Andrés Pichon-Riviere: “Budget impact should not decide what is adopted, but how”

VOICES · HEALTH TECHNOLOGY ASSESSMENT · ARGENTINA

Andrés Pichon-Riviere, a physician, is director general of the Institute for Clinical Effectiveness and Health Policy (IECS), a post to which he was re-elected in April for 2026–2030. He also co-directs its Pan American Health Organization (PAHO)/WHO Collaborating Center on health technology assessment and health economics. He is a full professor of public health at the University of Buenos Aires, holds the rank of principal researcher at the National Scientific and Technical Research Council (CONICET) and was vice president of the International Network of Agencies for Health Technology Assessment (INAHTA) until 2014.

He speaks as Argentina argues over who decides which medicines get paid for. In March 2025 the government announced a National Agency for the Evaluation of Health Technology Financing; 19 months later it is still not operating, according to press reports, although in July the Inter-American Development Bank approved an $800,000 technical-cooperation grant that includes support for creating it. The provinces are moving first: Mendoza’s assessment agency, created by law in 2024, issues opinions that bind the provincial health ministry and the province’s own insurers, and in June Neuquén created one whose opinions bind the provincial social insurer. In August PAHO held a workshop in Neuquén where four provinces discussed cooperation on health technology assessment. Coverage of high-cost drugs is under strain too: the fund that pays for it, which also finances disability services, has delayed disability payments, and disputes keep reaching the courts.

In this written interview, he sets out which decisions an agency could inform or take, how to keep budget limits from becoming hidden rationing and why price should be anchored in value.

The interview

In April 2025 you told a congressional health committee that many countries in the region had focused “on how to produce reports rather than on how to make decisions.” What concrete decisions, not just reports, should a national agency be empowered to take in a system as fragmented as Argentina’s?

My warning was about a sequence that tends to get reversed. Before discussing an agency’s budget, makeup or powers, you have to define which concrete decision it will inform or take.

In a country without a single health system (unlike the United Kingdom or, in part, Brazil), the question is not only what is decided but whom the decision binds. The same technology is assessed, paid for and litigated by the national and provincial public sector, the social health insurers (obras sociales), PAMI (the national retirees’ insurer) and the prepaid private insurers. The design has to specify, for each type of decision, which payers it binds and for which it is only advisory.

The range of possible decisions is wide, and there is no single obvious model. It runs from coordinating the assessments that provinces and payers already produce and setting common methodological standards (the initial role of Canada’s agency, geared more to coordinating than deciding) to deciding what goes into the mandatory benefits package, and under what conditions of use and for which indications, not just yes or no; setting the maximum price the system will pay, or risk-sharing and price-volume agreements; prioritizing disinvestment from low-value technologies; and giving technical grounding to court cases.

Creating the agency is necessary and long overdue. But a powerful agency whose decision-making mandate has not been precisely defined reproduces, under another name, the fragmentation it is meant to correct.

Meanwhile the provinces are moving: Mendoza’s agency issues binding opinions, and in June Neuquén created one. Is this the basis of a federal assessment system, or a risk that the same technology gets different answers depending on the province?

Argentina has long had many bodies assessing technologies for different subsectors, from a unit at the Superintendency of Health Services and units in provincial ministries to academic institutions such as IECS and units within social health insurers and prepaid insurers. The new provincial agencies are the most institutionalized expression of something to be expected with fragmented financing.

I do not see that as a problem in itself. Several countries have more than one assessment body, and horizontal coordination between provinces is already beginning here. One example is RedARETS, Argentina’s Public Health Technology Assessment Network, a federal non-profit network created in 2012 that brings together public health-sector and academic institutions. So “federal system or different answers” is not the right way to frame it: a federal assessment system allows for, and even expects, different answers.

The risk is not that two provinces decide differently. The same technology can justify different decisions depending on local epidemiology, costs or priorities. The risk is that they decide differently because they started from different evidence. That is what must not happen. A second risk is duplication: dozens of units assessing the same technologies, multiplying the effort and the chance of error.

