AUTHOR: Michael Vlessides
If there’s one clear message from a preliminary analysis of the Obstructive Sleep Apnea Death and Near Miss Registry, it’s this: Monitoring matters.
The data showed that effective monitoring solutions for the ward and home could reduce the significant perioperative adverse outcomes associated with obstructive sleep apnea (OSA), including severe brain damage and even death.
“OSA in the perioperative period is an important clinical problem,” said Karen Posner, PhD, the Laura Cheney Professor in Anesthesia Patient Safety at the University of Washington, in Seattle. “Although the risk of OSA in perioperative mortality is unclear, legal case reviews have identified OSA as an important factor in medical malpractice.
Posner and her colleagues analyzed data from 75 case reports of OSA-related adverse events reported to the OSA registry. There were several inclusion criteria:
- adults 18 years of age and older;
- OSA diagnosed or suspected;
- the event occurred after 1992; and
- the event occurred within 30 days of surgery and was suspected to be related to OSA.
Three researchers evaluated each case for the part that OSA may have played in the adverse event. Cases were then grouped according to outcome, either severe brain injury or death, or critical events such as respiratory arrest or ICU transfer without significant injury.
Of the 75 patients, OSA was diagnosed in 83% (by a sleep study) and suspected in the remaining 17% (by screening or patient history). Patients were middle-aged (52±15 years), ASA physical status III (58%) and obese (body mass index, 38±9 kg/m2). The majority underwent inpatient (76%) elective (85%) procedures under general anesthesia (91%). More than one-third of patients had severe OSA (36%).
At the 2018 annual meeting of the American Society of Anesthesiologists (abstract BOC01), Posner reported that 49% of OSA-related events occurred on the hospital ward, compared with 20% in the PACU, step-down unit or ICU; and 19% occurred at home. “The other 12% occurred during the procedure or immediately afterwards before handing the patient off to recovery,” she said.
Events on the ward occurred between two and 64 hours after the end of anesthesia (mean, 16 hours); 54% occurred within 12 hours of anesthesia’s conclusion. By comparison, events at home occurred from three hours to four days after anesthesia’s termination, with six events (43%) occurring within 24 hours. Perhaps not surprisingly, most patients (76%) were receiving opioids at the time of the event.
“When we look at outcomes,” Posner said, “63% of cases had death or severe brain injury [n=47], while the remaining 37% were critical events without significant patient injury [n=28].” Outcomes did not vary by age, sex, body mass index, ASA physical status, OSA severity (for patients with this diagnostic detail) or opioid use.
Perhaps not surprisingly, the analysis also revealed that monitoring on the ward at the time of the event was often absent (43%) or was classified as intermittent pulse oximetry (35%). In contrast, most patients in the PACU, step-down unit or ICU were monitored by continuous pulse oximetry with central monitoring (69%). No patients at home were monitored.
“Of course, we know that these different locations have different monitoring patterns,” Posner explained. “So, we put both monitoring and location into a logistic regression model, and only the presence or absence of respiratory monitoring was significantly associated with outcomes, where rescue was six times more likely to occur in locations with respiratory monitoring in place compared to no respiratory monitoring.”
“Effective monitoring solutions for the ward and home could potentially reduce these perioperative adverse outcomes associated with OSA,” she pointed out.
“I wonder how the critical incidents discussed in the study were detected,” commented Paul Myles, MBBS, MPH, DSc, a professor of anesthesia and perioperative medicine at Monash University School of Medicine in Melbourne, Australia. “For instance, it would make sense to me that they’re more likely to be picked up in the PACU or operating room environment, but much less likely to be picked up on the ward or at home. Would that explain the differences in the odds ratio in your data?”
“I think certainly at home, patients were less likely to get picked up until much later, when rescue opportunities were much rarer,” Posner replied. “So certainly that’s very relevant. We also looked at some events that were witnessed, both at home and in the ward,” she added. “We’re looking at those data as well.”