Vaccine allocation

 Discussion Draft of the Preliminary Framework for Equitable Allocation of COVID-19 Vaccine

Guiding Principles from Allocation Frameworks Developed for the COVID-19 Pandemic:

• Ensure that allocation maximizes benefit to patients, mitigates inequities and disparities, and adheres to ethical principles.

• Promote the common good through fairness, transparency, accountability, and trustworthiness.

• Save the greatest number of lives possible—while respecting rights and fairness—to

maximize benefit to the community as a whole.

• Use the best available evidence to assess benefit to communities and address uncertainty.

• Allocate scarce resources responsibly to reduce risk while providing benefit.

• Provide clear and transparent criteria for prioritization strategies.

• Ensure that allocation policies are flexible, responsive to the concerns of the affected

population, and proportionate to the epidemiological situation and the vaccine supply relative to need.

How to proceed in practical terms? Who knows...

 Hockney

The Good Doctor_ Why Medical Uncertainty Matters

 The Good Doctor_ Why Medical Uncertainty Matters

The narrative of pandemics (2)

Información científica especializada, información pública y medios de comunicación durante la crisis del coronavirus

Today you'll find our article on communication in pandemic times in Blog Economía y Salud AES, how markets of attention and radical uncertainty drive current situation.

David Hockney

Healthcare built around you?

Bezonomics
How Amazon Is Changing Our Lives, and What the World's Companies Are Learning from It

Amazon Care

Atul Gawande departure from Haven, the alliance between Amazon, JP Morgan and Berkshire for developing health services for their workers has created uncertainty about what is really Amazon going to do.
Anyway, if you look at the web you can check what is already doing for their employees: Amazon Care. Here you'll find the FAQS. Up to now everybody was thinking about health goods and businesses that Amazon could provide (details in this report). Right now they have started a pilot of health system that may be developed anywhere. A platform business, that starts slowly with and app and a physicians group.
If you want to undestand what Amazon really means for the economy (and for healthcare) than a new book can provide you some answers: Bezonomics.

The global business world will eventually divide into two camps—those who adopt their own version of Bezonomics, and those who don’t. Alphabet, Facebook, Netflix, Alibaba, JD.com, and Tencent have built huge, powerful businesses based on their ability to collect and analyze data, and keep applying those learnings to make their businesses smarter and their offerings to customers more attractive. In their pursuit of AI-driven technologies such as voice and facial recognition, the Internet of Things, and robotics, they’re creating automated business models that will crush traditional businesses that fail to adapt to this new world. And the emergence of 5G technology, which will replace our current digital networks, will only widen the gap. Experts predict that this next generation of Internet connectivity will be as much as a hundred times faster than today’s web.

The impact that Bezonomics is having on society is just as profound. Some of the big tech companies are sowing discord with fake news, interfering with elections, and violating personal privacy. As Apple CEO Tim Cook put it: “If you’ve built a chaos factory, you can’t dodge responsibility for the chaos.” The global wealth gap has become so out of kilter that politicians in America and Europe have singled out Amazon and other big tech companies for blame. These wealth-creation machines have become so efficient at creating riches for their top employees and shareholders that they’re likely to engender more public outrage and become easy targets for regulators—perhaps in some cases even be broken up.

A must read. In my opinion, what really brings Bezonomics to healthcare is the largest expression of commercialism. In other words, healthcare built around excedent appropriation, not around the patient. If this is so (and Atul Gawande departure is a signal) then we all have to stand up against this model and create value and platforms based on professionalism.

How epidemic-macroeconomic models of pandemic create uncertainty

Dealing with Covid-19: understanding the policy choices

A model is as good as its assumptions!. This is obvious and the application requires good data. Both issues, assumptions and data are the reasons why many models doesn't fit in this pandemic. Bad assumptions and bad data give bad conclusions. Have a look at this paper and in p.5 you'll find the different health and economic impact of models under different assumptions. So different that require a clever explanation if somebody wants to use them to take a decision.

VSL-based and SIR-macro models have helped to inform policy decisions in the early stages of the Covid-19 pandemic. However, the existing models are subject to a number of caveats, particularly relating to the uncertainty of their underlying epidemiological projections and stylised economic foundations.

