Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Out-of-Pocket Maximums and Catastrophic Medical Cost Exposure

Abstract. An out-of-pocket maximum is a contractual or statutory limit on specified patient cost sharing within a defined period. Its financial effect depends on the expenses eligible for accumulation, the treatment of network and coverage exclusions, individual and family allocation rules, and enforcement at the point of claim. Catastrophic medical cost exposure is a wider household phenomenon: illness, disability, a medical episode, or an accumulation of care needs can generate expenses, foregone consumption, debt, and financial distress even when a health plan has a stated maximum. This monograph examines the relationship through legal and contractual analysis, health economics, empirical studies of cost-sharing incentives, and international household financial-protection indicators. It distinguishes eligible insured spending from total household health payments, the nominal cap from the effective liability boundary, and measured catastrophic expenditure from unmet need. A transparent piecewise liability model and six explicitly hypothetical cases illustrate how deductibles, coinsurance, network exclusions, medical billing protections, income, and liquidity shape exposure. Comparative sections analyze U.S. regulated insurance, Medicare, German statutory insurance, Swiss compulsory insurance, and the British NHS. The study concludes that a cap is meaningful only within its legally enforceable benefit perimeter and that household financial protection must be assessed jointly through benefits, residual liability, resources, access, and time. Its original contribution is an integrated source-critical and contract-level interpretation, without claiming a new theory, an original empirical dataset, or a causal effect outside the cited studies.

Keywords: Out-of-pocket maximum; catastrophic health expenditure; health insurance; deductible; coinsurance; essential health benefits; balance billing; provider networks; household financial protection; health financing; nonlinear insurance; affordability; unmet medical need; cost-sharing regulation; health expenditure measurement.

Introduction

Health insurance joins two distinct financial questions. The insurer or public scheme specifies which claims it will pay, while the household seeks to know how much illness can ultimately cost. The questions overlap, but their accounting units are different. An annual out-of-pocket maximum often protects covered in-network cost sharing; it may leave premiums, uncovered care, certain out-of-network services, balance billing, and service-specific obligations outside its accumulation rule. The HealthCare.gov maximum definition provides an unusually clear public statement of this boundary. A plan with a modest headline limit may still permit a substantial total household cash demand when the boundary is crossed.

The corresponding scientific problem is to identify both the protection supplied by the cap and the residual tail of liability. This problem has legal, actuarial, behavioral, and distributive dimensions. The legal dimension determines which liabilities are enforceable. The actuarial dimension concerns the distribution of covered and excluded costs and the share transferred to a risk pool. The behavioral dimension concerns how a nonlinear schedule changes care-seeking and provider choice at different points in the year. The distributive dimension concerns the relationship between liability and a household's remaining resources. These dimensions interact: a formal ceiling can be generous in dollars and unaffordable for a low-income family, while a system with small statutory copayments can still expose patients to excluded services, delayed care, or the cost of acquiring coverage.

The present inquiry treats the policyholder or insured individual, the household, the provider, and the financing institution as distinct units. Contractual accumulation is generally measured for an insured person or covered family over a benefit period. National catastrophic-spending statistics are constructed from household expenditures and welfare denominators. These concepts should therefore be connected through explicit mappings rather than substituted for one another. The OECD financial-hardship chapter confirms that international comparisons depend on household survey methods, the costs included, and the definition of resources remaining after basic needs.

The method is a structured documentary synthesis of authoritative legal and institutional texts and selected original quantitative research. Statutory rules and official program guidance establish formal obligations in specified jurisdictions and years. Randomized experiments and observational or quasi-experimental studies establish narrower causal or descriptive claims about care use and financial consequences. Mathematical cases demonstrate identities and conditional implications without representing measured population outcomes. The literature review is source-critical and purposive; it is not a systematic review with a preregistered search or pooled effect estimate.

Foundations and Source-Critical History

The object of an expenditure ceiling

A maximum is an upper bound only for an expressly defined variable. Let the variable be eligible member cost sharing: the accumulated deductible, copayment, and coinsurance amounts recognized by a plan during a specified coverage period. A ceiling on that variable can be legally binding while a household's total medical and coverage spending exceeds it. The official health coverage glossary distinguishes an allowed amount from billed charges and specifies that premiums, balance-billed charges, and noncovered health care lie outside the usual out-of-pocket limit. The allowed amount is a contractual or administrative price basis; the provider's billed charge may differ.

This distinction determines the epistemic status of any assertion that a family is protected from catastrophic medical costs. The assertion is accurate for the cost class to which the cap applies, subject to plan compliance and services actually received. It cannot be extended to all costs merely because the plan uses the word maximum. The evidence needed is therefore a benefit specification, a covered-service definition, an accumulation rule, the relevant regulatory text, and ideally claims-level verification that the adjudicated liabilities respect those provisions.

The upper bound must also be separated from an annual limit on insurance benefits. Under U.S. rules, the essential-benefit annual-limit prohibition addresses dollar limits on the amount a plan will pay for essential health benefits, subject to its stated scope and exceptions. A patient out-of-pocket maximum instead limits qualifying patient cost sharing. Confusing the two would invert the direction of the protection: one concerns the insurer's benefit obligation, the other the insured's payment obligation.

From cost-sharing experiments to household protection

The modern economics of medical cost sharing developed around the response of health-service use to the immediate price facing a patient. The original RAND insurance demand study, reported by Manning, Newhouse, Duan, and collaborators (1987), compared randomly assigned contracts that varied coinsurance and bounded patient expenditure. Its results document a substantial response in use and spending to the insured consumer's price. They do not establish that a maximum alone produces financial security across every medical expense category.

