P1.163. Early Postoperative Pain Trajectory Phenotypes After Oesophagectomy Within a Multimodal Analgesia Pathway

Topic Esophageal Cancer: Surgical Treatment of Esophageal Cancer Background Postoperative pain after oesophagectomy remains heterogeneous despite multimodal analgesia. Identifying early pain trajectory phenotypes may help recognise patients who require targeted analgesic escalation. We derived postoperative day 1–5 pain trajectory phenotypes and assessed associations with length of stay and readmission. Methods Consecutive oesophagectomy patients managed within a multimodal analgesia pathway were reviewed. Maximum daily pain scores were recorded contemporaneously. Patients with complete overall pain scores (0–10 NRS) recorded on postoperative days (POD) 1–5 were included; for each POD the maximum recorded overall score was used. Pain trajectory phenotypes were identified using k-means clustering (k=3) applied to z-standardised POD1–5 pain profiles, using multiple random initialisations and retaining the solution with the lowest within-cluster sum of squares. Phenotypes were summarised using, peak daily pain (maximum POD1–5), and pain burden quantified by area-under-the-curve (AUC) across POD1–5. Total length of stay (LOS) was compared across phenotypes using Kruskal–Wallis testing. Readmission rates were compared using chi-square testing. Results Consecutive oesophagectomy patients managed within a multimodal analgesia pathway were reviewed. Maximum daily pain scores were recorded contemporaneously. Patients with complete overall pain scores (0–10 NRS) recorded on postoperative days (POD) 1–5 were included; for each POD the maximum recorded overall score was used. Pain trajectory phenotypes were identified using k-means clustering (k=3) applied to z-standardised POD1–5 pain profiles, using multiple random initialisations and retaining the solution with the lowest within-cluster sum of squares. Phenotypes were summarised using, peak daily pain (maximum POD1–5), and pain burden quantified by area-under-the-curve (AUC) across POD1–5. Total length of stay (LOS) was compared across phenotypes using Kruskal–Wallis testing. Readmission rates were compared using chi-square testing. Conclusion Despite multimodal analgesia, nearly one in five patients demonstrated a persistent-high pain trajectory through POD5, with substantially greater peak pain and overall pain burden than low and moderate trajectories. However, pain phenotype was not associated with total LOS or readmission in this cohort. Trajectory phenotyping captures clinically meaningful heterogeneity beyond single-day comparisons and may still support targeted analgesic escalation; future work should link phenotypes to opioid use, mobilisation, complications and patient-reported outcomes.