Allocating a US$12 Million Public Health Budget
A state health department has an additional US$12 million for one budget cycle and must choose between expanding two programmes. Option A is school-based HPV vaccination of 12-year-old girls: two doses at an estimated US$30 per fully vaccinated girl delivered, a cohort of 60,000 girls, 80% coverage, and 70% vaccine efficacy against cervical cancer death. Assume a 1.5% lifetime risk of cervical cancer death among unvaccinated women and 25 DALYs lost per death. Option B is community hypertension screening and treatment for adults aged 40 and over: screening costs US$3 per person, 40,000 people are screened, 25% are hypertensive, 60% of hypertensives are retained on treatment for 10 years at US$40 per patient-year, treatment reduces 10-year cardiovascular events by 25%, the untreated event rate is 10% over 10 years, and each event costs 5 DALYs. Cold-chain capacity is stretched; the state has a strong primary care network and an election in 14 months. The cost-effectiveness threshold is US$1,500 per DALY averted.
1. Calculate the cost per DALY averted for each option and interpret against the US$1,500 threshold, stating your assumptions.
Read the worked answer
Option A: 60,000 × 0.8 = 48,000 girls vaccinated; cost = 48,000 × US$30 = US$1.44 million. Deaths averted = 48,000 × 0.015 × 0.7 = 504; DALYs = 504 × 25 = 12,600; cost per DALY = 1.44 million/12,600 = US$114. Discounting at 3% to present value, with benefits spread over roughly 30 years and an average factor of about 0.65, gives about 8,190 DALYs and about US$176 per DALY - still far below the threshold.
Option B: screening = 40,000 × US$3 = US$120,000. Hypertensives = 10,000; retained = 6,000. Treatment = 6,000 × US$40 × 10 = US$2.4 million; total = US$2.52 million. Events averted = 6,000 × 0.10 × 0.25 = 150; DALYs = 750; cost per DALY = US$3,360 - above the threshold.
Assumptions matter: vaccine efficacy and duration, cervical cancer incidence, coverage, and herd effects for A; retention, adherence, event rates and side-effect costs for B. Even with generous assumptions - HPV efficacy 50% and discounting - Option A remains an order of magnitude cheaper. Option B would need retention above about 90% and higher event rates to approach the threshold. On efficiency grounds the comparison is not close, but the calculations are only as good as the local parameters and should be tested in sensitivity analysis.
2. The health secretary prefers Option B because strokes are happening now and vaccinated girls will not benefit for 20 years. How do you respond, and what allocation do you recommend?
Read the worked answer
Acknowledge the legitimate concern: strokes are occurring now, and a programme with visible results has political and equity value. But immediacy is a time-horizon argument, not a cost-effectiveness argument, and discounting already penalises HPV. At 3% discounting, HPV still costs roughly US$176 per DALY, about one-twentieth of the hypertension programme as designed.
The hypertension programme is poor value because it treats everyone screened and retains only 60%; much of the spending goes to people who will not benefit. A better version - targeting people with existing cardiovascular disease, diabetes or very high blood pressure - would have a much larger absolute risk reduction and could fall below the threshold.
Recommendation: fund Option A fully at US$1.44 million, add a catch-up campaign for girls out of school because school-based delivery reaches the easiest-to-reach and can widen inequity, and redesign rather than abandon hypertension work through risk stratification, fixed-dose combinations and task-sharing. Use the remaining budget for other cost-effective interventions. Commit to monitoring and cervical cancer screening, since vaccination does not remove the need for screening. Present the numbers transparently and offer a phased plan with a 12-month review, so the political demand for visible action is met without sacrificing efficiency.
3. What non-cost factors and process safeguards should shape a defensible decision?
Read the worked answer
Non-cost criteria: equity - rural girls, out-of-school girls and adults with uncontrolled hypertension are different constituencies; budget impact and sustainability - vaccination is a one-time cost per cohort, while treatment is recurrent and grows; health-system capacity - cold chain, school platform, laboratory and drug supply; acceptability and trust, including vaccine hesitancy and rumours; legal and human-rights considerations; and political accountability. Consider opportunity cost: the US$12 million cannot fund something else, so compare against the marginal alternative, not against zero.
Process safeguards: publish the criteria and weights in advance; convene an independent technical advisory group with declared conflicts of interest; conduct probabilistic sensitivity analysis and state what would change the decision; consult community representatives and disclose the trade-offs. Guard against the visible-versus-statistical bias and against election-cycle pressure to favour short-term outputs. Document dissent rather than suppressing it. Plan an evaluation with process and outcome indicators - coverage, completion of the two-dose schedule, cold-chain failures, hypertension retention and blood pressure control - and include a review clause so the decision can be corrected. Transparency about assumptions and uncertainty is what makes the choice defensible, not the precision of the cost-per-DALY figure.