Catastrophic Health Expenditure

Catastrophic Health Expenditure Thresholds, Headcount and Overshoot

Catastrophic health expenditure is estimated from household surveys by dividing each household's out-of-pocket payments for care by a measure of its resources and counting the households above a cut-off. The usual cut-offs are 10% or 25% of total consumption under the budget share approach and 40% under capacity-to-pay approaches, which first set aside spending on basic needs. The measure is a main test of the financial protection side of universal health coverage and was the original form of Sustainable Development Goal (SDG) indicator 3.8.2.

Why a large health payment counts as a welfare loss

O'Donnell and colleagues, in a World Bank guide to health equity analysis, start from the risk of large medical bills faced by households without full insurance. If the bills are large relative to household resources, the disruption to material living standards may be considered catastrophic. Paying a large share of the budget for care comes at the expense of other consumption, either now, if care is financed by cutting current consumption, or later, if it is financed from savings, the sale of assets or credit.

Xu and colleagues, in a 59-country analysis of household surveys, identified health services that require payment, low capacity to pay, and the lack of prepayment or health insurance as the three preconditions for catastrophic payments. The measure therefore belongs to health system financing. Risk pooling and prepayment spread the cost of illness, and catastrophic spending records where they fail.

Two ways to define a household's ability to pay

Every method uses out-of-pocket payments as the numerator. Cylus, Thomson and Evetovits define these as formal and informal payments at the point of using any health good or service, net of third-party reimbursement; premiums, contributions and taxes are left out because they are the prepaid, pooled sources that protect against financial risk. The methods differ mainly in the denominator.

The budget share approach divides payments by total household consumption (or income), assuming the whole budget is available for care. Capacity-to-pay approaches first deduct an amount for basic needs, which households must meet before paying for care. Cylus and colleagues compare four methods, set out in the table.

Consumption is generally preferred to income as the denominator. Cylus and colleagues report that research favours it as a better welfare indicator that is easier to measure, and the SDG 3.8.2 metadata recommend it whenever both are available.

The headcount, overshoot and mean positive overshoot

O'Donnell and colleagues define the measures by analogy with poverty measurement. Incidence is a headcount of households above the threshold. Intensity is how far they exceed it.

$$E_i = \mathbf{1}\left(\frac{T_i}{x_i} > z\right), \qquad H = \frac{1}{N}\sum_{i=1}^{N} E_i$$

where $T_i$ is the out-of-pocket payment of household $i$, $x_i$ is its total expenditure (or $x_i - f(x_i)$ under a capacity-to-pay approach, with $f(x_i)$ its spending on basic needs), $z$ is the threshold, $E_i$ equals 1 above the threshold and 0 otherwise, $N$ is the number of households and $H$ is the headcount.

$$O_i = E_i\left(\frac{T_i}{x_i} - z\right), \qquad O = \frac{1}{N}\sum_{i=1}^{N} O_i, \qquad MPO = \frac{O}{H}$$

where $O_i$ is the overshoot of household $i$, $O$ is the mean overshoot across all households and $MPO$ is the mean positive overshoot, defined only when $H > 0$. Because $O = H \times MPO$, the overshoot is incidence multiplied by intensity. Complex surveys need weights, strata and sampling units. SDG reporting counts people, so each household is also weighted by its number of members.

Worked example with four households and two definitions

The households and every number below are illustrative. Four households of equal size and sampling weight report annual consumption and out-of-pocket payments. For the capacity-to-pay method a standard basic-needs amount of £6,000 is deducted from every household, as in the normative method; a household below that amount has no capacity to pay, and any payment counts.

1. Budget share at 10%. Households C and D are above 10%, so $H = 2/4 = 0.50$. Their overshoots are 0.2167 minus 0.10, or 0.1167, and 0.125 minus 0.10, or 0.025.

$$O = \frac{0 + 0 + 0.1167 + 0.025}{4} = 0.0354, \qquad MPO = \frac{0.0354}{0.50} = 0.0708$$

where the numerator of $O$ lists the overshoots of households A to D. The counted households exceed the threshold by about 7.1 percentage points on average, so they spend about 17.1% of their budgets.

