Dongwoo Chae (1,2), Dong Woo Han (3), and Kyungsoo Park (1)
1. Department of Pharmacology, Yonsei University College of Medicine, Seoul, Korea 2. Brain Korea 21 Plus Project for Medical Science, Yonsei University, Seoul, Korea 3. Department of Anesthesiology and Pain Medicine, Anesthesia and Pain Research Institute, Seoul, Korea
Objectives:
Our study aims to develop predictive models of the time course of postoperative pain and nausea under patient controlled analgesia (PCA) treatment.
Methods:
Serial postoperative visual analogue scale (VAS) pain scores and severity of nausea scale (N) ranging from 1 to 10 of 28,656 patients were retrospectively collected. The typical time course of VAS is characterized by an initial surge of pain severity due to diminishing post-anesthetic effect followed by an eventual pain relief. A0, Kane, Kon, and Emax denote the percentage reduction of VAS relative to 10 at baseline, rate constant of disappearance of post-anesthetic effect, rate constant of pain relief, and the maximal percentage reduction of VAS. Covariate search was carried out to identify factors affecting the model parameters. Basal infusion rate was tested for its effect on Kon and Emax. N was described as a sum of two surge functions, each representing the early nausea occurrence due to surgical procedures and post-anesthetic effect and delayed nausea development due to PCA exposure. Covariates that significanly increased the probability of nausea occurrence were identified. The final VAS and nausea models were fitted using a training dataset consisting of 10,000 patients and validated using two test datasets each consisting of 10,000 and 8,656 patients.
Results:
General and spinal anesthesia showed distinct pain profiles with the latter characterized by a lower baseline VAS. In IV PCA patients, typical A0 estimates of general and spinal anesthesia were 60% and 90%, respectively, with similar values of 60% and 92% in epidural PCA patients. Kon was lower in epidural PCA compared to IV and PCA patients, with typical values of 0.087/h and 0.12/h, respectively. Emax was higher in IV PCA patients, with its typical estimate of 75% compared to 69% in epidural PCA patients. Younger age was significantly associated with higher A0, Kon, and Emax. Female gender was associated with lower Kon and Emax. In IV PCA patients, longer duration of anesthesia was associated with lower A0, Kon, and Emax. Higher basal infusion rate was positively correlated with Kon (IV PCA) and Emax (epidural PCA), suggestive of an analgesic dose-effect relationship. Female gender, older age, increased prescription frequencies of Keromin, Tridol, and Pethidine were associated with higher risk of nausea in both IV and epidural PCA patients. Higher basal infusion rate of PCA regimen was associated with a higher risk of nausea in IV PCA patients. The developed model was successfully validated using the test datasets.
Conclusions:
Our model successfully predicted the time courses of VAS and N under PCA infusion. The developed model would be useful in devising individualized PCA regimens under widely different situations to optimize pain and side effects management.
Reference: PAGE 27 (2018) Abstr 8569 [www.page-meeting.org/?abstract=8569]
Poster: Drug/Disease Modelling - Other Topics