Back to 2026 Abstracts
5x-Multiplier vs. 3-Tier Model: A Pragmatic Randomized Clinical Trial for Discharge Opioid Prescriptions After Intra-Abdominal Cancer Surgery
*Ching-Wei D. Tzeng1, *Brittany C. Fields
1, *Heather A. Lillemoe
1, *Anneliese Hierl
1, *Jose A. Karam
2, *Surena Matin
2, *Larissa Meyer
4, *Zhouxuan Li
3, *Wei Qiao
3, Jean-Nicolas Vauthey
1, Matthew H. Katz
1, *Timothy Newhook
11Surgical Oncology, MD Anderson Cancer Center, Houston, Texas; 2Urology, MD Anderson Cancer Center, Houston, Texas; 3Biostatistics, MD Anderson Cancer Center, Houston, Texas; 4Gynecologic Oncology, MD Anderson Cancer Center, Houston, Texas
Objectives: Implementation of standardized prescribing models is associated with reduced discharge opioid prescription volumes, mitigating potential for chronic dependence and diversion after major abdominal surgery. Comparison between prescribing strategies has not been performed, and thus the ideal discharge model remains unknown. This study's objective was to determine which of two validated discharge opioid prescribing models resulted in both fewer prescribed opioids and less post-discharge consumption.
Methods: This was a pragmatic single-center, multi-specialty, phase II randomized clinical trial comparing two discharge opioid prescribing models: the linear "5x-multiplier" algorithm (e.g., last 24 hour oral morphine equivalents [OME] multiplied by 5) versus the capped "3-tier" model (5-15-30 pills depending on 0, 1-29, ≥30mg OME in last 24 hours), applied at hospital discharge (Fig. 1). Adult patients without opioid dependence undergoing open abdominal (liver, pancreas, ovary, sarcoma, kidney) cancer operations, performed by 25 surgeons from 5 specialties, were included. All patients received a non-opioid analgesic bundle perioperatively and at discharge. Co-primary endpoints were discharge opioid prescription volume and 14-day post-discharge consumption. Secondary endpoints included refill rates and symptom inventory scores. Using the two-sample t test and assuming a standard deviation of 70mg for discharge OME and 35mg for 14-day consumption with 20% dropout, this required 75 patients/arm to have 80% power in detecting a mean difference in OME.
Results: From April-December 2024, 150 patients (52% female; median age 63 years) were randomized: 73 to 5x-multiplier; 77 to 3-tier model. Last follow-up was June 2025. Operations performed included hepatectomy (32%, n=48), pancreatectomy (29%, 43), nephrectomy (13%, 20), thoracoabdominal sarcoma resection (15%, 22), and ovarian cytoreductive surgery (11%, 17). Inpatient opioid use was similar between arms. Median discharge OME was 25mg in 5x-multiplier and 75mg in 3-tier arms (p<0.001; Fig. 2), with 44% of 5x-multiplier patients discharged opioid-free (vs 1% in 3-tier). Median post-discharge 14-day opioid consumption was 0mg for 5x-multiplier versus 10mg for 3-tier (p=0.496). Refill rates were 24% for 5x-multiplier versus 18% for 3-tier (p=0.426), consistent with historical rates. There were no differences between study arms in validated symptom inventory scores at 14-, 30-, or 90 days (Fig. 3).
Conclusion: In this pragmatic randomized clinical trial, the 5x-multiplier algorithm resulted in less opioids prescribed at discharge with similar 14-day opioid consumption, refill rates, and symptom inventory scores, compared to a 3-tier model after major abdominal cancer surgery.