Aron-Dine, Einav, and Finkelstein (2013), in their reassessment of the RAND experiment, warn against treating the effects of nonlinear contracts as though they were described adequately by one constant elasticity. A patient who expects to reach a cap may face a different effective annual price from the price displayed on an initial bill. The timing of care, expectations about later illness, and the current position relative to the deductible and maximum all matter. Their methodological reassessment also discusses threats to straightforward interpretation of the original randomized effects, reinforcing the need to distinguish study design from generalized policy conclusions.

There is a parallel history in global health-financing research. Catastrophic payment measures were developed to investigate whether health expenditures consume resources otherwise needed for basic living standards. Their denominators and thresholds are analytical conventions, not contractual limits. The WHO 2025 financial-hardship metadata specifies a revised SDG 3.8.2 construction using out-of-pocket payments relative to a household discretionary budget. The two histories meet in the household accounts: plan rules define one part of actual payment, and financial-protection research assesses the burden of the resulting payment within the budget.

Evidence and jurisdictional boundaries

This study uses several levels of evidence. First, formal U.S. regulatory requirements and official public-program descriptions are read as statements of law or administration only for the plans and dates they cover. Second, ministry and regulator explanations for Germany and Switzerland are evidence of those national program designs; they are not a substitute for applying the current statute to a particular claim. Third, WHO and OECD indicators provide specified comparative measures whose scope is governed by their metadata. Fourth, original research supports empirical claims within the populations and designs examined. Fifth, worked examples provide reproducible conditional arithmetic.

The distinction is consequential for historical interpretation. A 2013 U.S. experiment does not measure the 2026 marketplace maximum. The financial protection created by Medicaid enrollment in Oregon cannot be assumed to equal the effect of reducing a deductible in a commercial plan. A national aggregate out-of-pocket share cannot establish the risk faced by a particular family. Where evidence is limited to a public scholarly abstract or official overview, the monograph uses the supported conclusion rather than inferring additional outcomes or significance.

Annual Caps and Eligible Expenses

Legal and contractual components

In U.S. private insurance subject to the relevant federal provisions, 45 CFR § 156.130 regulates annual cost-sharing limitations and describes how network cost sharing is treated. The provision states that network plans generally need not count cost sharing for out-of-network benefits toward the annual limitation. The legal applicability of the provision, including exceptions and interaction with other federal and state requirements, must be determined for the plan and benefit year. The consumer-facing 2026 Marketplace explanation states an upper bound of $10,600 for individual coverage and $21,200 for family coverage for that plan year. These are regulatory outer limits for the identified category of plan, not the amount that every enrollee necessarily pays or an unconditional household expenditure ceiling.

A deductible is the amount or specified cost-sharing stage the insured must satisfy before the plan pays on particular services. Coinsurance is a proportional contribution after the applicable deductible or coverage threshold; a copayment is a fixed amount for a covered service. An out-of-pocket maximum may aggregate some or all of these according to the contract and law. The amount recognized is ordinarily calculated from covered services under a permitted pricing basis. Payment for a noncovered service cannot automatically be treated as satisfying a covered-service limit. Preventive services may have special no-cost-sharing treatment, and drug benefits can be governed by separate accounting.

The legally relevant record includes the policy or certificate, formulary, network status, Summary of Benefits and Coverage, explanation of benefits, provider billing documents, and applicable program rules. The Summary of Benefits and Coverage requirement is designed to make important benefit terms comparable; a summary should be read alongside the complete governing document for exceptions and accumulation instructions. A single number displayed during shopping is insufficient to reconstruct the payment function.

Formal allocation of eligible liability

Consider a deliberately simplified annual in-network contract with eligible allowed spending G, deductible D, coinsurance rate c after the deductible, and maximum recognized patient cost sharing M. Under the assumptions that all included services use the same benefit schedule, no separate copayments apply, D is no greater than M, and G is nonnegative, the pre-cap patient liability is

p(G)=min⁡(G,D)+cmax⁡(G−D,0),p(G)=\min(G,D)+c\max(G-D,0),

and the capped liability is

C(G)=min⁡{M,p(G)}.C(G)=\min\{M,p(G)\}.

This equation is an accounting exposition of a conventional deductible-plus-coinsurance contract, not a new model of health behavior. It describes the covered allowed-amount component only. The insurer's net payment for these allowed claims is G minus C(G), abstracting from coordination of benefits, third-party payments, risk adjustment, and medical management. When c is strictly positive and M exceeds D, the eligible allowed spending required to reach the ceiling is

G∗=D+M−Dc.G^{*}=D+\frac{M-D}{c}.

The threshold is a function of both the cap and the coinsurance rate. If c equals zero, liability typically stops increasing after the deductible and the nominal maximum may never be reached under this simple contract; if D already equals M, the ceiling can be exhausted by the deductible. The equation should not be applied to claim-specific copayments, tiered drugs, mixed benefit classes, or complex family accumulators without an additional schedule.

Actual benefit processing is sequential. Each allowed claim passes through eligibility, network, coordination, deductible, coinsurance or copayment, patient-liability recognition, and the remaining accumulator. Reversals, retroactive adjustments, and midyear changes can affect the final recognized amount. Thus an end-of-year algebraic equality is a useful audit check, while a claim-by-claim ledger is necessary to establish operational compliance.

Individual and family accumulation

An individual maximum and a family maximum address distinct allocation questions. A family plan can contain a family-wide aggregate and individual embedded limits, depending on its design and regulatory obligations. Consider two family members with uncapped eligible liabilities q1 and q2 and individual eligible ceilings m1 and m2. In a simple family contract with a combined ceiling F, the final combined eligible liability can be represented by min(F, min(q1,m1)+min(q2,m2)). This expression assumes the same accumulation period, mutually compatible benefit rules, and that individual limitations are applied correctly. It is not a statement that every family contract has precisely this formula.