 Juan Genovés

QALYs and COVID

Revisiting the Economic Cost of Flattening the Curve

The Incidental economist blog provides information regarding QALY in the current pandemic. Forget the cost per QALY (so difficult to estimate in my opinion) and take only the 6,4 QALYs per death avoided.

It updates previous estimates and says:

The table below summarizes the previous calculations and current updates. Our revisited analysis shows that, as the shutdown continues, the cost per QALY gained increases exponentially due to the exponential growth in the total cost of both forgone productivity and business failure.

We previously emphasized that a key challenge in making calculations of this type is the uncertainty around the data inputs. Six weeks later, this still holds true, particularly for the range of QALY losses without a shutdown, i.e. the predicted corona-related deaths in the absence of intervention.

One interesting aspect of this analysis is that as time goes on, the cost per QALY gained will become higher and higher. This is because the net gains will diminish — the lives saved remains constant, but the offsetting life years lost due to other factors increase — while the costs increase exponentially. The key number that remains unknown is the relationship between the length of the lockdown and the number of lives lost.
In our first post, we concluded that the shutdown would meet conventional standards of cost effectiveness only if the deaths avoided was on the high end of the possible range and the costs on the low end — an outcome that seemed unlikely. Revisiting the issue, it is now clear that the cost per QALY gained from the shutdown will be outside the conventional range of acceptability even at the high end of deaths avoided. How far outside the range the shutdown policy will ultimately prove to be is unknown.

What is going on here?

 Radical Uncertainty
Decision-Making Beyond the Numbers

The question ‘What is going on here?’ sounds banal, but it is not. In our careers we have seen repeatedly how people immersed in technicalities, engaged in day-to-day preoccupations, have failed to stand back and ask, ‘What is going on here?’ We have often made that mistake ourselves.

This is precisely the question that Mervyn King and John Kay pose in their new book Radical Uncertainty. Terrific reading for lockdown days. Below, I've selected some statements:

 The difference between risk and uncertainty was the subject of lively debate in the inter-war period. Two great economists – Frank Knight in Chicago and John Maynard Keynes in Cambridge, England – argued forcefully for the continued importance of the distinction. Knight observed that ‘a measurable uncertainty, or “risk” proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all’

The title of this book, and its central concept, is radical uncertainty . Uncertainty is the result of our incomplete knowledge of the world, or about the connection between our present actions and their future outcomes. Depending on the nature of the uncertainty, such incomplete knowledge may be distressing or pleasurable. I am fearful of the sentence the judge will impose, but look forward to new experiences on my forthcoming holiday. We might sometimes wish we had perfect foresight, so that nothing the future might hold could surprise us, but a little reflection will tell us that such a world would be a dull place.
We have chosen to replace the distinction between risk and uncertainty deployed by Knight and Keynes with a distinction between resolvable and radical uncertainty. Resolvable uncertainty is uncertainty which can be removed by looking something up (I am uncertain which city is the capital of Pennsylvania) or which can be represented by a known probability distribution of outcomes (the spin of a roulette wheel). With radical uncertainty, however, there is no similar means of resolving the uncertainty – we simply do not know. Radical uncertainty has many dimensions: obscurity; ignorance; vagueness; ambiguity; ill-defined problems; and a lack of information that in some cases but not all we might hope to rectify at a future date. These aspects of uncertainty are the stuff of everyday experience.
Radical uncertainty cannot be described in the probabilistic terms applicable to a game of chance. It is not just that we do not know what will happen. We often do not even know the kinds of things that might happen.

Our ability as humans to deal with radical uncertainty is the product of our much greater capacity for social learning and greater ability to communicate relative to other species. We are social animals; we manage radical uncertainty in a context determined by the knowledge we have acquired through education and experience, and we make important decisions in conjunction with others – friends, family, colleagues and advisers.
Reference to the ‘wisdom of crowds’ makes an important point while missing another. The crowd always knows more than any individual, but what is valuable is the aggregate of its knowledge, not the average of its knowledge.

How testing market fails during a pandemic

What Caused the COVID-19 Testing Deficit?

The evidence of market failure during this pandemic is everywhere. Shortages, excessive prices, unavailable capacity...It is a clear example of mismatch between demand and supply. The question is, Can we do it otherwise?. In this article there are some hints for resource allocation for testing activities.

