In Reply

We appreciate the thoughtful comments engendered by the updated Guidelines for Perioperative Management of Patients with Obstructive Sleep Apnea (OSA).¹  As Dr. Corso et al. mention, there are numerous clinical methods for assessing the likelihood that a patient suffers from obstructive sleep apnea. Chung et al. demonstrated that the STOP and Berlin questionnaires as well as the checklist proposed in the American Society of Anesthesiologists Guidelines are all valid screening tools for OSA in surgical patients,²  and no statistically significant differences regarding specificity and positive predictive values were found among the three screening questionnaires. While STOP and STOP-BANG scores were shown to correlate with the probability of sleep apnea, it was not established that they correlate with its severity.³  Because we did not find clear evidence of improved benefit from using any one particular OSA screening questionnaire, the key consideration is obtaining information relevant to an assessment of the potential for sleep apnea from the patient or the patient’s family.

The primary role of the “assessment of risk” checklist in the Guidelines is to encourage anesthesiologists and surgeons to consider the various factors which might predispose an individual patient to develop complications in the perioperative period. Patients with a score of 5 or 6 may be inappropriate candidates for outpatient surgery, and in some cases may warrant continuous monitoring of ventilatory function for several days postoperatively. Unfortunately, there is no guarantee that patients with a score of “4” will always have a favorable outcome following discharge to an unmonitored setting; residual anesthetics, sedatives, and opioid analgesics all increase the risk that a patient will fail to “self-resuscitate” from an obstructive episode. While there is no guarantee, observing patients while breathing room air in an unstimulated environment following emergence from anesthesia offers some degree of assurance that residual anesthetic effects will not contribute to an adverse outcome following discharge.

From a safety perspective, one could argue that any patient with even the slightest risk of sleep apnea should be monitored postoperatively in an acute care setting for several days, so that any episode of severe hypoxemia could be recognized and treated by trained personnel, as suggested by Dr. Rothfield. Such a requirement could mandate a significant increase in monitoring capabilities for hospitalized OSA patients, but also preclude the possibility of conducting outpatient surgery for many patients with suspected or confirmed OSA. While some hospitals have instituted continuous oximetry and capnography with central station monitoring for all postoperative patients diagnosed with OSA, others may be reluctant to do so because of cost and implementation issues. Until we have a reliable method of selecting the small fraction of OSA patients most likely to fail to “self-resuscitate” from a postoperative sleep apnea episode, we must either (1) admit all at-risk patients—even those who are undiagnosed or would normally qualify for outpatient surgery—to a monitored hospital setting for up to 3 days (because of rapid eye movement–rebound risk) or (2) accept the risk that in the absence of universal monitoring rare but serious postoperative complications are bound to occur.

The Guidelines state that, “An RCT indicates improved ventilatory function for OSA patients when postoperative CPAP is compared with no postoperative CPAP.”⁴  Dr. Roesslein correctly notes in his letter that this RCT did not compare “CPAP” with “no CPAP.” Rather, it compared early (immediate) post-extubation CPAP with 30-min post-extubation initiation of CPAP, and respiratory measurements were not taken within the 30-min “non-CPAP” phase of the comparison group. Although information obtained from a few case reports^5–7  suggests that postoperative institution of CPAP may improve postoperative pulmonary function in OSA patients, we agree that more well-controlled studies are needed to correctly evaluate the beneficial effects of postoperative CPAP. On the basis of this limited information and our survey findings, the assessment of risk included in the Guidelines assigns a decreased risk of perioperative complications to OSA patients who are compliant with CPAP and will be using this modality during the postoperative period.