An embedded individual ceiling protects a single member of a family against using all of the family maximum alone; the aggregate family ceiling limits recognized cost sharing across members. Claim timing matters when multiple members cross separate thresholds, especially under per-service copayments or mixed networks. Family composition changes, births, coverage loss, and changes of employment raise additional questions concerning accumulator continuity. An annual statutory limit therefore cannot be interpreted without the participant-level rule.

A concrete design comparison follows. Suppose one member incurs $40,000 of covered allowed spending and a second member incurs $2,000; the nominal family maximum is $12,000 and the applicable individual maximum is $6,000. If the first member's uncapped liability is $9,000 and the second's is $2,000, the simplified embedded structure yields $6,000+$2,000=$8,000, which remains below the family ceiling. Without the individual embedded limit, the first member might face a higher amount subject to applicable law. The outcome is determined by eligibility and plan design, rather than the family maximum in isolation.

Medicines, premiums, and temporal boundaries

Prescription drug cost sharing may be integrated into a plan's general maximum or governed by a different program-specific threshold. Medicare Part D supplies a clear example of a distinct statutory structure: the CMS 2026 Part D instructions establish a $2,100 annual out-of-pocket threshold for covered Part D drugs in 2026. Some payments made on a beneficiary's behalf count under Part D's defined incurred-cost rules. It would be incorrect to add a Part D threshold to a Medicare Advantage medical out-of-pocket maximum as if the two were one accumulator, or to interpret the prescription threshold as a ceiling on all medical expenses.

Premiums normally purchase the insurance entitlement and are economically relevant even though they do not count toward the standard medical out-of-pocket maximum. This creates a distinction between cost-sharing protection and total household financing burden. For a family purchasing an individual-market policy, annual fixed premiums can be substantial even in a year with no claims. Conversely, a generous premium subsidy may reduce fixed expense without changing the plan's cost-sharing maximum. A full affordability analysis therefore treats premiums and eligible patient payments separately before examining total resources.

Annual ceilings have a time dimension. They typically reset at a plan-year or calendar-year boundary, which can expose a family to two deductible and maximum cycles during a continuous episode spanning two periods. A single illness may also generate uncovered rehabilitation, caregiving, travel, or income loss that does not enter a medical claims accumulator. The legal cap belongs to the covered-benefit period; the economic shock can persist across several years. Longitudinal assessment is therefore essential for evaluating recurring illness and complex chronic treatment.

Out-of-Network and Uncovered Costs

The network boundary

Provider networks change the price and sometimes the eligibility of care. A plan may pay a negotiated in-network allowed amount but use a different payment schedule, higher cost sharing, separate accumulators, or no benefit at all for particular out-of-network encounters. The general network exception in federal regulation establishes why the phrase maximum out-of-pocket has to be read together with the phrase for which services. In-network status must be verified for the particular provider, location, facility, and episode, rather than assumed from the appearance of a hospital or a referral.

There are at least three different financial amounts in a medical bill: the provider's charge, the allowed or recognized amount used by the financing arrangement, and the member responsibility after coverage rules. The difference between the first two can be written off by contract, paid by a supplementary scheme, or become a patient balance-billing liability where permitted. These alternatives are legally different. A nominal $8,500 in-network maximum does not alone disclose the patient's liability if a noncovered provider bills $15,000. The appropriate analysis begins with whether the service is covered, whether an applicable balance-billing prohibition operates, and what amounts count toward the ceiling.

Coverage exclusions form a related boundary. An excluded drug, treatment, or provider category can create spending outside the covered-benefit formula even when the service is medically important to the patient. A denied claim also requires distinguishing a legitimate exclusion, medical-necessity dispute, authorization failure, coding error, or administrative mistake. The same cash payment may be initially outside the accumulator and later reclassified on appeal. Scientific studies of protection therefore need final claim disposition and the timing of the financial burden.

Surprise-billing protections and their scope

The United States enacted specific protections against certain unexpected out-of-network charges through the No Surprises Act. The official CMS insured-patient rights guidance identifies protected emergency services and certain other situations in which a patient generally faces in-network cost-sharing standards despite an out-of-network provider. It also identifies limitations: ground ambulance services are generally outside federal No Surprises Act billing protections, subject to other applicable law, and particular notice-and-consent circumstances can change coverage of post-stabilization services.

This is a decisive qualification to a broad claim that every out-of-network expense is uncapped. Where the law requires in-network-equivalent cost sharing and accumulation, the ordinary out-of-network exclusion cannot simply be imposed. Conversely, the existence of surprise-billing protections does not transform all elective out-of-network care into protected care. The correct order is to identify the service, legal protection, plan type, date, and applicable jurisdiction; determine the protected cost-sharing calculation; and then establish the accumulator treatment. For comparative studies, the legal classification is part of the measured exposure rather than background narrative.

A transparent catastrophic episode

Consider one hypothetical insured person within one benefit period. The deductible is $2,000, post-deductible coinsurance is 20%, and the covered in-network maximum is $8,500. The person receives $75,000 of eligible in-network allowed care. Before the cap, the member's charge would be $2,000 plus 20% of $73,000, or $16,600. The eligible member payment therefore stops at $8,500. For these claims, the plan pays $66,500. This is a numerical illustration of the contract formula, not an estimate of typical utilization or a statement about any real insurer.

Now assume the same person separately incurs $7,200 of legally permissible out-of-network liability excluded from the accumulator and $4,300 for services outside the policy. Annual premiums are $5,600. The person's payment for care is $8,500 + $7,200 + $4,300 = $20,000. Including premiums, cash committed to coverage and care becomes $25,600. The $8,500 figure is a valid maximum for the eligible component while the total care-related payment is more than twice that amount. The scenario deliberately assumes the $7,200 is outside applicable surprise-billing protection; if protection applies, the legal and arithmetic classification must change.