Globally, the development of diagnostics has long been left to markets, many of which are highly specialized. But while there are diagnostics markets for major infectious and non-infectious diseases, and even neglected tropical diseases, there is none for pandemic diseases.
Governments can of course counteract market deficiencies, but the commonly used mechanisms still require a trace level of demand, which does not exist for pandemic-disease diagnostics until the brink of an outbreak. And national governments, subject as they are to political and ideological constraints, cannot be relied upon always to create markets with the same swiftness demonstrated by South Korea. Reactive market creation is therefore not the way forward.
Instead, national governments should support the creation of a global coordinating platform for pandemic preparedness. Such a platform can take the lead in raising and pooling capital to channel toward rapid development, production, and distribution of diagnostics for pandemic diseases.
The blueprint for such a platform already exists. The Coalition for Epidemic Preparedness Innovations (CEPI) is a coordinating mechanism focused on advancing vaccine development and facilitating clinical validation, mass-scale manufacturing, and stockpiling. By reducing uncertainty and minimizing disruptions, CEPI makes vaccine markets more secure, accessible, and dynamic.
CEPI relies on both traditional financing (large grants from governments and foundations) and innovative financing (the returns from instruments like the International Finance Facility for Immunization, or IFFIm). In the event of an outbreak, CEPI uses instruments like Advanced Market Commitments (AMCs) or volume guarantees – which can be structured through mechanisms like the Global Health Investment Fund and InnovFin, or as conditional pledges to IFFIm and Gavi, the Vaccine Alliance – to enable it to scale up production quickly.
This blueprint can easily be replicated for diagnostics. All that is needed is a specialized entity – an institution or initiative that couples research and development with market access. 

Health vs. wealth in a pandemic

HEALTH VS. WEALTH? PUBLIC HEALTH POLICIES AND THE ECONOMY DURING
COVID-19

A NBER paper says:

A pandemic can impact an economy in many ways: reductions in people’s willingness
to work, dislocations in consumption patterns and lower consumption, added stress on the financial system, and greater uncertainty leading to lower investment. These are
respectively referred to as (labor) supply shocks, demand shocks, financial shocks and
uncertainty shocks. Connected economies and epidemiological communities also move in synch. Even a healthy economy, or an economy that has not mandated a shutdown, may feel the impact of external events. With the exception of the 1918 influenza, recent
pandemics have neither had as large of a global impact, nor has there been as much real
time data available to empirically assess the economic and public health impact of NPIs.
We study outcomes during the Covid-19 pandemic.
We have three main results. First, our analysis shows NPIs may have been effective
in slowing the growth rate of confirmed cases of Covid-19 but not in decreasing the growth rate of cumulative mortality. Second, we find evidence of spillovers. NPIs may have impacts on other jurisdictions. Finally, there is little evidence that NPIs are associated with larger declines in local economic activity than in places without NPIs.

Behavioral response to the virus

Using Behavioural Science to Help Fight the Coronavirus

Main topics of the paper:

(1) Evidence on handwashing shows that education and information are not enough. Placing hand sanitisers and colourful signage in central locations (e.g. directly beyond doors, canteen entrances, the middle of entrance halls and lift lobbies) increases use substantially. All organisations and public buildings could adopt this cheap and effective practice.

(2) By contrast, we lack direct evidence on reducing face touching. Articulating new norms of acceptable behaviour (as for sneezing and coughing) and keeping tissues within arm’s reach could help.

(3) Isolation is likely to cause some distress and mental health problems, requiring additional services. Preparedness, through activating social networks, making concrete isolation plans, and becoming familiar with the process, helps. These supports are
important, as some people may try to avoid necessary isolation.

(4) Public-spirited behaviour is most likely when there is clear and frequent communication, strong group identity, and social disapproval for those who don’t comply. This has implications for language, leadership and day-to-day social interaction.

(5) Authorities often overestimate the risk of panic, but undesirable behaviours to watch out for are panic buying of key supplies. Communicating the social unacceptability of both could be part of a collective strategy.  

(6) Evidence links crisis communication to behaviour change. As well as speed, honesty and credibility, effective communication involves empathy and promoting useful individual actions and decisions. Using multiple platforms and tailoring message to
subgroups are beneficial too.

