Background : Respiratory suppression is observed during endoscopy under sedation. If respira tory suppression can be predicted before endoscopy, incidental complications can conceivably be prevented. In the present study, we focused on the relation between respiratory suppression from sedation and lung function. Methods : A total of 211 patients underwent respiratory function tests before the surgical operation and gave written informed consent individually to participate in this study. We investigated the relation between respiratory suppression from sedation and lung function. During the endoscopic procedure, when blood oxygen saturation (Sp02) fell to below 90%, the patient was evaluated as 'respiratory suppression present'. Results : Sedation lowered Sp0
2 by an average of 6.0%, and was significantly lower than the prior to sedation blood oxygen saturation (PreSpO2). Compared to patients with Sp0
2 maintained up to 90%, patients with Sp0
2 fallen below 90% were significantly older, shorter in stature, lighter in bodyweight, and more commonly female. Furthermore, respiratory suppression from sedation was influenced by vital capacity (VC) and PreSpO
2. Multivariate analysis was performed, and the receiver operating characteristic (ROC) curve constructed for the respiratory suppression prediction model based on age, height, VC and PreSpO
2 yielded area under the curve (AUC) of 0.79. As VCpredjct can be calculated from age and height, the three variables of age, height and VC in the above model were substituted with VCpredict resulting in a two-factor model based on VCpredict and PreSpO
2. The ROC curve of the two-factor model had AUC of 0.77, which was slightly decreased but by no means inferior. Conclusion : Predicting respiratory suppression from VCpredict and PreSpO
2 is clinically rele vant with the additional benefit of simplicity.
View full abstract