The illustration also identifies the limits of the concept of a catastrophic ceiling. Suppose the out-of-network and excluded categories have no effective finite contractual upper bound. Then a finite M bounds C(G) but need not bound C(G)+N+U, where N denotes lawful residual network charges and U denotes uncovered spending. For fixed M, the latter expression increases without a global finite upper bound as N or U increases. This conclusion is elementary contract algebra. It says nothing about the probability of arbitrarily large bills, which must be estimated separately from actual claims, patient payment, and coverage data.

Dynamic incentives under nonlinear contracts

A consumer's marginal cost of a covered service generally changes with the accumulator. Before the deductible the marginal share may approach 100% of allowed charges; after the deductible it can fall to coinsurance c; after the eligible maximum it may fall to zero. These segments create a nonlinear budget set. Observed behavior is shaped by the price at the point of use, the expected end-of-year marginal price, liquidity, information, and health status. The Medicare Part D nonlinear-contract study by Einav, Finkelstein, and Schrimpf (2015) analyzes prescription responses to changing price regions using administrative data and an explicit dynamic framework. Its empirical setting is historical Part D design; later statutory redesign must not be read back into its treatment environment.

Brot-Goldberg, Chandra, Handel, and Kolstad (2017) exploit a high-deductible employer-plan switch. The reported firm-wide spending reduction of roughly 11.8%–13.8% arose from lower quantities rather than demonstrated price shopping; reductions affected potentially useful as well as less valuable care. This finding cautions against equating higher cost sharing with efficient consumer selection. A patient approaching the maximum may also make care decisions on the basis of the immediately quoted price despite anticipating a lower effective annual price. Providers' billing opacity and the uncertainty of disease trajectories make this behavior understandable within the observed constraints.

An out-of-pocket cap thus has two roles that must be analyzed together. It transfers the highest eligible expense to the risk pool and changes the patient's marginal price when the ceiling is reached. The first role addresses insurance against severity; the second affects utilization incentives. Actual welfare implications depend on health benefit, forgone needed care, insurer financing, possible premium responses, and who bears the marginal payment. No single elasticity establishes the preferred cap or deductible for every population. A source-grounded evaluation must place these incentive results next to clinical effects and distributional protection rather than treating lower spending as sufficient evidence of success.

Household Financial Protection

Financial catastrophe as a household measure

Household financial catastrophe is defined relative to resources, not merely to an insurance contract. The WHO revised SDG indicator measures financial hardship when positive out-of-pocket health spending exceeds 40% of a defined discretionary household budget. Its detailed 2025 indicator description explains that the discretionary budget subtracts a societal poverty line from total household consumption or income. The revised methodology was adopted in March 2025. This measure includes impoverishing spending as an explicitly characterized component and treats households with little or no discretionary budget according to the published methodological conventions.

The OECD's 2025 financial-hardship methodology also uses a 40% threshold but defines resources available to pay for care through a standardized basic-needs subtraction based on food, rent, and utilities. The numerical threshold therefore does not make the two indicators automatically interchangeable. Different poverty lines, consumption aggregates, survey years, household weights, currency conversion, and service classifications can assign the same observed payment to different hardship states. Every cross-country graph should declare the indicator version, denominator, and year.

For a transparent illustrative comparison, suppose a household has annual consumption of $72,000 and a prescribed illustrative basic-needs benchmark of $42,000. Available resources are $30,000. If household direct out-of-pocket care payments are the $20,000 from the earlier example, the ratio is 66.7%. It exceeds a 40% capacity-to-pay threshold. The calculation is a generic scenario, not an official WHO or OECD estimate: the agency-specific societal poverty line, survey definitions, and any adjustments may generate a different denominator. The $5,600 insurance premium was excluded from the direct payment numerator but remains relevant to the family's cash budget. If the family has $10,000 of available resources instead, the same $8,500 contract maximum alone would consume 85% of those resources if reached.

Distribution, liquidity, and affordability

The distribution of liabilities matters more than an average when evaluating catastrophic exposure. Let Y be actual household out-of-pocket health payment, W a consistent household income or consumption measure, and B the basic-needs benchmark defined by the chosen indicator. One transparent exceedance statistic is Pr[Y > 0.4(W−B)] for positive discretionary resources, with the published indicator's conventions applied to zero or negative values. Its purpose is to measure the share crossing a financial-hardship threshold. It must not be confused with Pr[Y > M], because M caps a different variable. The expected amount E[Y] and the tail of Y can move in opposite directions if reduced covered cost sharing is offset by uninsured service needs, premiums, or unmet care.

A cap of $7,000 could be simultaneously more protective than one of $10,000 for a covered high-cost episode and unaffordable for a household with $9,000 remaining after basic consumption. Ability to prepay also matters. The household may owe a substantial deductible early in the year, even when a subsequent episode will eventually exhaust the maximum. A legally guaranteed reimbursement can arrive after a debt has already caused distress. These timing effects require a cash-flow view rather than only annual benefit accounting.

The Commonwealth Fund 2024 coverage survey uses an underinsurance definition based on actual out-of-pocket payments relative to income, adjusted for lower incomes, and deductible size. It explicitly notes that its underinsurance classification does not incorporate all potential high costs associated with maximums, copayments, or uncovered services because those features were not asked. This is methodologically valuable: an operational affordability index may omit exactly the contract feature under review. The index should be described accurately and supplemented with benefit detail, rather than treated as a complete cap-exposure dataset.

The global scale of financial hardship also matters. The joint 2025 WHO monitoring report presents revised series and reports approximately 2.1 billion people experiencing financial hardship due to out-of-pocket health spending in 2022. This is a worldwide estimate under the report's methodology and reference year, not a count of people who exceeded an insurance out-of-pocket maximum in 2026. The distinction protects historical comparability and avoids converting a household-survey concept into a contractual claims statistic.