(7) Risk perceptions are easily biased. Highlighting single cases or using emotive language will increase bias. Risk is probably best communicated through numbers, with ranges to describe uncertainty, emphasizing that numbers in the middle are more likely. Stating a maximum, e.g. “up to X thousand”, will bias public perception. 

A known unknown

Coronavirus and the Limits of Economics
Why standard economic theories have no answers for this kind of crisis

You'll find an interesting article in FP

Economists have long made the distinction between uncertainty and risk. Uncertainty is typically understood as involving outcomes that cannot straightforwardly be assigned a probability, unlike risk. Economics offers limited resources to understand how to make decisions in the presence of fundamental uncertainty. But a still deeper form of uncertainty is one in which the possible outcomes cannot easily be anticipated at all. Such a wildly unpredictable outcome has come to be popularly known in recent years as a black swan event.

 The coronavirus pandemic might at first appear to have been such a black swan event, but that claim does not withstand scrutiny: The possibility of such a threat was long recognized by experts. This recognition led to scenarios being discussed at the highest levels of governments. The possibility of a pandemic was therefore a “known unknown” rather than an “unknown unknown.”

Consider that an economy cannot be separated from society: It is socially embedded. The notion that the economy can be analyzed independently of the public health, political, or social processes—often promoted by the dominant tradition in economics and reflected in general equilibrium theory—is shown by the pandemic to be not merely fragile but false.

PS D Rumsfeld stated:

Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don't know we don't know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones.

Galeria Marlborough

Designing a market for lab tests in a pandemic

THE HANDBOOK OF  MARKET DESIGN

Who gets What - and Why

Alvin Roth in his book sheds light on the field known as Market design. Given a set of agents, market design seeks to identify the game rules that might be implemented and that would produce the desired behaviors in the players. In some markets, prices may be used to induce the desired outcomes—these markets are the study of auction theory. In other markets, prices may not be used—these markets are the study of matching theory.

Until recently, economists often passed quickly over matching and focused primarily on commodity markets, in which prices alone determine who gets what. In a commodity market, you decide what you want, and if you can afford it, you get it.The price does all the work, bringing the two of you together at the price at which supply equals demand.

The first question is, can we consider lab tests for a pandemic a commodity?. My impression is that once a virus pandemic has appeared, we need the sequencing for the genome of the virus and after that, several suppliers may offer different options with a high level of uncertainty. Buyers are blind about the options for testing. Suppliers don't know about the extent of the outbreak and the need of tests. The market is rising. In such situations the criteria should be to allocate most of the production to the place where the outbreak has started, in order to prevent contagion. Price may distort this allocation because other countries and buyers may strategically move first to achieve strategic reserves.

Alvin Roth says:

The first task of a successful marketplace is bringing together many participants who want to transact, so they can seek out the best transactions. Having a lot of participants makes a market thick.

Congestion is a problem that marketplaces can face once they’ve achieved thickness. It’s the economic equivalent of a traffic jam, a curse of success. The range of options in a thick market can be overwhelming, and it may take time to evaluate a potential deal, or to consummate it. Marketplaces can help organize potential transactions so that they can be evaluated fast enough that if particular deals fall through, other opportunities will still be available. In commodity markets, price does this well, since a single offer can be made to the entire market (“Anyone can buy a pint of my raspberries for $5.50”), but in matching markets, each transaction may have to be considered separately.

Decisions that depend on what others are doing are called strategic decisions and are the concern of the branch of economics called game theory. Strategic decision making plays a big role in determining who does well or badly in many selection processes. Often when we game theorists study a matching process, we learn how participants “game the system.” Well-designed matching processes try to take into account the fact that participants are making strategic decisions.

When a market doesn’t deal effectively with congestion and participants may not be able to find the transactions they want, it might not be safe for them to wait for the marketplace to open if some opportunities are available earlier. Even when going early isn’t an option, the marketplace might force participants to engage in risky gambles.

The range of tests may be overwhelming, as it is right now with coronavirus. How can we manage such congestion?

There is a chapter in the Handbook of Market Design by Paul Klemperer about The Product-Mix Auction: A New Auction Design for Differentiated Goods. I've read what he proposes and I think that fits quite well with the market for lab tests in a pandemic. Of course, additional details are needed. The research question is:

How should goods that both seller(s) and buyers view as imperfect substitutes be sold, especially when multi-round auctions are impractical?

