Comparison of total morphine milligram equivalents at hospital discharge between opioid-naive and opioid-experienced surgical patients: a single-centre retrospective cohort study.

Background Perioperative pain management is a key concern amid the growing opioid pandemic, particularly for opioid-experienced patients. This retrospective single-centre cohort study aimed to compare morphine milligram equivalents (MME) at hospital discharge between opioid-naive and opioid-experienced adults undergoing surgery with postoperative patient-controlled analgesia (PCA). We hypothesised that opioid-experienced patients would require higher MME at discharge, and greater intraoperative remifentanil and postoperative PCA use.Methods We retrospectively analysed 406 patients from 2016 to 2023 who received intravenous PCA for acute postoperative pain management. Trauma and neuraxial/regional block cases were excluded; emergency non-trauma cases included. Opioid-experienced patients were defined as chronic use of opioids for ≥3 months before surgery. The primary outcome was opioid dose at discharge in MME. Secondary outcomes were total intraoperative remifentanil dose and total PCA use in MME, analysed using multiple linear regression with permutation testing.Results Opioid-experienced patients had a 15.4 MME day-1 higher discharge opioid dose (95% confidence interval [CI] 7.4-23.4 MME day-1; P<0.001), received 6.7× more opioids at discharge than opioid-naive patients (63.5 vs 9.4 MME day-1; P<0.001) and nearly doubled their own preoperative use (63.5 vs 30 MME day-1). Opioid-experienced patients also required 52.0 MME day-1 more via PCA (95% CI 13.1-90.8 MME day-1; P=0.009). Each additional preoperative MME was associated with a 0.9 MME day-1 increase in PCA use during the hospitalisation (95% CI 0.2-1.6 MME day-1; P=0.017).Conclusions Preoperative opioid experience strongly predicted postoperative opioid requirements and discharge prescribing. Early identification of opioid-experienced patients and tailored multimodal strategies may improve individualised pain management. However, the retrospective single-centre design and lack of non-opioid analgesia data limit generalisability.