Unrecognized Obstructive Sleep Apnea and Postoperative Cardiovascular Complications: A Wake-up Call.

Abstract

In the United States, obstructive sleep apnea (OSA) affects 14% of adult men and 5% of adult women, with higher rates among obese individuals and older adults.1 The majority of individuals with OSA remain undiagnosed.2,3 Although OSA has repeatedly been associated with unfavorable long-term cardiovascular outcomes,4 it is increasingly recognized that patients with OSA are at risk for adverse events in the postoperative setting.5,6 Perioperative clinicians have traditionally equated the presence of OSA with complications affecting the respiratory system, but based on a small number of mostly retrospective studies, it is unclear whether patients with unrecognized OSA are at higher risk of postoperative cardiovascular events. In addition, because most patients with OSA will be undiagnosed at the time of surgery, questions remain regarding the importance of identifying this population preoperatively, as well as of risk-stratifying individuals for clinically relevant adverse outcomes. Quality data are needed to address this issue, because the results could significantly influence clinical care pathways for patients undergoing surgery. The study by Chan and colleagues7 in this issue of JAMA provides new information about some of these important issues. The Postoperative Vascular Complications in Unrecognized OSA (POSA) study was a large, multicenter, prospective observational cohort study designed to evaluate the association between the results of preoperative sleep testing for OSA and 30-day postoperative cardiovascular outcomes among patients undergoing noncardiac surgery. The primary outcome was a composite of myocardial injury, heart failure, thromboembolism, atrial fibrillation, stroke, and cardiac death. Among the study population of 1218 patients, the authors identified high rates of undiagnosed OSA; 37.1% of patients had mild OSA, 19.3% moderate OSA, and 11.2% severe OSA. Postoperative cardiovascular events occurred in 235 patients (19.3%). Multivariable analysis, controlling for perioperative factors known to adversely affect outcomes, revealed that severe OSA was significantly associated with a higher rate of postoperative cardiovascular events (adjusted hazard ratio [HR], 2.23 [95% CI, 1.493.34]; P = .001). However, this association was not seen among patients with mild OSA (adjusted HR, 1.36 [95% CI, 0.97-1.91]; P = .08) or moderate OSA (adjusted HR, 1.47 [95% CI, 0.98-2.09]; P = .07). Overnight oximetry monitoring was performed on the first 3 postoperative nights and found that higher risk for postoperative cardiovascular events was associated with longer duration of postoperative oxygen desaturation less than 80% (P < .001). No significant interaction was observed between the type of anesthesia, use of postoperative opioids, and supplemental oxygen therapy with perioperative outcomes. This study has several strengths that distinguish it from previous work in the field: the study is large, prospective, and well-organized; has predefined outcomes; and uses rigorous methodology and follow-up. A few of these strengths are worth discussing, as they suggest a causal relationship between OSA and postoperative cardiovascular complications. First, the study used standardized objective sleep testing (also known as home sleep apnea testing [HSAT]) using a portable type 3 sleep testing device in the 30 days before the operation to determine the presence and severity of OSA. Arguments can be made that this type of testing, because of an inability to measure sleep, may underestimate the severity of OSA and should not be used for population screening.2 However, the accuracy of HSAT is greater than the accuracy of questionnaire screening and is viewed as a reasonable alternative to expensive and resource-intensive polysomnography, particularly when time constraints exist (ie, preoperatively) and when quality oversight is ensured.2 HSAT is already widely used in clinical practice, although its feasibility for preoperative testing is institution-dependent, largely because of the infrastructure (eg, personnel) needed to support testing in the context of a comprehensive preoperative program. HSAT might best be used as a second risk-stratification step in patients identified as high risk for OSA by questionnaire screening. Regardless, the use of objective data to define OSA in the POSA study confirms that the rate of undiagnosed and unrecognized OSA in the surgical population remains high (67.6% in this study). Of interest, the majority of patients in this study were Asian and had a lower body mass index (mean, 26.8) than that seen in the population of the United States. Even though Asian patients have been shown to have similar rates of OSA at lower body mass index compared with white patients,8 many if not most of these patients would be missed by routine preoperative screening.9 Second, the study used predetermined clinically relevant outcomes, including a composite outcome as a primary outcome, which may initially be cause for criticism. However, when the data were analyzed independently for each indiRelated article page 1788 Opinion