Causal evidence and its limits

The Oregon Medicaid randomized evaluation illustrates the financial effects of coverage expansion in a particular population. Baicker and colleagues (2013), in the Oregon clinical and financial evaluation, used lottery-based assignment and reported that Medicaid coverage nearly eliminated catastrophic out-of-pocket medical expenditures within the study's measured outcome and period. It also increased service use and reduced reported financial strain. The published two-year results did not find statistically significant changes in several physical-health outcomes such as mean glycated hemoglobin or the measured prevalence of hypertension and hypercholesterolemia. These statements must be kept together to preserve the actual evidentiary pattern.

The Oregon intervention was access to Medicaid coverage, with its associated eligibility, benefits, and copayments. It was not random assignment to two otherwise identical commercial plans with different maximums. The causal claim is therefore about the observed expansion and its studied outcomes, subject to the instrument and sample, and does not directly estimate the marginal effect of lowering an annual commercial out-of-pocket cap by $1,000. A comparison of cap policies requires data capable of isolating that variation or an explicit structural assumption.

Wharam and colleagues (2013), examining high-deductible emergency-care patterns, report concerning reductions in high-severity emergency care among lower-socioeconomic-status enrollees following high-deductible exposure. This observational study addresses a different margin from randomized Medicaid expansion and cannot be generalized as a universal effect for every deductible design. Alongside the RAND and firm-switch results, it supports a narrower conclusion: price exposure can alter the quantity and kind of care obtained, and that response can interact with income and clinical severity.

A measured low incidence of catastrophic payments can arise in two radically different environments. One is an environment of accessible covered care with effective protection. The other is an environment where patients avoid needed care because they cannot pay. A household survey observing little spending may record both as low financial outlay; additional measures of unmet need, foregone treatment, health outcomes, and access are necessary to evaluate protection. This is a methodological implication of combining payment-based indicators with utilization research, rather than a claim that any particular country's indicator is invalid.

Stress-testing alternative cap designs

A cap comparison should vary at least five inputs: eligible covered spending; the share and legal status of charges outside the cap; household discretionary resources; time within and across coverage periods; and the price of acquiring coverage. For the earlier single-member example, reducing M from $8,500 to $6,000 decreases eligible member liability by $2,500 if the original uncapped liability exceeds both limits. If the uncovered and outside-network costs remain $11,500, the family's direct health payments fall from $20,000 to $17,500. The change is real but smaller than the remaining burden. If the person's eligible cost sharing never reaches $6,000, changing M alone does not alter that year's realized payment.

The distributional effect also depends on illness prevalence. People who incur little covered cost receive no direct claims payment from a lower maximum in that year; people whose illness crosses the threshold receive greater risk-pool support. A financing institution might adjust premiums, taxes, benefits, or contributions in response, but the size and direction of equilibrium incidence require data and a defined financing setting. The worked cases hold these other terms constant so that the identity being demonstrated is transparent.

A second stress test concerns the coverage boundary. Keeping M unchanged while converting a legally permissible outside-network $7,200 payment into covered in-network-equivalent liability can materially lower residual cash demand, depending on deductible credit and the cap already reached. In the episode above, if the person has exhausted the recognized $8,500 cap, fully protected additional allowed in-network care would generally create no additional recognized cost sharing in that same covered period. This result is conditional on coverage, cost-sharing law, and claims adjudication. Network and billing reform can therefore be at least as important as the numerical headline maximum for an affected episode.

Data, Measurement, and Validation

Administrative and household evidence

A complete empirical study requires two linked but often separately available information systems. Administrative enrollment and claims data can reveal benefit design, allowed charges, covered services, accumulators, paid claims, member liabilities, service dates, and providers' network classification. Household survey data can reveal direct payments, income or consumption, household composition, debt, borrowing, asset sales, and unmet care. The administrative dataset describes the financing contract and recognized claims; the household survey captures a broader financial experience, although with recall and reporting limitations.

For administrative analysis, the minimum record is plan identifier, benefit year, service category, claim date, adjudication date, billed and allowed charge, in-network status, insurer payment, member responsibility, remaining accumulator, and adjustments or reversals. Drug claims and medical claims may reside in different adjudication systems. Payments made by third parties or copay assistance programs require special handling under applicable rules. Coverage disputes and denied claims need their final disposition. Administrative data without the patient-billed balance may understate lawful amounts owed outside the plan's allowed price.

The survey requires its own explicit conventions: household unit; interview recall period; current versus annual income; cash spending versus accrued bills; insurance premiums inside or outside health spending; informal payments; and valuation of essential consumption. An out-of-pocket health payment may have been financed by a loan and not yet settled. Alternatively, a respondent may report payments for services never captured as an insurance claim. These are not necessarily errors; they are differences in observation. The OECD indicator definition expressly addresses the composition of household spending and the need for a consistent comparative denominator.

Identification, comparison, and measurement error

To estimate the causal effect of a cap change, a researcher must distinguish the cap from concurrent deductible, premium, network, subsidy, and formulary changes. Cross-sectional differences in plans are especially vulnerable to selection: people anticipating expensive treatment may choose richer plans, and employers or public schemes may offer different benefits to different populations. A credible design could use an exogenous regulatory change, randomized assignment, a defensible eligibility threshold, or a carefully tested quasi-experiment. The identifying assumptions and untreated counterfactual would need to be stated and assessed.

In contrast, comparing the arithmetic consequences of two contract schedules for one fixed set of hypothetical claims requires no causal identification. It answers the conditional legal-financial question: how much would the plan allocate to the patient if those exact services and allowed amounts occurred? That accounting comparison should not be mistaken for the behavioral or equilibrium outcome of imposing the new schedule. Actual spending can change through utilization, coding, negotiated prices, or plan selection. The original Brot-Goldberg employer study shows why such mechanisms cannot be assumed away in empirical policy evaluation.

