQ was created by dividing BMI into ranges and adding the serum bicarbonate ranges.

Results: The study included 196 obese subjects, of whom 17 had normal polysomnography. Of the remaining subjects, 105 had pure OSA and 74 had OHS with OSA. Both the original and modified SBQs scores were higher for the OHS subjects than for those with pure OSA (p < 0.001). An original SBQ score of ≥6 gave a satisfactory discrimination for OHS diagnosis (sensitivity 71.6 %, specificity 59.1 %, positive predictive value (PPV) 55.2 %, and negative predictive value (NPV) 74.7 %). The diagnostic OR for an original SBQ score of ≥6 for predicting OHS was 3.7. The sensitivity and NPV were increased for the modified SBQ (sensitivity 89.2 %, specificity 47.6 %, PPV 54.6 %, NPV 86.2 %), and the OR was 7.5. Both the original and modified SBQ scores were moderately correlated with ESS, AHI, ODI, lowest SpO2, and sleep time spent with SpO2 <90 %.

Conclusions: The modified SBQ can be used to screen for OHS in obese subjects.

Keywords: Obesity; Obesity hypoventilation syndrome; STOP-BANG questionnaire.

MeSH terms

  • Diagnosis, Differential
  • Humans
  • Obesity / complications*
  • Obesity / diagnosis
  • Obesity Hypoventilation Syndrome / diagnosis*
  • Polysomnography
  • Predictive Value of Tests
  • Sensitivity and Specificity
  • Sleep Apnea, Obstructive / diagnosis*
  • Statistics as Topic
  • Surveys and Questionnaires*