Association of Obstructive Sleep Apnea With Difficult Intubation: Prospective Multicenter Observational Cohort Study.

BACKGROUND: Obstructive sleep apnea (OSA) has been found to be associated with difficult airway, although there is a paucity of prospective studies investigating thresholds of OSA severity with difficult airway outcomes. The aim of this study was to examine the association between OSA and difficult intubation or difficult mask ventilation. We also explored the utility of the Snoring, Tiredness, Observed apnea, high blood Pressure, Body mass index, Age, Neck circumference, and Gender (STOP-Bang) score for difficult airway prediction. METHODS: The Postoperative Vascular Complications in Unrecognized Obstructive Sleep Apnea (POSA) trial was an international prospective cohort study of surgical patients 45 years or older with one or more cardiac risk factor presenting for noncardiac surgery, with planned secondary analyses of difficult airway outcomes. Multivariable logistic regression analyses tested associations between OSA severity and predictors of difficult airway with difficult intubation or difficult mask ventilation. Overall, 869 patients without prior diagnosis of OSA were screened for OSA risk with the STOP-Bang tool, underwent preoperative sleep study, and had routine perioperative care, including general anesthesia with tracheal intubation. The primary outcome analyzed was difficult intubation, and the secondary outcome was difficult mask ventilation. RESULTS: Based on the sleep studies, 287 (33%), 324 (37%), 169 (20%), and 89 (10%) of the 869 patients had no, mild, moderate, and severe OSA, respectively. One hundred and seventy-two (20%) had a STOP-Bang score of 0-2 (low risk), 483 (55%) had a STOP-Bang score of 3-4 (intermediate risk), and 214 (25%) had a STOP-Bang score 5-8 (high risk). The incidence of difficult intubation was 6.7% (58 of 869), and difficult mask ventilation was 3.7% (32 of 869). Multivariable logistic regression demonstrated that moderate OSA (odds ratio [OR] = 3.26 [95% confidence interval {CI}, 1.37-8.38], adjusted P = .010) and severe OSA (OR = 4.05 [95% CI, 1.51-11.36], adjusted P = .006) but not mild OSA were independently associated with difficult intubation compared to patients without OSA. Relative to scores of 0-2, STOP-Bang scores of 3-4 and 5-8 were associated with increased odds of difficult intubation (OR = 3.01 [95% CI, 1.13-10.40, adjusted P = .046] and 4.38 [95% CI, 1.46-16.36, adjusted P = .014]), respectively. OSA was not associated with difficult mask ventilation, and only increasing neck circumference was found to be associated (adjusted P = .002). CONCLUSIONS: Moderate and severe OSA were associated with difficult intubation, and increasing neck circumference was associated with difficult mask ventilation. A higher STOP-Bang score of 3 or more may be associated with difficult intubation versus STOP-Bang score of 0-2. Anesthesiologists should be vigilant for difficult intubation when managing patients suspected or diagnosed with OSA.

Factors influencing unsafe behaviors: A supervised learning approach.

Despite its potential, the use of machine learning in safety studies had been limited. Considering machine learning's advantage in predictive accuracy, this study used a supervised learning approach to evaluate the relative importance of different cognitive factors within the Theory of Reasoned Action (TRA) in influencing safety behavior. Data were collected from 80 workers in a tunnel construction project using a TRA-based questionnaire. At the same time, behavior-based safety (BBS) observation data, % unsafe behavior, was collected. Subsequently, with the TRA cognitive factors as the input attributes, six widely-used machine learning algorithms and logistic regression were used to develop models to predict % unsafe behavior. The receiver operating characteristic (ROC) curves show that decision tree provides the best prediction. It was found that intention and social norms have the biggest influence on whether a worker was observed to work safely or not. Thus, managers aiming to improve safety behaviors need to pay specific attention to social norms in the worksite. The study also showed that a TRA survey can be used to extend a BBS to facilitate more effective interventions. Lastly, the study showed that machine learning algorithms provide an alternative approach for analyzing the relationship between the cognitive factors and behavioral data.