2. Budget share at 25%. No household is above 25%, so $H = 0$ and the mean positive overshoot is undefined.

3. Capacity to pay at 40%. Household C spends £2,600 of a £6,000 capacity to pay, or 43.3%, and is counted. Household A is below the basic-needs amount and pays £300, so it is counted. B and D are below 40%, so again $H = 0.50$.

The headcount is the same at 10% of the budget and 40% of capacity to pay, but the households differ. The budget share flags the richest household and misses the poorest, whose £300 comes out of money needed for basic needs.

Choices that change the estimate

O'Donnell and colleagues describe the threshold as a matter of judgement and advise presenting results for a range of values. The threshold also depends on the denominator, since 10% of total expenditure might be considered catastrophic, but 10% of nondiscretionary expenditure probably would not. They report 10% as the most common threshold with total expenditure, and 40% of capacity to pay as the level used by WHO researchers.

The denominator carries a value judgement about savings and borrowing. Two households with equal income and payments have the same payment-to-income ratio, but the ratio to total expenditure is larger for the one without savings that cuts current consumption. Under capacity-to-pay methods, the definition of basic needs matters as much as the threshold. Cylus and colleagues show that deducting actual food spending gives households with similar budgets very different effective thresholds, since a household that cuts food to pay for care appears to have more capacity to pay. Other choices include household or person as the unit, rescaling recall periods to one period, and payments reimbursed after the survey recall period, which the SDG 3.8.2 metadata flag for schemes with retrospective reimbursement.

Who gets counted and the equity problem with a fixed budget share

Cylus and colleagues applied the four methods to household budget surveys from 14 European countries. With the budget share method, catastrophic spending was largely experienced by richer households, whereas the normative food, housing and utilities method found it concentrated in the poorest quintile in every country. The budget share applies one threshold to all households, while capacity-to-pay methods create effective thresholds that vary across households. The authors liken the budget share to a flat income tax and conclude that the SDG budget share overestimates hardship among rich households and underestimates it among poor ones.

Wagstaff and van Doorslaer developed indices that capture whether catastrophic payments fall disproportionately on poor households, and O'Donnell and colleagues weight the headcount by the concentration index in the following way.

$$H^W = H\left(1 - C_E\right)$$

where $H^W$ is the rank-weighted headcount and $C_E$ is the concentration index of $E_i$ across the income distribution. The implied weights fall linearly from 2 for the poorest household to 0 for the richest, so a negative $C_E$, with the worse-off more likely to exceed the threshold, makes $H^W$ larger than $H$.

From catastrophic spending to financial hardship in SDG 3.8.2

The original SDG indicator 3.8.2 was the proportion of the population whose out-of-pocket spending exceeded 10% and 25% of total household consumption or income. On that basis Wagstaff and colleagues estimated global incidence at the 10% threshold as 9.7% in 2000, 11.4% in 2005 and 11.7% in 2010, with 808 million people affected in 2010.

In 2025 the UN Statistical Commission approved a revision proposed by WHO and the World Bank. The revised indicator counts people with positive out-of-pocket spending above 40% of the household discretionary budget, which is consumption or income minus the societal poverty line; below that line the discretionary budget is negative, so any payment counts. The 2025 global monitoring report states that the budget share version tended to concentrate catastrophic spending among the better-off, that the original and revised trends are not directly comparable, and that the large but non-impoverishing component of the new indicator has deteriorated in line with the original. The new measure combines large spending with impoverishing health expenditure without double counting.

