Published in:
01-03-2017 | Original Contributions
Pre-Operative Ability of Clinical Scores to Predict Obstructive Sleep Apnea (OSA) Severity in Susceptible Surgical Patients
Published in: Obesity Surgery | Issue 3/2017
Login to get accessAbstract
Background
Severe obstructive sleep apnea (OSA) is an independent risk factor for perioperative complications. Clinical scores such as Snoring, Tiredness, Observed apnea, high blood Pressure, Body Mass Index (BMI) higher than 35 kg m−2, Age older than 50 years, Neck circumference larger than 40 cm, and male gender (STOP-Bang), perioperative sleep apnea prediction (P-SAP), and OSA50 have been proposed for detecting OSA. We recently proposed a new score based on morphological metrics only, the DES-OSA score. This study compared the DES-OSA score to the three other ones with regard to their ability to detect OSA. Obese patients are particularly at risk of OSA.
Methods
Following informed consent and institutional review board (IRB) approval, 1584 consecutive adults were. Should the STOP-Bang be indicative of increased risk of severe OSA, the patient was referred to complementary polysomnography (PSG). Eventual already existing recent PSG data were also collected. The abilities of the four scores to predict OSA severity were compared using sensitivity, specificity, Cohen’s kappa coefficient (CKC), and area under ROC curve (AUROC) analysis.
Results
PSG was performed in 150 patients. For detecting severe OSA, OSA50 had the highest sensitivity [value (95 % CI) 0.98 (0.90–1)]. STOP-Bang was significantly less sensitive than P-SAP and OSA50. In that respect, DES-OSA was significantly more specific than the three other ones [0.75 (0.65–0.83)]. The AUROC of DES-OSA was significantly the largest [0.9 (0.84–0.95)]. The highest CKC at detecting severe OSA was 0.62 (0.49–0.74) for DES-OSA. Similar results were obtained for moderate to severe OSA prediction.
Conclusions
DES-OSA, which is the only exclusively morphological score available, appears to surpass the three other scores in their ability to predict moderate to severe and severe OSA, at least in our setting and in our screened population.
Clinical Trial Registration
ClinicalTrial.gov NCT02051829