PURPOSE: To examine the evidence linking the presence of OSA in patients undergoing elective surgery with perioperative complications.
METHODS: A systematic review was conducted in Pubmed-Medline, Web of Science, Scopus, the Cochrane Database of Systematic Reviews & the Cochrane Central Register of Controlled Trials. Our search was restricted to cohort or case-control studies in adult (>18 years-old) patients until the end of October 2010, published in any language. Primary postoperative outcomes were postoperative desaturation, respiratory failure (RF), reintubation, myocardial infarction, and or ischemia, arrhythmias and cardiac arrest. Other major primary outcomes were ICU transfer and length of stay (LOS). Patients may have been diagnosed with OSA by a screening questionnaire, oximetry or Polysomnography. Studies without information on controls (patients without OSA), those involving upper airway surgery and those with diagnosis of OSA was on the basis of ICD-9 codes alone were excluded from the analysis.
RESULTS: Out of 6247 records, 13 studies were included in the final analysis (n=3942). OSA was associated with significantly higher odds of any postoperative cardiac events (45 /1195 [3.76%] vs. 24 /1420 [1.69%],respectively; OR 2.07; 95% CI 1.23-3.50, p=0.007 & respiratory failure (33/1680 [1.96%] vs. 24/3421 [0.70%]; OR 2.43, 95% CI 1.34-4.39, p=0.003. Effects were not heterogeneous among studies for these outcomes(I2=0 to 15%, p>0.3). Presence of OSA was also significantly associated with a higher odds of desaturation (189/1764 [10.71%] vs. 105/1881 [5.58%]; OR 1.96, 95% CI 1.48-2.60, p<0.00001,and ICU transfer (105/2062 [5.09 %] vs. 58/3681 [1.57%],respectively; OR 2.29, 95% CI 1.62-3.24,p<=0.00001. Both of these outcomes showed a significant degree of heterogeneity of effect among studies (I2=57 to 68%, p≤0.02). OSA was not associated with a higher LOS in two studies.(MD 1.02 days, 95% CI -0.94 to 2.97, p=0.3; I2=0%, p=0.5). We did not find evidence of publication bias from the evaluation of the funnel plots.
CONCLUSIONS: Incidence of postoperative respiratory desaturation, respiratory failure, postoperative cardiac events and ICU transfers was higher in patients with OSA.
CLINICAL IMPLICATIONS: Postoperative outcomes in patients can be improved if patients are screened for OSA before elective surgery.
DISCLOSURE: Frances Chung: Grant monies (from sources other than industry): ResMed Inc study of patients with Sleep Disordered breathing
Peter Gay: Grant monies (from industry related sources): RespMed Inc. Study of patients with sleep disordered breathing
The following authors have nothing to disclose: Roop Kaw, Vinay Pasupuleti, Jigesh Mehta, Adrian Hernandez
No Product/Research Disclosure Information