 My design is straightforward in concept—each bidder can make one or more bids, and each bid contains a set of mutually exclusive offers. Each offer specifies a price (or, in the Bank of England's auction, an interest rate) for a quantity of a specific "variety." The auctioneer looks at all the bids and then selects a price for each "variety." From each bid offered by each bidder, the auctioneer accepts (only) the offer that gives the bidder the greatest surplus at the selected prices, or no offer if all the offers would give the bidder negative surplus. All accepted offers for a variety pay the same (uniform) price for that variety.

The idea is that the menu of mutually exclusive sets of offers allows each bidder to approximate a demand function, so bidders can, in effect, decide how much of each variety to buy after seeing the prices chosen. Meanwhile, the auctioneer can look at demand before choosing the prices; allowing it to choose the prices ex post creates no problem here, because it allocates each bidder precisely what that bidder would have chosen for itself given those prices. Importantly, offers for each variety provide a competitive discipline on the offers for the other varieties, because they are all being auctioned simultaneously.

Compare this with the "standard" approach of running a separate auction for each different "variety." In this case, outcomes are erratic and inefficient, because the auctioneer has to choose how much of each variety to offer before learning bidders' preferences, and bidders have to guess how much to bid for in each auction without knowing what the price differences between varieties will turn out to be; the wrong bidders may win, and those who do win may be inefficiently allocated across varieties. Furthermore, each individual auction is much more sensitive to market power, to  manipulation, and to informational asymmetries than if all offers compete directly with each other in a single auction. The auctioneer's revenues are correspondingly generally lower. All these problems also reduce the auctions' value as a source of information. They may also reduce participation, which can create "second-round"  feedback effects further magnifying the problems.

The rules of the auction are as follows:
1. Each bidder can make any number of bids. Each bid specifies a single quantity and an offer of a per-unit price for each variety. The offers in each bid are mutually exclusive.

2. The auctioneer looks at all the bids and chooses a minimum "cut-off" price for each variety

3. The auctioneer accepts all offers that exceed the minimum price for the corresponding variety, except that it accepts at most one offer from each bid. If both price offers in any bid exceed the minimum price for the corresponding variety, the auctioneer accepts the offer that maximizes the bidder s surplus, as measured by the offer's distance above the minimum price.

4. All accepted offers pay the minimum price for the corresponding variety—that is, there is "uniform pricing" for each variety

It is easy to include additional potential sellers (i.e., additional lenders of funds, in our example). Simply add their maximum supply to the total that the auctioneer sells, but allow them to participate in the auction as usual. If a potential seller wins nothing in the auction, the auctioneer has sold the sellers supply for it. If a potential seller wins its total supply back, there is no change in its position 

My impression is that lab tests in a pandemic require a market design, current allocation methods are relying in a free market that doesn't allows to create value where it it most needed.
Just a final words by Alvin Roth:

Because economics touches on just about everything, economists have an opportunity to learn something from just about everyone, and I’ve met and worked with some remarkable people in each of the markets I’ve helped design.

Market design is giving new scope to the ancient profession of matchmaking. Consider this book a tour of the matching and market making happening around us. I hope it will give you a new way to see the world and to understand who gets what—and why. 

To test or not to test (for coronavirus) (2)

Escaping epidemics through migration? Quarantine measures under incomplete

information about infection risk

Some days ago I was explaining the rationale for coronavirus testing regarding clinical decision making. However, as we all know, individual behavior is also capable to produce health and disease contagion. Therefore, in case of coronavirus, behavioral externalities are crucial and nowadays we have denominated them "social distancing". 

Having said that, there is an additional behavioral value from testing to take into account. If all individuals in a population have access to the test, maybe everybody is aware of social distancing than in a situation than only suspected cases receive the test. Behaviors may change, and quarantine strategies more successful. In such situation it is much more feasible to restrict mobility. Let's take for example what this article explains:

This paper studies the effect of public policies to restrict migration by individuals suspected of carrying disease, when those individuals do not know for certain whether they have the disease but may have more information than the authorities about their probability of being carriers. It has long been known that migration affects the spread of
disease, and this influence has for centuries been used to justify placing restrictions on the movement of individuals suspected of carrying infections.

