Choosing a Colorectal Cancer Screening Strategy Under a Hard Budget Cap
The Ministry of Health must choose a colorectal cancer screening strategy for 1.2 million people aged 50–74 over ten years. The technical unit appraises three options against a 'no organised screening' baseline using local cost and epidemiological data. Option A, biennial faecal immunochemical testing with colonoscopy for positives: cost $18 million, 1,850 life-years gained, projected uptake 52%. Option B, colonoscopy every ten years: cost $48 million, 2,170 life-years gained, projected uptake 28%. Option C, a media campaign with no organised testing: cost $4 million, 180 life-years gained. GDP per capita is $28,000 and the ministry's informal threshold is one GDP per capita per life-year. The screening budget is capped at $40 million over the period, and three of the six governorates have no endoscopy unit. Past campaigns achieved uptake well below projections.
1. Calculate the cost per life-year gained for each option against baseline, and the incremental cost-effectiveness ratio of B versus A. Which options are cost-effective at the ministry's threshold?
Read the worked answer
Average cost per life-year gained versus no screening
- A: $18,000,000 / 1,850 = $9,730
- B: $48,000,000 / 2,170 = $22,120
- C: $4,000,000 / 180 = $22,222
All three fall below the $28,000 threshold, which is exactly where naive reading goes wrong. Average ratios compare each option to doing nothing and conceal the comparison that matters: what does the extra money buy relative to the next best alternative?
Incremental analysis (rank by cost)
- C to A: ($18m − $4m) / (1,850 − 180) = $14m / 1,670 = $8,383 per life-year - A strongly buys worthwhile gains over C. C is dominated in extended terms and can be dropped.
- A to B: ($48m − $18m) / (2,170 − 1,850) = $30m / 320 = $93,750 per life-year - more than three times the threshold.
Conclusion. A is the cost-effective choice. B looks acceptable on an average ratio only because most of its benefit is shared with A; the additional 320 life-years cost $93,750 each, which the ministry cannot justify at $28,000. Note also that B is unaffordable anyway (§48m > $40m cap), and that affordability and efficiency are separate questions that must both be answered.
2. Given the $40 million cap and the missing endoscopy units, what would you actually recommend, and what are the equity and budget-impact consequences?
Read the worked answer
Recommendation. Adopt A as the national platform, costed at $18 million, and use part of the remaining $22 million deliberately: endoscopy capacity, pathology, quality assurance, a kit-return logistics system, a call–recall register, and a guarantee that every positive FIT is colonoscoped within 60 days. A screening test without a functioning follow-up pathway produces cost without benefit.
Why not B. It exceeds the cap, fails the incremental threshold, and its effectiveness depends on 28% uptake - the test is accurate, but participation is the binding constraint. Spending the whole envelope on a test two-thirds of the population declines is poor value.
Equity. The three governorates without endoscopy are precisely where FIT has an advantage, because it can be delivered through primary care with specimens transported to a central laboratory. But equity needs active management: rural populations have lower literacy for kit instructions, less reliable postal return, and longer travel. Plan community health worker distribution and return, and language-appropriate instructions, or the programme will widen the urban–rural gap in outcomes.
Budget impact. Phase rollout over 24 months, with the first wave in the underserved governorates to correct the existing imbalance. A 'phased B in one or two governorates' proposal should be resisted unless it is framed as a time-limited evaluation with a funded plan to equalise - introducing a superior service in the richest areas is the classic equity failure in screening policy.
Watch items. FIT positivity rates, colonoscopy completion, adenoma detection rate, interval cancers, and stage shift at diagnosis.
3. Critics raise two objections: life-years ignore quality of life, and the uptake projections are optimistic given past campaigns. How do these affect the decision, and how would you respond analytically?
Read the worked answer
Objection 1 - quality of life. A life-year is not a quality-adjusted life-year. Colorectal screening gains accrue mostly in people who are otherwise well and return to full function, so utility weights would be close to 1 and the ICERs would move little. The one place utilities bite is the harm side: colonoscopy complications, false positives, and the anxiety of repeated testing. The honest response is to redo the analysis with QALYs, apply sensitivity ranges to the utility decrements for complications, and discount costs and effects at 3% per year - undiscounted ten-year totals flatter all options, particularly the expensive one whose costs are front-loaded. The gap between $93,750 and $28,000 is large enough that plausible utility weights will not close it.
Objection 2 - optimistic uptake. Test it, do not argue about it. Run one-way sensitivity analyses on uptake for both arms, then probabilistic sensitivity analysis with distributions reflecting observed campaign performance. The threshold question is precise and answerable: B would need to deliver 30m/28,000 = 1,071 additional life-years over A, i.e. 2,921 in total, to reach the threshold. Given that uptake is the main constraint on B, that is implausible and should be stated with the calculation attached.
Decision impact. Neither objection reverses the ranking; both sharpen the implementation risk. Build a monitoring plan with pre-agreed stopping or modification rules: if FIT uptake falls below 40% at 18 months, or colonoscopy completion within 60 days falls below 80%, return to the technical unit rather than continuing to spend. State explicitly that the process used average ratios, incremental ratios, affordability and equity, and publish all four.