Coverage data have several recurrent error sources. A billed charge is mistakenly used instead of an allowed amount; a service is classified as out of network despite an applicable protected-billing rule; a claim paid after year-end is assigned to the wrong benefit period; a family accumulator is treated as a sum of individual limits; member cost sharing is counted twice after a claim reversal; or an uncovered service is erroneously added to the maximum. Each can alter estimated exceedance and tail behavior. The audit should preserve the raw and corrected records, record the reason, and recompute outcomes.

Cross-jurisdictional comparison adds conversion and institutional problems. German limits depend on qualifying income and household status, Swiss obligations combine a chosen deductible with capped retention and hospital contributions, the British NHS relies on tax-financed services with service-specific charges and exemptions, and U.S. health plans can have several separate accumulators. A dollar ceiling converted at exchange rates does not create a like-for-like measure of protection. Analysts should compare the underlying obligation schedule and available household resources, then report monetary conversion only with a disclosed date and price-level adjustment.

Sensitivity and validation protocol

An empirical investigation of the right tail should report the incidence of eligible cap exhaustion, the incidence and distribution of all directly paid medical charges, and the incidence of household financial hardship under the chosen methodology. It should further stratify by income, age, chronic care need, coverage type, service class, and network access when data permit. The population denominator must be specified: enrolled members, service users, households, or residents. These denominators cannot be interchanged.

Suggested sensitivity dimensions are the basic-needs benchmark; handling of zero or negative discretionary resources; alternative thresholds, such as 20%, 25%, or 40% where methodologically permitted; premium inclusion for a separate household cash-demand measure; out-of-network classification; benefit-year boundary; and missing payments. The central WHO indicator should be reported according to its official definition, with alternative calculations explicitly labeled as robustness exercises rather than silently substituted for the indicator.

Validation proceeds from legal and textual to numerical and empirical checks. First, map every benefit schedule to the controlling primary document and effective date. Second, rebuild the member liability on sample claims and reconcile it to allowed amounts and insurer payment. Third, reconcile claim-level and household-level spending without double counting. Fourth, recompute the cash-flow and catastrophic-exceedance arithmetic. Fifth, compare selected observations with independent plan summaries or household survey tabulations. Sixth, check that reported uncertainty reflects the sampling design and the limitations of the linkage.

The scope of validation performed for this monograph is documentary and mathematical. Original studies have been evaluated by their published sources or abstracts; no restricted microdata, insurer ledger, original policyholder interviews, or nationwide person-level linkage was acquired. Hence this monograph offers established causal results and auditable conditional financial analysis, not an original estimated risk distribution or statistical confidence interval for its illustrative households.

International Synthesis and Scientific Conclusions

United States: a regulated eligible ceiling amid multiple financing boundaries

The U.S. structure is distinguished by plan-specific annual limits, network contracting, substantial heterogeneity in premiums and benefits, and federal and state rules that change the classification of particular bills. For eligible regulated marketplace coverage in 2026, the official maximum guidance gives $10,600 individual and $21,200 family outer limits, while the full contractual maximum may be lower. A household can nevertheless incur premiums, noncovered services, lawful outside-network charges, and expenses in another benefit period. The regulatory maximum should therefore be reported alongside deductible, coinsurance, covered network, and exclusions.

The Medicare programs show that even one national jurisdiction has different maximum mechanisms. The 2026 Part D annual drug threshold is $2,100 for covered outpatient drugs under the specified program. The independent 2026 Medicare Advantage review documents required limits for covered Parts A and B services in Medicare Advantage and the absence of a general annual maximum for traditional Medicare's Parts A and B; it reports 2026 regulatory maximums of $9,250 in network and $13,900 combined in- and out-of-network for applicable Medicare Advantage plans. Those are Part A/B maxima, distinct from the Part D threshold, and actual plan limits can be lower. Traditional Medicare beneficiaries may have supplementary insurance, so the absence of a standard Parts A/B maximum does not itself establish every beneficiary's eventual payment.

The CMS surprise-billing guidance demonstrates that effective protection depends on legal overrides of an out-of-network label. The central analytical question is not whether a provider is formally out of network in an abstract sense, but whether the services, patient, and plan fall under an applicable billing protection. U.S. comparisons should therefore distinguish insured expenses by at least covered in-network, covered out-of-network protected, covered out-of-network unprotected, and uncovered categories.

Germany: income-linked statutory copayment protection

Germany's statutory health insurance employs an income-related limit on recognized statutory copayments rather than an identical fixed-dollar maximum for all enrollees. The Federal Ministry of Health's official copayment guidance describes a general burden limit of 2% of relevant annual gross household income and a 1% limit for qualifying chronically ill insured persons receiving continuing treatment, subject to detailed rules. Relief from further qualifying copayments follows the statutory burden-limit mechanism. Which co-payments qualify, household allowances, and application procedures are central to the result.

The economic consequence is that the maximum recognized liability varies with a defined income base, while the contractual classification still matters. The 2% and 1% figures should not be interpreted as a guarantee that every form of household health-related spending will remain below those percentages. Unreimbursed products, noncovered care, supplementary insurance premiums, and other expenses can be governed by different rules. The principal comparison with U.S. fixed caps is a structural one: an income-linked burden threshold builds some affordability differentiation directly into the recognized copayment limit. Whether it yields better observed protection for a specified household requires the complete schedule and actual cost data.