Using the measure in policy analysis and economic evaluation

Catastrophic spending rates are used to compare countries and find groups left exposed by benefit design and cost sharing. Wagstaff and colleagues found that incidence correlated negatively with the share of health spending channelled through social security funds and other government agencies. They concluded that the share of the population supposed to be covered is a poor indicator of financial protection, and that raising the share of GDP spent on health is not enough without raising the prepaid share of health spending.

In economic evaluation, extended cost-effectiveness analysis adds financial risk protection to health gains and costs. Verguet, Kim and Jamison list cases of catastrophic health costs averted as one metric, in which individuals whose direct medical and non-medical (mainly transport) costs exceed a threshold share of income are counted before and after a policy such as public financing of treatment.

Boundaries with nearby measures

Impoverishing health expenditure asks whether payments push a household below a poverty line or further below it; catastrophic spending asks whether payments are large relative to the budget. A household just above the poverty line can be impoverished by a payment below a 10% budget share threshold, and a wealthy household can cross a threshold while staying well above the poverty line. The out-of-pocket share of total health expenditure is a national ratio, not a household measure; in the Republic of Moldova, where it was 41% in 2013, Cylus and colleagues found the budget share at 25%, actual food and partial normative food methods each giving catastrophic spending in fewer than 5% of households.

What a catastrophic spending rate cannot show

The measure sees only households that pay. O'Donnell and colleagues note that it ignores households that cannot meet the cost and forgo treatment, which probably suffer a greater welfare loss, and the SDG metadata note that people facing access barriers report no spending, which lowers the rate. A low rate therefore needs to be read alongside service coverage and unmet need.

Illness also causes lost earnings, which the measure omits, and cross-section data cannot show whether payments were financed by cutting consumption or by savings, assets and credit. A single cut-off treats households just below it as unaffected; the SDG metadata report that the threshold affects the level of indicator 3.8.2 but not its trend. Survey data are self-reported and rarely record health care needs. A catastrophic payment headcount is not the number of households made poor, and estimates with different denominators, thresholds or units cannot be compared directly.

Sources

• Cylus J, Thomson S, Evetovits T. Catastrophic health spending in Europe: equity and policy implications of different calculation methods. Bulletin of the World Health Organization. 2018;96(9):599-609. https://doi.org/10.2471/BLT.18.209031
• O'Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Analyzing health equity using household survey data: a guide to techniques and their implementation. Washington, DC: World Bank; 2008. Chapter 18, Catastrophic payments for health care. https://openknowledge.worldbank.org/handle/10986/6896
• United Nations Statistics Division. SDG indicator metadata: indicator 3.8.2, proportion of population with positive out-of-pocket household expenditures on health exceeding 40 per cent of household discretionary budget. New York: United Nations; last updated 27 March 2026. https://unstats.un.org/sdgs/metadata/files/Metadata-03-08-02.pdf
• Verguet S, Kim JJ, Jamison DT. Extended cost-effectiveness analysis for health policy assessment: a tutorial. PharmacoEconomics. 2016;34(9):913-923. https://doi.org/10.1007/s40273-016-0414-z
• Wagstaff A, Flores G, Hsu J, Smitz MF, Chepynoga K, Buisman LR, van Wilgenburg K, Eozenou P. Progress on catastrophic health spending in 133 countries: a retrospective observational study. The Lancet Global Health. 2018;6(2):e169-e179. https://doi.org/10.1016/S2214-109X(17)30429-1
• Wagstaff A, van Doorslaer E. Catastrophe and impoverishment in paying for health care: with applications to Vietnam 1993-1998. Health Economics. 2003;12(11):921-934. https://doi.org/10.1002/hec.776
• World Health Organization, International Bank for Reconstruction and Development / The World Bank. Tracking universal health coverage: 2025 global monitoring report. Geneva: WHO and World Bank; 2025. https://www.who.int/teams/health-financing-and-economics/global-monitoring-report/2025
• Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJL. Household catastrophic health expenditure: a multicountry analysis. The Lancet. 2003;362(9378):111-117. https://doi.org/10.1016/S0140-6736(03)13861-5