A1727
October 22, 2008
9:00 AM - 11:00 AM
Room Hall E2-Area D,

Postoperative Complications in Surgical Patients at Risk for Obstructive Sleep Apnea
Kevin J. Finkel, M.D., Lini Zhang, M.D., Heidi Hahn, M.H.S., Adam C. Searleman, B.S., Michael S. Avidan, M.D.
Anesthesiology, Washington University, St Louis, MO

Introduction:
Obstructive sleep apnea (OSA) affects approximately 20% of the U.S. adult population¹ with up to 90% of cases undiagnosed.² In a previous study, we demonstrated that OSA was common among adult surgical patients at a tertiary care university hospital. Patients with OSA are thought to be more likely to experience peri-operative complications, but this has not been demonstrated in a large prospective study. We therefore designed a prospective, observational study of adult surgical patients to determine if there was an association between screening high risk for OSA and peri-operative complications.

Methods:
Consecutive adult surgical patients were screened for OSA risk with the Berlin Questionnaire³ and the Flemons' Index.⁴ We determined OSA risk with three approaches: i) Berlin Questionnaire; ii) Flemons' Index; and iii) Both screens positive. All three approaches included patients with a known history of OSA. The medical records were audited during their hospital stay for perioperative complications and in-hospital mortality. Odds ratios were calculated using logistic regression.

Results:
We screened 4427 surgical patients over seven months: 1716 (39%) had a positive Berlin questionnaire, 1270 (29%) had a positive Flemons' Index, 959 (22%) were positive with both screens, and 549 (12%) had a history of OSA. Patients who screened high risk for OSA by any of the three methods were more likely to have diabetes mellitus, arrhythmias, coronary artery disease, stroke, congestive heart failure, or chronic obstructive pulmonary disease than those who did not screen high risk for OSA (ps < 0.05). Each approach was associated with increased risk of ICU admission or postoperative intubation but not mortality; the greatest risks were with the combined screen.

Discussion:
Our preliminary data suggest that 22-39% of patients undergoing surgery are at high risk for OSA. Patients who screened high risk had more co-morbidities and had a higher incidence of postoperative intubation and ICU admission. OSA screening may help to identify patients who are at increased risk for certain peri-operative complications. Further research is required to determine whether OSA risk is directly associated with certain peri-operative complications or whether it represents an epiphenomenon. The most reliable and cost effective method to identify and quantify OSA risk among surgical patients needs further exploration.

References:
1. Young T et al. NEJM 1993; 328:1230-1235.
2. Young T et al. Sleep 1997; 20:705-706.
3. Netzer N et al. An Intl Med 1999; 131:485-91.
4. Flemons W, et al. Am J Resp Crit Med 1994; 150: 1279-85.