Switzerland: deductible choice, capped retention, and a continuing hospital contribution

Swiss compulsory health insurance combines a standard annual deductible with percentage retention and a hospitalization contribution. The Swiss Federal Office of Public Health's co-payment explanation states a standard adult deductible of CHF 300 and a 10% retention fee above that deductible up to CHF 700 per year; people may choose higher deductibles under the regulated system. It also states a CHF 15 daily hospital contribution without a general time limit, subject to specified exemptions, and describes particular medication-related variations.

Under a standard adult deductible of CHF 300, the deductible plus ordinary maximum retention is CHF 1,000, but describing CHF 1,000 as the total annual patient ceiling would be inaccurate. Premiums, applicable hospital contributions, excluded items, and the effect of chosen higher deductibles are separate. The Swiss design illustrates why a cap may apply to one cost-sharing component without creating a universal household maximum. Optional deductible choice adds a prospective tradeoff between premium savings and early-year exposure.

United Kingdom: financing through services and targeted charges

The British NHS illustrates a different institutional form. Much core publicly provided care is financed through public funds and available without the kind of general annual deductible familiar in U.S. commercial contracts; the scope and charging rules vary within the United Kingdom and by service. The GOV.UK NHS costs service specifically describes eligibility-dependent help with prescriptions, dentistry, eye tests, and other costs. The appropriate comparative unit is therefore the entitlement to services, charge exemptions, and residual patient payments, rather than inventing an NHS-wide deductible-plus-coinsurance maximum.

A low catastrophic-payment rate can be consistent with extensive public financing, but it does not by itself establish service access, waiting time, or the absence of unmet need. OECD's 2025 comparative analysis reports low measured catastrophic-payment incidence for the United Kingdom among countries covered by the specified method. This is a financial-spending observation; assessing the full welfare outcome also requires service-access and outcome evidence. The analysis here stays within the verified payment and entitlement scope.

Integrated interpretation and research agenda

These international mechanisms can be compared through four observable questions. First, what service and patient eligibility does the financing arrangement establish? Second, what portion of a recognized claim becomes eligible patient liability? Third, what income, family, network, time, and service conditions constrain that liability? Fourth, what household payments and foregone care remain after all contractual and public protections operate? This is an analytic checklist drawn from the evidence and is not proposed as a new doctrine or proprietary conceptual system.

The most important substantive result is that there are several types of boundedness. A legally stated copayment limit bounds an identifiable set of contributions. An income-based burden threshold changes that set's proportional relationship to household resources. A network and billing rule determines which charges can escape the protection. A system-level entitlement determines whether charges arise at all for particular services. Finally, household financial-hardship measurement detects whether actual payments burden the resources available for basic living. Each is a necessary part of a precise comparison, and no one measure exhausts the others.

The empirical research agenda follows directly. The strongest next study would link claims and benefit accumulators to household payment and welfare data, preserve service-level legal classifications, and record changes across benefit years. It would estimate both eligible cap exhaustion and actual financial-hardship risk, including utilization forgone because of expected payment. It should separately evaluate lower caps, stronger network protection, premium support, and wider service inclusion, with clear counterfactual assumptions. The resulting estimates would tell policymakers and researchers which legal and financing mechanisms actually reduce catastrophic exposure for specified populations.

Conclusion

An out-of-pocket maximum is a precise mechanism of risk transfer within a defined health-benefit boundary. It limits qualifying patient cost sharing, shifts subsequent covered allowed expense to the insurer or public pool, and changes the patient's marginal price at the ceiling. Its protection is legally and economically meaningful. The amount displayed on a plan's summary acquires its scientific meaning only after the covered services, allowed-price basis, individual or family accumulator, network protections, benefit period, and claims administration are identified.

Catastrophic medical cost exposure is broader. A household can exhaust the recognized maximum and still confront lawful out-of-network payments, uncovered services, annual premiums, charges in a subsequent coverage period, debt, and the consequences of delayed treatment. Under a contract with unbounded excluded components, no finite eligible cap alone establishes a finite bound on all medical payments. Under an income-linked or highly subsidized system, protection may operate through a different financial arrangement. Such distinctions explain the observed plurality of insurance designs without collapsing them into one numerical ranking.

The empirical literature supports careful, bounded conclusions. RAND and later nonlinear-contract studies establish that cost sharing can change utilization and expenditure; the observed reductions do not automatically mean that patients found lower prices or avoided low-value care. Oregon's randomized Medicaid expansion demonstrates substantial financial-protection gains from a coverage intervention in its studied population, while offering no isolated estimate for a commercial maximum. WHO and OECD indicators describe financial hardship relative to household resources under documented methods; they do not count claims that exceed plan limits.

The scientific conclusion is therefore operational and distributive: assess the entire eligible-liability schedule, classify residual costs under actual law, validate payments across data sources, and evaluate the resulting burden using a consistent household resource denominator and evidence on access to care. This integrated comparison is the monograph's original analytical synthesis. It creates no new theory, asserts no unverified scientific novelty, and leaves the question of an optimal numerical ceiling to jurisdiction-specific empirical study.

Technical Appendix: Contract and Household Illustrations

Eligible cost-sharing arithmetic

Assume in-network eligible allowed spending G=$75,000, D=$2,000, c=0.20, and M=$8,500. Uncapped patient payment is min(75,000,2,000)+0.20 max(75,000−2,000,0)=$16,600. Capped payment is min(8,500,16,600)=$8,500. Insurer payment for these allowed claims is $66,500. The spending point at which the cap is reached is G*=2,000+(8,500−2,000)/0.20=$34,500.

Complete illustrative cash demand

Add a legally payable, unprotected out-of-network amount N=$7,200, uncovered care U=$4,300, and annual premiums P=$5,600. Direct health payments equal C+N+U=$20,000; payments including premiums equal $25,600. These are constructed figures in U.S. dollars, without a claim that any described service escaped mandatory protections.

