Regional Anaesthesia
1699
Prediction of the need for additional analgesia in the postoperative period.
Pain,postoperative, Safety,techniques
S. Kananchuk1, Ilukevich G.V., Abramovich M.S., Seliun Y.A.
111-18 - Minsk (Belarus)

Materials and Methods: Based on the NCCB, a prospective, randomized study was conducted between 2013 and 2016, analyzing the adequacy of postoperative analgesia after radical cystectomy in 107 patients with bladder cancer aged 61.0 years (95% CI: 56.0; 67.0). All patients in the postoperative period received long-term epidural or paravertebral analgesia. Patients were divided into two groups (Gr), depending on the level of the subjective verbal analogue scale (VAS). 1st Gr with VAS ≤ 3 points. 2nd Gr with VAS ≥ 4 points. 1st Gr consisted of 91 patients who did not need additional analgesia in the postoperative period, and 2nd Gr consisted of 16 patients requiring additional analgesia. To assign patients to one of the groups, decision rules were built using decision trees. classifier in the study used decision trees built using the C&RT algorithm. The method of decision trees makes it possible to develop a model and to predict the belonging of observations to a particular class of categorical function, even at the preoperative stage, depending on the corresponding values of one or several prediction characteristics. As a criterion of agreement, the Gini criterion was used. The algorithm was terminated by the percentage of unclassified observations at terminal vertices. To assess the suitability of decision trees for the classification of the examination sample, a V-cross-check was conducted. As information signs for the construction of the "decision trees" classifier, all the initial (preoperative) indices were taken and those were chosen whose p-values were less than 0.05 (mean arterial pressure (MAP), heart rate (HR), pH, intraoperative hypotension (IH)).
Results and Discussion: Based on the constructed classifier tree, we present an algorithm for assigning patients to 1st or 2nd Gr. Step 1. If the symptom of HR is at the most traumatic moment < 65.5, then the patient refers to 1st Gr. Step 2. If the symptom of HR is at the most traumatic moment > 65.5, and the sign of MAP < 96.5, then the patient belongs to the 1st Gr. Step 3. If the symptom of HR is at the most traumatic moment > 65.5 and the sign of MAP > 96.5, and the sign of IH=1, then the patient refers to the 2nd Gr. Step 3. If the symptom of HR is at the most traumatic moment > 65.5 and the sign of MAP > 96.5, and the sign of IH = 2, then the patient belongs to the 1Gr.
Conclusion(s): The developed algorithm allows to accurately predict the need for additional analgesia in the postoperative period in 92.52% of cases.


09 / Prediction of the need for additional analgesia in the postoperative period.