In the article “Patients with a high risk for obstructive sleep apnea syndrome: Postoperative respiratory complications” by Pereira et al.,1 the authors aim to evaluate the incidence of STOP-BANG score≥3, in surgical patients admitted to the Post-Anesthesia Care Unit. It is rather alarming that the reported prevalence of patients with characteristics suggestive of obstructive sleep apnea (OSA) as evaluated by the STOP-BANG questionnaire was 52% (177 from 340). This number suggests that the majority of patients in the surgical population have undiagnosed OSA; however, it is unfortunate that in this study there was no polysomnography analysis of the patients and so, there was no confirmation of this estimated prevalence. If, however, we report to the study done by Chung et al.,2 author of the STOP and STOP-BANG questionnaires,3 which was also with a surgical population, we see that the reported percentage of patients with STOP-BANG≥3 is of 57.5% (429 from 746), which is close to the number from the study by Pereira et al.1 However, in the study by Chung et al.,2 these patients were analyzed with polysomnography and 75.3% of the group with SOTP-BANG score >3 were confirmed to have OSA (43.3% from the total of 746). If we accept that both of these surgical populations, from the study of Pereira et al.1 and the study of Chung et al.,2 are similar, we can infer that the prevalence of undiagnosed OSA in patients with STOP-BANG≥3 in the Pereira et al. study is 39% (177*0.75/340), somewhat lower than the reported estimated prevalence of 52%.
It is interesting to note that the authors also conclude that these patients with High Risk OSA (HR-OSA) have a higher incidence of postoperative respiratory complications and that mild/moderate hypoxia was the most frequent adverse respiratory event (ARE) to occur in these patients. If we analyze the reported ARE it seems that, HR-OSA patients were not associated with overall statistically significant higher number of ARE (p=0.140). Also, out of the 6 defined types of ARE, only the previously mentioned mild/moderate hypoxia (the definition of which is included in the methods section) was statistically more frequent in this HR-OSA group. It is also noteworthy that the HR-OSA population had a statistically significant greater ASA score, more frequent neuromuscular residual blocking, high-risk surgery and major surgery and, as we can see in Table 3 of the article, out of the 8 analyzed “predictors of adverse respiratory events” only HR-OSA was not a statistically significant predictor, contrary to others, such as: gender, intra-abdominal surgery, high risk surgery, major surgery, use of neuromuscular blocking drugs and residual neuromuscular blockade (and the latter was the only independent risk factor for ARE).
In conclusion, this study exposes a high percentage of patients with probable undiagnosed OSA that would probably benefit from a polysomnography analysis. However, we feel this study has failed to show that surgical patients with a high risk of OSA, as determined by STOP-BANG score≥3, are associated with higher incidence of ARE in the postoperative period or, that HR-OSA is a predictor of these ARE. However, previous papers have shown that OSA itself is associated with a statistically significant higher incidence of respiratory complications,4 increased intensive care unit admissions,5 and greater duration of hospital stay.6