That is why a national agency is decisive. Its value is not to centralize every decision but to guarantee a common base: distributing assessments so they are not duplicated, setting methodological standards, including cost-effectiveness and budget impact thresholds, and above all producing the components common to all, starting with clinical effectiveness. That is Europe’s path: since 2025 the European Union has carried out joint clinical assessments, each one for all member states, to which each country adds its own economic and local considerations to reach its own decision, which need not be the same. The basis of a federal system is already there. What is missing is the common layer of evidence that turns it into a system rather than an overlap.

According to the Civil Association for Equality and Justice (ACIJ), the 2027 budget bill would raise health spending 16% in real terms, though ACIJ notes that the sector’s growth since 2023 owes much to PAMI funds now being recorded as Treasury transfers. Your recent work sets out a framework for budget impact thresholds. How should an Argentine payer set that limit without it becoming implicit rationing?

Good decisions cannot be made without economic evidence, nor with economics alone. Our two papers in The Lancet Global Health, on cost-effectiveness thresholds in 2023 and budget impact thresholds in 2025, aim to give countries explicit economic rules on which to then deliberate.

For a budget impact threshold not to become implicit rationing, it must not be confused with the cost-effectiveness threshold. One tells you whether a technology is worth what it costs; the other, whether the system can absorb its total cost over a period. A technology can be highly cost-effective and have a high budget impact, typically in common diseases. Applied as a yes-or-no barrier, the budget threshold becomes implicit rationing again, this time not for lack of a rule but from applying the wrong one: it leaves out valuable technologies just because they are big, precisely those that benefit the most people. Budget impact should not decide what is adopted, but how. Used well, it is a signal, not a wall: a high impact should trigger more intense deliberation and a different negotiation, such as risk-sharing, price-volume agreements or phased entry, not an automatic rejection.

In Argentina that requires three conditions. Express it as a share of the health budget, not a nominal figure, which inflation turns into a hidden cut. Anchor it to the resources actually available for new technology, not to an accounting aggregate that grows partly because PAMI funds are now recorded as Treasury transfers while specific programs shrink in real terms, and so overstates the fiscal space. And define it by payer: the same technology has a very high impact on a small social insurer and a moderate one on PAMI.

Above all, the threshold has to be explicit and public. Rationing is illegitimate when it works without being named. A declared threshold, used to structure negotiation rather than to exclude, turns implicit rationing into transparent prioritization, informed by economics but not decided by economics alone.

The Solidarity Redistribution Fund finances both disability services and reimbursements for social insurers’ high-cost treatments. In August, according to press reports, the Superintendency of Health Services blamed delays in its transfers on lower tax collection. What design changes would stop high-cost coverage from depending on each month’s tax take?

It is a delicate question, but the core of the problem is one of design, and it can be framed independently of the current situation. A benefits package makes sense only if people can count on it. Today high-cost care is paid out of what is left in a payroll-based fund that rises and falls with formal employment. That spending has the opposite profile: infrequent and large, hard to predict case by case but easy to pool. It is exactly the kind of risk that should be pooled and smoothed, not paid out of current cash.

So, first, pool the high-cost risk at the widest possible scale and ring-fence it from other subsidies competing for the same monthly collections, as Uruguay’s National Resources Fund does. Second, diversify its financing beyond payroll contributions, with co-financing from general revenues or earmarked taxes, so it does not follow the cycle of formal employment. Third, give it a countercyclical reserve and a multi-year budget horizon. Fourth, make the obligations predictable: define in advance, through technology assessment, what the high-cost basket includes, and negotiate prices centrally.

For society and decision-makers to commit funds, they need clear rules on how much is spent and on what. It would be desirable for Argentina, like the whole region, to spend more on health; there are few more efficient investments for a country than its population’s health and preventing tomorrow’s costs. But that is sustainable only if resources are shown to be used efficiently and do not end up in a bottomless pit, something that happens in health more often than one would like.

“Rationing is illegitimate when it works without being named.”

On July 7th the Supreme Court overturned, and sent back for a new decision, a ruling that had ordered a prepaid insurer to supply an imported cystic fibrosis drug when a locally made version with the same active ingredients, approved by the national regulator, was available. What role should health technology assessment play so that coverage of high-cost medicines stops being decided case by case in the courts?

