Back to 2025 Abstracts
FACTORS ASSOCIATED WITH EARLY VS. DELAYED PCA DISCONTINUATION AFTER PANCREATECTOMY
Yifan Wang
*, Brittany C. Fields, Anneliese N. Hierl, Zhouxuan Li, Laura R. Prakash, Morgan L. Bruno, Elsa M. Arvide, Whitney L. Dewhurst, Jessica E. Maxwell, Naruhiko Ikoma, Rebecca A. Snyder, Michael P. Kim, Matthew Katz, Ching-Wei D. Tzeng
Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX
Background: Intravenous patient-controlled analgesia (PCA) accounts for the majority of inpatient opioid exposure after pancreatectomy. While standardization of initial PCA settings can limit rapid cumulative opioid exposure, parameters to guide timing of PCA discontinuation are lacking. The aim of this study was to identify the prevalence and risk factors for extended PCA use.
Methods: Patients who underwent pancreatoduodenectomy (PD) or distal pancreatectomy (DP) from 11/2020-05/2024, were identified from a prospectively maintained quality improvement database. This period coincided with the start of "version 3" of our risk-stratified pancreatectomy clinical pathways, requiring immediate recovery room use of a 3-drug multimodal non-opioid analgesia and preset initial PCA settings, with recommended PCA discontinuation by the end of postoperative day (POD) 2. "Early"
vs. "delayed" PCA discontinuation (PCA-stop) was defined as by POD2
vs. on POD 3 or later, respectively. Patients undergoing central or total pancreatectomy, multivisceral resections, or those with preoperative long-acting opioid use were excluded. Multivariable logistic regression was performed to examine factors associated with early
vs. delayed PCA-stop.
Results: Of 510 consecutive patients, 399 (78.2%) received an intravenous PCA postoperatively. PCA was prescribed in 275 (85.9%) patients undergoing PD
vs. 124 (65.3%) patients undergoing DP. A minimally invasive approach was less common for PD compared to DP (12/275 [4.4%]
vs. 29/124 [23.4%], p<0.001). Cumulative PCA opioid exposure on POD 0-2 was higher after PD (70
vs. 57mg, p=0.029), but comparable between groups on POD3 and beyond (13
vs. 11mg, p=0.24). The rate of delayed PCA-stop was significantly higher after PD (73.5%
vs. 47.6%, p<0.001). Once discontinued, PCA was restarted in only 1.3% (5/399) cases.
On adjusted analysis, PD (OR 2.14, p=0.002), higher PCA use during POD 0-2 (OR 1.07, p=0.002), and delayed gastric emptying (OR 3.22, p=0.012) were independently associated with delayed PCA-stop, whereas minimally invasive surgery (OR 0.34, p=0.004) was associated with early PCA-stop. Decision tree analysis showed that the most important determinant of timing of PCA-stop was PD
vs. DP. In patients undergoing PD, a second-level node dichotomized patients based on POD 0-2 PCA use above or below a cutoff of 51 mg (equivalent to ten tramadol 50-mg pills).
Conclusion: Delayed PCA discontinuation occurred in nearly 75% of patients after pancreatoduodenectomy and 50% of patients after distal pancreatectomy, despite equally limited PCA cumulative use after POD2. Patients undergoing minimally invasive surgery, distal pancreatectomy, with early return of gastric function, and/or with low cumulative POD 0-2 PCA use may tolerate early PCA discontinuation to reduce inpatient opioid exposure, with a low risk for PCA restart.
Table 1. Overview of demographic characteristics and postoperative opioid requirements.
Table 2. Univariable and multivariable logistic regression analyses to identify factors predictive of delayed PCA discontinuation.
Back to 2025 Abstracts
