Projected effect of fecal immunochemical test threshold for colorectal cancer screening on outcomes and costs for Canada using the OncoSim microsimulation model. (September 2017)

• Record Type: Journal Article
• Authors: Coldman, Andrew; Flanagan, William; Nadeau, Claude; Wolfson, Michael; Fitzgerald, Natalie; Memon, Saima; Gauvreau, Cindy; Miller, Anthony; Earle, Craig
• Is Part Of: Journal of cancer policy. Volume 13(2017)
• Page Start: 38
• Page End: 46
• Publication Date: 2017-09
• Subjects: FIT Threshold -- CRC Screening; Cancer -- Government policy -- Periodicals; Cancer -- Patients -- Services for -- Periodicals; Medical Oncology -- Periodicals; Public Health -- Periodicals; Cancer; Periodicals; 362.196994
• DOI: 10.1016/j.jcpo.2017.07.004
• Languages: English
• ISSNs: 2213-5383

Abstract

Highlights: The effect of FIT threshold on colorectal cancer screening outcomes is modelled. Modelling uses reported results on FIT threshold and clinical sensitivity. Lower FIT thresholds are predicted to result in reduced cancer mortality. Cost effectiveness increases with declining FIT threshold. Abstract: Background: Immunochemical faecal testing (FIT) is used in Canada for colorectal cancer (CRC) screening. The threshold for FIT abnormality is under operator control and this study uses a simulation model to predict the effect of threshold selection on outcomes including years-of-life gained, CRC incidence and mortality, and direct health system costs. Method: The OncoSim Model was used to predict outcomes of biennial screening between ages 50–74 years in a cohort aged 45 years followed until death for eight FIT threshold values between 50 and 225 ng/ml. The literature on FIT performance was reviewed and seven parameter sets of sensitivity and specificity values by FIT threshold were created to span published variation in test performance in subjects with CRC, adenomas or no colorectal neoplasia. Results: Reducing the FIT threshold reduced both projected CRC incidence and mortality for all parameter sets, although cost impacts and cost-effectiveness varied compared to no screening. Biennial FIT was projected to be cost-effective at all thresholds considered with a maximum of CAD$ 5400 per QALY over the seven parameter sets and a maximum of CAD$ 6800 per QALY for a one level change in the eight threshold levels considered. Demand for colonoscopy varied strongly with FIT threshold and was greatest for the lowest threshold (50 ng/ml) but the magnitude varied across the 7 parameter sets. Conclusions: Compared to no screening, all thresholds of FIT examined were predicted to be cost-effective in the prevention and management of CRC. Threshold choice strongly influences predicted demand placed upon colonoscopy resources.