Facilitators  ·  Economic Evaluation in Health  ·  23 September 2026

Poll question bank

Two or three questions at the end of each session, answered on phones. Correct answers are in green. Launch the question, give about 40 seconds, show the distribution, then read the explanation. Where the room splits, that is the discussion; do not rush it.

Every number here already appears on a slide. Two questions are opinion questions with no right answer; they give a before-and-after on confidence.

Session 1  ·  Where economic evidence enters health policy

Q1Before today, how many published economic evaluations in your own field have you read closely? opinion

  1. None
  2. One
  3. Two to five
  4. More than five

Baseline only. Re-ask at the close in a different form (the confidence question).

Q2Who starts an HTAIn assessment?

  1. The HTAIn Secretariat, on its own initiative
  2. A programme, ministry department or state that nominates a topic
  3. The manufacturer of the technology
  4. A Regional Resource Centre that wants to publish

The user nominates; the recommendation returns to the user, who decides. HTAIn advises, it does not command.

Session 2  ·  What economic evaluation is, and is not

Q3A study reports the cost of treating cervical cancer by stage, with no comparison of options. What is it?

  1. A full economic evaluation
  2. A cost description
  3. A cost-effectiveness analysis
  4. A cost-minimisation analysis

Costs only, no alternatives compared: Drummond's cost description box. Useful, but it cannot say what to fund.

Q4A paper titled "Cost-effectiveness of X versus Y" compares only the costs of X and Y. What is it really?

  1. A cost-effectiveness analysis, as titled
  2. A cost-utility analysis
  3. A cost analysis, a partial evaluation
  4. A cost-benefit analysis

Two alternatives, costs only. A word in the title is not a method. One paper in the pack does exactly this.

Q5Cost-minimisation analysis is legitimate only when:

  1. the new option is the cheaper one
  2. equivalent outcomes have been demonstrated, not assumed
  3. all costs are in rupees
  4. a decision model was used

If outcomes differ, ignoring them is not minimising cost; it is ignoring the question.

Session 3  ·  Framing the question

Q6The Indian Reference Case asks for which base-case perspective?

  1. Provider
  2. Payer only
  3. Abridged societal
  4. Full societal, including productivity

Payer and patient direct costs, including out-of-pocket. Payer perspective reported separately.

Q7The comparator in an Indian evaluation should normally be:

  1. doing nothing
  2. current practice
  3. the best practice anywhere in the world
  4. the cheapest option available

Do-nothing is right only when it is genuinely what happens now. A weak comparator flatters every result downstream.

Q8HPV vaccination of 14-year-olds evaluated over a one-year horizon would look:

  1. cost-effective
  2. like a waste of money
  3. unaffected by the horizon
  4. cost-saving

All the cost falls in year one; the cancers prevented come decades later. The horizon decides the answer.

Session 4  ·  Costing your own service

Q9A ₹5,00,000 machine will last ten years. What should one year of a costing study carry?

  1. ₹5,00,000
  2. ₹50,000
  3. about ₹58,600, annuitised at 3%
  4. Nothing: it has already been bought

5,00,000 ÷ 8.530. Straight-line ₹50,000 ignores discounting; zero ignores opportunity cost.

Q10A PM-JAY package rate is best described as:

  1. the cost of the procedure
  2. a price
  3. an out-of-pocket payment
  4. an ICER

What the payer pays. In 2018 about 42% of packages were priced below half their measured cost.

Q11Costing a service top down, dividing the department's spend by its patients, tends to miss:

  1. overheads
  2. variation between patients
  3. salaries
  4. the department's budget

Top down sees every rupee but makes every patient average. Bottom up sees variation but misses what nobody counted.

Session 5  ·  Outcomes, discounting and the ICER

Q12Two years lived at a utility of 0.5 are worth how many QALYs?

  1. 0.5
  2. 1
  3. 2
  4. 2.5

0.5 × 2. The QALY's power and its controversy in one line.

Q13A programme costs ₹3,400 more than the comparator and gains 0.13 QALYs. What is the ICER?

  1. ₹204 per QALY
  2. ₹26,154 per QALY
  3. ₹4,420 per QALY
  4. ₹442 per QALY

3,400 ÷ 0.13. ₹204 is the average ratio, total cost over total QALYs, meaningless for a decision.

Q14India's discount rate for costs and outcomes is:

  1. 0%
  2. 3%
  3. 5%
  4. 10%

3%, varied from 0 to 5% in sensitivity analysis, with undiscounted results also reported.

Session 6  ·  Thresholds, opportunity cost and budget impact

Q15The rule that cost per DALY below one to three times GDP per capita is cost-effective is:

  1. India's official threshold
  2. WHO's current recommendation
  3. a heuristic that its own authors have argued should be retired
  4. derived from India's health budget

Demand-side, not linked to any budget. Two papers in the pack still use it; ask whether it changes their verdict.

Q16India's recent estimate of willingness to pay per QALY is closest to:

  1. ₹54,881
  2. ₹2,12,307
  3. ₹20,000
  4. ₹10,00,000

Chugh et al. 2026. ₹54,881 is the ICER for VIA screening from Session 2, a distractor on purpose.

Q17An intervention is cost-effective. Which question does that leave unanswered?

  1. Is it good value per QALY?
  2. Can we afford it for everyone this year?
  3. Is its ICER below the threshold?
  4. Does it produce health?

Budget impact is a separate test. Cost-effective is not the same as funded.

Session 7  ·  Modelling in half an hour

Q18Drug A costs ₹10,000 and cures 80%. Those not cured need second-line treatment costing ₹20,000. Expected cost per patient?

  1. ₹10,000
  2. ₹12,000
  3. ₹14,000
  4. ₹30,000

10,000 + 0.2 × 20,000. Roll back: multiply along each path, then add.

Q19The Markov assumption means:

  1. nobody can move back to a healthier state
  2. the next move depends only on the current state, not on history
  3. every cycle is one year
  4. every patient has the same costs

Memoryless. If time since surgery changes the risk, the model must be built to remember it.

Session 8  ·  Handling uncertainty

Q20Probabilistic sensitivity analysis deals with:

  1. parameter uncertainty
  2. structural uncertainty
  3. heterogeneity between patients
  4. all three

Only parameters. If the model is built wrong, all ten thousand simulations are wrong together.

Q21Which distribution suits a cost parameter in a PSA?

  1. Normal
  2. Gamma or log-normal
  3. Beta
  4. It does not matter

Costs cannot be negative and are right-skewed. A normal distribution lets the model draw a negative cost.

Q22"An 80% probability of being cost-effective" means:

  1. 80% of patients will benefit
  2. the intervention works in 80% of trials
  3. in 80% of simulations it was good value at that threshold, given the model
  4. we can be sure it is cost-effective

And a one-in-five chance of being wrong. 'Probably', not 'yes'.

Close  ·  Close

Q23How confident are you now that you could appraise a published economic evaluation in your field? opinion

  1. Not at all
  2. A little
  3. Fairly
  4. Very

Compare with the Session 1 baseline. Opinion; no right answer.