← Read Edition 74: Stop asking what healthcare costs

Interactive tool · Healthcare Pulse

The Value Lens

Nine ways to ask whether a health intervention is worth it. Enter two options and every lens judges them at once. They will not agree, and where they disagree is the actual decision.

9 evaluation lensesChange one input, watch the verdict moveIllustrative, not a pricing tool
Six words worth knowing before you start
OutcomeWhatever unit of health you are counting: an admission avoided, a life-year gained, a case detected. You choose it.
ComparatorWhat you are measuring against. Change it and every number here changes with it.
ThresholdThe most you would pay for one extra unit of outcome. India has not set an official figure, so this is your judgement.
ICERIncremental cost-effectiveness ratio. Extra cost divided by extra benefit, read as a cost per unit of outcome.
QALYQuality-adjusted life year. One year in full health. Half a year in full health, or a year at half quality, both count as 0.5.
Budget impactWhat adoption costs in total across everyone eligible, which is a different question from whether it is good value.
Verdict
Enter your assumptions
What this is and is not. A structured way to see that the same numbers produce different answers depending on which question you ask. It is not evidence about any real product, and the figures you enter are your own assumptions rather than findings. India has not adopted an official cost-effectiveness threshold, so the willingness-to-pay figure above is a reference point you set, never a rule. For a real decision, the machinery already exists: Health Technology Assessment in India, under the Department of Health Research.

What each lens actually measures

Nine methods, and the differences between them are not academic. Each answers a different question, and each can be used to reach a conclusion the others would reject.

CEACost-effectiveness analysis
ΔCost ÷ ΔEffect = ICER, the cost per extra unit of outcome
What it measuresWhat one extra unit of health costs, using a natural clinical outcome: an admission avoided, a life-year gained, a case detected, a blood pressure controlled.
When to use itWhen the outcome is clinically specific and both options aim at the same thing. This is the workhorse of health technology assessment worldwide.
Where it breaksIt cannot compare a diabetes drug with a cataract surgery, because the outcome units differ. And the verdict depends entirely on a threshold that India has never officially set.
What to watch forA vendor quoting a low ICER without naming the comparator. Change the comparator and the ratio changes completely.
CUACost-utility analysis
ΔCost ÷ ΔQALY, the cost per extra healthy year
What it measuresThe same ratio, but the outcome is a quality-adjusted life year, or QALY, which folds length of life and quality of life into one number.
When to use itWhen you need to compare interventions that do completely different things. A QALY is the common currency that makes that possible.
Where it breaksThe quality weights come from surveys of preference and are contested. Whose preferences, measured how, and does a year of life at 0.7 quality really equal 0.7 of a year at full health?
What to watch forQALY figures imported from a British or American study and applied to India unchanged. Preference weights are not portable across health systems.
CBACost-benefit analysis
Monetised benefit − Cost
What it measuresEverything converted to rupees, including the health outcome, then subtracted. A positive number means the benefit exceeds the cost.
When to use itFor broad programmes where benefits fall outside health: productivity, education, environment. Also when a finance ministry needs one comparable number across sectors.
Where it breaksIt requires putting a rupee value on a life or a year of health. That is doable, and it is a political act rather than a technical one.
What to watch forThe monetisation assumption buried in an annexe. That single number often decides the whole result.
CMACost-minimisation analysis
Lower total cost wins
What it measuresWhen two options genuinely produce the same outcome, the analysis collapses to a single question: which costs less?
When to use itOnly when there is credible evidence that outcomes are meaningfully equivalent. Generic versus branded of the same molecule is the classic valid case.
Where it breaksThe equivalence claim does all the work. If outcomes differ even slightly, this lens hides the difference rather than weighing it.
What to watch forThe commonest misuse in the field. Someone applies it to options that are not equivalent, because it produces the tidiest answer.
BIABudget impact analysis
ΔCost × eligible population × uptake
What it measuresWhat adoption actually costs the payer over a defined period, at realistic uptake, for the whole eligible population.
When to use itAlways, alongside a cost-effectiveness result. It answers a different question and is the one that stops programmes.
Where it breaksIt says nothing about whether the intervention is worth doing. A cheap and useless intervention passes this test easily.
What to watch forUptake assumed at 100 per cent, or an eligible population defined narrowly to keep the total small.
ROIReturn on investment
(Savings − Spend) ÷ Spend
What it measuresFinancial return on the money put in, usually as a percentage or a payback period. Counts only cash, never health.
When to use itFor an internal business case, a procurement committee, or a services proposal. It is the language those rooms already speak.
Where it breaksIt ignores health outcomes entirely. An intervention can show strong ROI while producing no clinical benefit at all.
What to watch forSavings claimed from avoided events that were never going to happen, or counted in a budget line nobody actually controls.
COICost of illness
Direct + indirect + productivity cost
What it measuresWhat a disease costs the system and society as it stands today, before anyone intervenes.
When to use itTo size a problem and make the case for attention. It is a burden measure, useful for prioritisation and advocacy.
Where it breaksIt tells you nothing about whether any specific intervention is good value. A large burden does not mean a good solution exists.
What to watch forA large cost-of-illness figure used as though it were evidence that a particular product is worth funding.
CCACost-consequence analysis
Every cost and outcome, listed separately
What it measuresAll the costs and all the outcomes side by side, deliberately not combined into a single ratio.
When to use itWhen the trade-offs are genuine value judgements and collapsing them into one number would hide the choice being made.
Where it breaksIt gives no verdict. Someone still has to weigh the rows, which means the decision stays with people rather than arithmetic.
What to watch forNothing much. Its weakness is honesty, which is why it is the least used and often the most appropriate.
EFFEfficiency analysis
Output ÷ input, benchmarked
What it measuresWhether the same result could come from fewer resources, and whether the resources are in the right place at all. Technical efficiency asks the first, allocative the second.
When to use itWhen capacity already exists and is underused. Data envelopment analysis benchmarks units against the best performers rather than an abstract standard.
Where it breaksBenchmarking against peers assumes the best performer is actually good. If the whole sector is inefficient, the benchmark is too low.
What to watch forEfficiency gains claimed from a headcount reduction that simply shifts the work somewhere it is not measured.
The single most useful habit. When someone presents an economic case, ask which lens produced the number, what the comparator was, and over what time horizon. Those three questions expose most of the weak arguments in this field, and none of them requires you to know the methods in detail.