If there were any doubts about the efficacy of the STOP-Bang questionnaire to predict perioperative complications, they can be put to rest, thanks to a meta-analysis by a group of Ontario researchers. The investigators found that patients who score high on the obstructive sleep apnea (OSA) STOP-Bang questionnaire are at higher risk for perioperative complications, justifying use of the tool to screen patients during the perioperative period.
According to Mahesh Nagappa, MD, assistant professor of anesthesia and perioperative medicine at the University of Western Ontario in London, it’s well documented that surgical patients who suffer OSA are at increased risk for perioperative complications. Since the STOP-Bang questionnaire helps identify high-risk OSA patients, the purpose of the meta-analysis was to compare perioperative complications in patients with high STOP-Bang scores (3 or greater) and low STOP-Bang scores (0-2).
Looking for Perioperative Complications
To that end, the investigators performed a literature search for the 2008-2016 period, yielding 119 citations. Studies were included if they used the STOP-Bang questionnaire as a screening tool for high-risk and low-risk OSA in an adult surgical population and mentioned perioperative complications associated with high and low scores.
“Perioperative complications were either cardiac or respiratory events, or an ICU admission,” Dr. Nagappa said.
In the end, 11 studies qualified for inclusion, comprising 20,482 patients, with 7,598 in the high STOP-Bang score group and 12,884 in the low STOP-Bang score group. “Because the STOP-Bang [questionnaire] has a very high negative predictive value, we can say with great confidence that patients in the low STOP-Bang group were not prone to moderate or severe OSA,” he explained.
As Dr. Nagappa reported at the 2016 annual meeting of the Canadian Anesthesiologists’ Society (abstract 152726), the analysis revealed that the odds of suffering perioperative complications were significantly greater in the high STOP-Bang score patients than their counterparts with low STOP-Bang scores (odds ratio [OR], 3.83; 95% CI, 1.75-8.36; P=0.0008).
“We found that even though there was huge variation between the included studies, the direction was the same in all of them,” Dr. Nagappa said. “They all showed that patients with high STOP-Bang scores had a greater risk of postoperative complications.”
Given these results, Dr. Nagappa noted that the next step for institutions is to incorporate the STOP-Bang tool into their practice to reap its full benefits. “A recent paper [Crit Care Nurs Clin North Am 2014;26:499-509] showed that using the STOP-Bang questionnaire for a period of two years brought down respiratory complications by more than 50%,” he said. “And the use of naloxone in the PACU also came down by more than 50%.
“So finally we can say with great confidence that the STOP-Bang [questionnaire] can be used in a very flexible way to bring down the complication rates in the PACU.”
An Action Plan for High-Risk Patients?
However, as session co-moderator Marie-Hélène Tremblay, MD, noted, the preponderance of high-risk patients makes adoption of some sort of screening tool imperative. “I see that there were some 7,000 high-risk patients; that’s quite a lot,” said Dr. Tremblay, professor of anesthesia at Laval University in Quebec City. “What would you suggest when it comes to managing those patients?”
“There is a gap in the literature with respect to what action plan should be implemented for the high-risk OSA patient,” Dr. Nagappa replied. “Simply using the STOP-Bang questionnaire can create a sense of awareness among the entire perioperative team—and that alone can decrease the rate of complications.
“We don’t have an answer in the literature with respect to what a clear action plan should be,” he said.
This phenomenon, he added, may be partly compounded by the many questions still unanswered with respect to OSA. “What we do not know is whether all OSA patients are the same,” Dr. Nagappa said. “Is it related to position—supine-related or non–supine-related OSA? Or is there a specific OSA phenotype? Who are more prone to complications? We do not have answers to a lot of questions in the literature, and further work needs to be done.”