The judicialization of health in Argentina is a symptom of a deeper problem: the lack of clear, explicit and predictable rules for coverage decisions. This ruling shows it in a way that is often overlooked: this time the Court sided with the payer. The problem is not the direction of the rulings, which sometimes widen coverage and sometimes narrow it, but that the package is still defined case by case, without a general rule anyone can know in advance.

The case had the courts settle two questions that are technical, not legal: what is guaranteed and for whom, and whether a local product with the same active ingredients is truly interchangeable with the imported one. The second was settled with an expert opinion on a single patient and the argument that our rules do not require bioequivalence studies to market the locally made similar product. But therapeutic interchangeability, especially for complex molecules, should be determined generally and in advance, on technical criteria, not presumed or decided patient by patient.

That is the role of technology assessment: define what the package includes and under what clinical conditions; establish, with the regulator, when an alternative is equivalent; and set the price and terms of adoption beforehand. Today, faced with a new, very effective high-cost molecule for a serious disease, nobody can say with certainty whether it will be available, for which patients and at what price.

The courts play an important role, as they do in many countries: ensuring that people actually get what their benefits package already guarantees. Their role should not be to define the package, whether to widen it or to cut it, as happens today in Argentina and elsewhere in the region. When there is an explicit, well-grounded coverage decision, the courts can check that the process was respected, instead of substituting their judgment for a decision that was never taken. Today the courts are filling, badly, a gap in the decision-making process.

“Today the courts are filling, badly, a gap in the decision-making process.”

A study in the journal Medicina this year found that 57% of the full-assessment recommendations by the National Commission for Health Technology Assessment and Clinical Excellence (CONETEC) favored conditional coverage, with suggested discounts of 50% to 83.6% on retail prices. In August an unpublished government report put retail prices for 17 active ingredients at a median 3.1 times those in five reference countries; the Industrial Chamber of Argentine Pharmaceutical Laboratories (CILFA) objected that list prices ignore what payers actually pay. Which instrument would most affect what payers and patients really pay?

A medicine’s list price has less and less to do with what health systems actually pay, and not only in Argentina. Payers negotiate, ideally on the value a technology brings their population, and those negotiations are rarely public. CILFA’s objection is partly right, but that opacity is itself the problem: without knowing the real price or the value that would justify paying it, there is no way to negotiate well or set policy.

The debate mixes up two things that need different instruments. One is what patients pay out of pocket for everyday medicines: the government report compares retail prices of old, off-patent drugs, and what brings those down is competition, meaning prescribing by active ingredient, generics and regulation that guarantees the alternatives are truly equivalent. The other is what payers pay for high-cost technologies, which is where the system’s sustainability is at stake.

There, the instrument with the most impact is value-based price negotiation, and its input is technology assessment. Knowing how much a technology really contributes lets you set the price at which it is cost-effective for that health system, and that price anchors any negotiation, risk-sharing agreement or conditional coverage. CONETEC already does it on a small scale, as the 57% figure shows. When coverage is tied to value, the price moves.

Global price-transparency initiatives, from WHO and others, help, but they are secondary to this. For Argentina, what another country pays matters less than how much a technology is worth to ours. A national agency able to answer that would give a fragmented system, negotiating blind, the anchor it lacks: value.

Disclosure: Pichon-Riviere has no contracts with drug or insurance companies in a personal capacity. IECS works with, and holds contracts with, pharmaceutical companies and insurers, as well as governments, universities and international agencies, among others.

Sources: IECS (re-election, April 2026); CONICET; ISPOR (INAHTA); Government of Argentina (national agency, March 2025); Cadena 3 (agency status, October 2026); Inter-American Development Bank (AR-T1385); Mendoza Law 9547; Noticias NQN (Neuquén agency); Noticias NQN (binding on the provincial insurer); PAHO (Neuquén workshop, August 2026); Superintendency of Health Services (Solidarity Redistribution Fund); Government of Argentina (fund and disability services); Cipo360 (payment delays); eHealth Reporter (congressional hearing, April 2025); The Lancet Global Health (budget impact thresholds, 2025); ACIJ (2027 budget); Supreme Court (G. B., R. v. OSDE, July 7th 2026); Medicina (Donato et al., 2026); La Nación (price report and CILFA, August 2026).