IMMEDIATE COLECTOMY VERSUS ENHANCED SURVEILLANCE FOR LOW-GRADE DYSPLASIA IN ULCERATIVE COLITIS

Immediate Colectomy versus Enhanced Surveillance for Low-Grade Dysplasia in Ulcerative Colitis

Elliot G Arsoniadis^1,2; Karen Kuntz³

¹Department of Surgery and ²Institute for Health Informatics; ³Division of Health Policy & Management, School of Public Health, University of Minnesota, Minneapolis, Minnesota

Purpose: Patients with ulcerative colitis (UC) are at increased risk of colorectal cancer and are recommended to undergo screening colonoscopy after 8-10 years with disease. Patients with colon cancer or high-grade dysplasia diagnosed by screening colonoscopy undergo colectomy (total proctocolectomy). However, it is unclear whether patients with low-grade dysplasia should undergo enhanced surveillance rather than immediate colectomy and how patient preferences affect this decision.

Methods: We developed a state-transition Markov model to determine the best approach for UC patients diagnosed with low-grade dysplasia on screening colonoscopy. Current literature was used to provide transition probabilities and utilities for our model. Base-case utilities for the colectomy and surveillance states were 0.887 and 0.94, respectively.  These were based on previously reported utilities from the literature for those patients having undergone colectomy with its inherent long-term morbidity and those with UC in remission but with the added burden of frequent surveillance colonoscopy.

Results: In the base case, immediate colectomy resulted in 2.0 more quality-adjusted life years (QALYs) compared with enhanced surveillance (25.5 vs. 23.5 QALYs, respectively). In order for the two strategies to have equivalent effectiveness (as measured by QALYs) the utility associated with colectomy would have to decrease to 0.762. The figure shows a two-way sensitivity analysis comparing the utilities for the colectomy and surveillance states, where the upper left-hand region favors colectomy and the lower right hand region favors enhanced surveillance. 

Conclusion: Although immediate colectomy yielded a greater benefit over enhanced surveillance, our model showed that this result was somewhat sensitive to patient preferences for the colectomy state relative to the enhanced surveillance state. Our results underscore the fact that for patients with long-standing UC and low-grade dysplasia, the decision to proceed with colectomy versus surveillance should consider patient preferences, particularly patients with strong aversion to undergoing immediate colectomy. This model is one tool that could be used to help engage patients in shared decision-making when faced with this issue.