Analytical approach:
This economic evaluation was based on a Markov model with a patient lifetime horizon. The authors stated that the analysis was carried out from the perspective of the health care payer, which was a provincial health authority in Canada.
Effectiveness data:
The clinical data were derived from a thorough literature review. The inclusion criteria and the approach used to pool the clinical estimates from individual sources were not described. The clinical outcomes for stent placement were obtained from a systematic review of 54 non-randomised studies, which enrolled a total of 1,198 patients. The transition probabilities, in the long-term, for patients receiving surgery, were taken from the Surveillance and Epidemiology and End Results database. Survival curves were used to assess the long-term risk of events. The key clinical inputs were the success rates of the two surgical procedures.
Monetary benefit and utility valuations:
The utility estimates were derived from a published study which used patient preferences, elicited using the standard gamble method.
Measure of benefit:
Quality-adjusted life-years (QALYs) were used as the summary benefit measure and they were discounted at an annual rate of 3%. Other outcomes such as short-term mortality, recurrence rates and risk of perforation were also reported, but were not combined with the costs.
Cost data:
The categories of costs were the hospital services related to the procedures and the subsequent hospital stay, stent apparatus, management of stoma, and chemotherapy for recurrence. The length of hospital stay and most of the resource use data were based on expert opinion. The costs were derived from multiple sources, such as the Ontario Health Insurance Plan schedule of physician benefits, the Ontario Case Costing Project, the literature, and expert opinion. All costs were in Canadian dollars (CAD) and were also presented in US dollars ($). The price year was 2006 and the long-term costs were discounted at a 3% annual rate.
Analysis of uncertainty:
The authors used three approaches to deal with the issue of uncertainty: one- and two-way deterministic sensitivity analyses on the model inputs using published or authors-based plausible ranges, probabilistic sensitivity analysis using a second-order Monte Carlo simulation, and different patient profile scenarios (a high-risk patient profile was compared with a low-risk one). Furthermore, in an alternative scenario, it was assumed that the patient’s physiologic status was minimally impaired by the large bowel obstruction and they did not have an elevated American Society of Anesthesiology score (which was two in the base-case analysis).