Change in the maximum

Hold all other terms fixed and reduce M from $8,500 to $6,000. Because uncapped recognized liability is $16,600, eligible member payment becomes $6,000; direct payments become $17,500; all-in payments including premiums become $23,100. The respective reductions are $2,500. If the person had eligible pre-cap liability of only $4,000, the change would instead save zero in that period.

Household-resource comparison

Assume annual household consumption W=$72,000 and an illustrative basic-needs benchmark B=$42,000. The residual budget is $30,000. With direct health payment Y=$20,000, the illustrative ratio is 20,000/30,000=66.67%. With Y=$17,500 after the cap change, it is 17,500/30,000=58.33%. Both exceed 40%. The calculation is intentionally not labeled as an official WHO or OECD classification; their detailed measurement conventions control official estimates.

Annual boundary

Suppose the insured incurs eligible cost sharing high enough to exhaust an $8,500 maximum in the final month of Year One and again in the opening months of Year Two. If the legally applicable accumulators reset between the two benefit periods, recognized eligible patient payment can total $17,000 across the episode. This is not a violation of either annual ceiling. It demonstrates why medical episodes and benefit-year accounting periods require separate identifiers.

Family embedded maximum

Assume family-wide ceiling F=$12,000, two individual ceilings of $6,000, and uncapped recognized family-member liabilities of $9,000 and $2,000. Under the simplified compatible embedded-benefit assumptions, combined recognized liability is min(12,000, min(9,000,6,000)+min(2,000,6,000))=$8,000. This example does not purport to determine any real family plan's accumulation rule.

Audit notes for the examples

All amounts are hypothetical, use one currency, and assume legally eligible claim classifications. They exclude taxes, supplemental reimbursement, interest, benefit coordination, provider refunds, currency conversion, and real claim timing unless explicitly identified. No illustrative number is a fitted distribution parameter. The claims tested are arithmetic and classification identities, not empirical incidence or causal estimates.

References

  • Aron-Dine, A., Einav, L., & Finkelstein, A. (2013). The RAND Health Insurance Experiment, three decades later. Journal of Economic Perspectives, 27(1), 197–222. Published scholarly article. https://doi.org/10.1257/jep.27.1.197.

  • Baicker, K., Taubman, S. L., Allen, H. L., et al. (2013). The Oregon experiment—Effects of Medicaid on clinical outcomes. New England Journal of Medicine, 368(18), 1713–1722. PubMed record. https://doi.org/10.1056/NEJMsa1212321.

  • Brot-Goldberg, Z. C., Chandra, A., Handel, B. R., & Kolstad, J. T. (2017). What does a deductible do? The impact of cost-sharing on health care prices, quantities, and spending dynamics. Quarterly Journal of Economics, 132(3), 1261–1318. Publisher article.

  • Centers for Medicare & Medicaid Services. (2025). Final CY 2026 Part D redesign program instructions. CMS program instructions.

  • Centers for Medicare & Medicaid Services. (n.d.; checked November 2026). Know your medical bill rights when using insurance. CMS consumer guidance.

  • Commonwealth Fund. (2024). The state of health insurance coverage in the U.S.: 2024 biennial survey. Survey and methodology.

  • Einav, L., Finkelstein, A., & Schrimpf, P. (2015). The response of drug expenditure to nonlinear contract design: Evidence from Medicare Part D. Quarterly Journal of Economics, 130(2), 841–899. Publisher article.

  • Federal Ministry of Health, Germany. (n.d.; checked November 2026). Zuzahlung und Erstattung von Arzneimitteln [Copayments and reimbursement for medicines]. Ministry guidance.

  • Federal Office of Public Health, Switzerland. (n.d.; checked November 2026). Health insurance: Co-payment for persons resident in Switzerland. Federal guidance.

  • HealthCare.gov. (n.d.; checked November 2026). Glossary of Health Coverage and Medical Terms. Official glossary.

  • HealthCare.gov. (n.d.; checked November 2026). Out-of-pocket maximum/limit. Official maximum definition.

  • HealthCare.gov. (n.d.; checked November 2026). Summary of Benefits and Coverage. Official consumer explanation.

  • Kaiser Family Foundation. (2026). Medicare Advantage in 2026: Premiums, out-of-pocket limits, supplemental benefits, and prior authorization. Program analysis.

  • Manning, W. G., Newhouse, J. P., Duan, N., et al. (1987). Health insurance and the demand for medical care: Evidence from a randomized experiment. American Economic Review, 77(3), 251–277. Original AER bibliographic record.

  • Organisation for Economic Co-operation and Development. (2025). Financial hardship and out-of-pocket expenditure. In Health at a Glance 2025. OECD comparative chapter.

  • United Kingdom Government. (n.d.; checked November 2026). Get help with NHS prescriptions and health costs. GOV.UK service guidance.

  • United States Government. (2026 edition, as currently displayed). 45 CFR § 147.126, No lifetime or annual limits. Federal regulatory text.

  • United States Government. (2026 edition, as currently displayed). 45 CFR § 156.130, Cost-sharing requirements. Federal regulatory text.

  • Wharam, J. F., Zhang, F., Landon, B. E., Soumerai, S. B., & Ross-Degnan, D. (2013). Low-socioeconomic-status enrollees in high-deductible plans reduced high-severity emergency care. Health Affairs, 32(8), 1398–1406. PubMed record. https://doi.org/10.1377/hlthaff.2012.1426.

  • World Health Organization. (2025). Financial hardship in health and components (SDG 3.8.2, 2025 definition): Indicator metadata. WHO indicator definition.

  • World Health Organization. (2025). Financial hardship in health and components (SDG 3.8.2, 2025 definition): Population in millions, methodological record. WHO methodological description.

  • World Health Organization & World Bank. (2025). Tracking universal health coverage: 2025 global monitoring report. WHO report and data portal.