 Epidemiological studies have addressed how individual behaviour, among other factors, affects the spread of infections. However, the study of how individual behaviour in turn
changes in response to the new incentives created by the occurrence of a disease is much less developed. The principal contribution of our paper is to bring the study of strategic behavior under uncertainty into the domain of epidemiology, and to analyze its impact, in interaction with public policies, on the overall impact of epidemic disease.

Migration as a form of preventive behaviour has received very little attention, although evidence has accumulated that migration behaviour and epidemics are intrinsically linked. Migration behaviour can respond very rapidly to changes in the health  environment, in particular when it suddenly deteriorates through epidemics.

In our model we show that:
• First, when the disease is concentrated in one place (the epicentre of an epidemic for instance), a decision to migrate away from the epicentre brings a potentially infected individual in contact with more uninfected individuals than she would have met had she
remained where she was. Thus the typical migrant imposes a net negative externality as a result of her decision to migrate, and the marginal migrant (for whom, by definition, private benefits of migrating just equal the private costs of doing so) has a negative
impact on social welfare. Laissez faire will therefore lead to excessive migration. This provides a rationale for the frequent (and frequently justified) public policy response to epidemics, which is to attempt to restrict migration away from the epicentre by those who may be infected.
• Secondly, and less obviously, not all policies that aim to restrict migration in fact do so. In particular, we distinguish two effects of quarantine policies. The first is that they raise migration costs, which lowers migration. For example, mandatory health certificates or test results may be required by health authorities to leave the epicentre of the disease.We call this a “type 1” effect of quarantine measures. The second effect is that they impose a utility cost on individuals of remaining in the city where quarantine measures are effective, since they face a chance of being subjected to awkward and possibly
dangerous restrictions on their movements. We call this a “type 2” effect of quarantine measures. Such measures impose a welfare cost on those who suffer them, which tends to increase migration by those who are not currently subject to quarantine but fear they may  become so if they remain where they are. Policies implemented without taking type 2 effects into account may therefore have results that are opposite from those intended.
• Thirdly, even policies that actually reduce migration may have an adverse impact on social welfare if they reduce migration “too much”, and specifically if they discourage those intra-marginal migrants whose private benefits from migration substantially
exceed their private costs of migration, by enough to outweigh the negative externality they impose on others. Overall disease prevalence may even increase if in the name of avoiding negative externalities the authorities discourage relatively low-risk individuals
from escaping the epicentre of the disease, thereby increasing the probability that they will catch the disease there from infected individuals.

When people have imperfect information about their own infection status, migration imposes net negative externalities by increasing the rate of exposure faced by the uninfected outside the epicentre of the epidemic. In and of itself, this our paper has highlighted the fact that although quarantine of individuals who have been identified as sick reduces (obviously) the propensity of these individuals to migrate and spread the disease, the threat of quarantine increases the propensity to migrate of other individuals who have not yet been fallen sick but who know themselves to be at risk.

Quarantine measures have all these effects. However, if information about contagion is confirmed, then behaviors may change, and mandatory health certificates can be issued. The case of the italian village of Vò confirms that population screening has been successful in stopping the outbreak. This could have been done at the beginning if diagnostics kits had been available. Right now it seems an unfeasible strategy. We know now that there is a behavioral value of test information, beyond the clinical value. And in the case of coronavirus, confirmatory tests provide only partial information. In case of non infection, incubation period is uncertain, and some days after can be confirmed. Therefore, quarantine measures have to be mandatory and strict for the whole population and for specific areas.

Confidential drug pricing without confidential prices

Performance-based managed entry agreements for new medicines in OECD countries and EU member states: How they work and possible improvements going forward

In this blog I've explained my position against confidential prices for drugs. However, there is an option to complicate it: confidential entry agreements. This is the current trend for high cost drugs with uncertain outcome. The report of the OECD explains the current situation in different countries and helps to shed light in this important issue. Just take this short statement and you'll be convinced of the complete mess:

It is difficult to assess to what extent performance-based MEAs have so far been successful. Few countries have formally evaluated their experience. Confidentiality of agreements continues to be a barrier to independent evaluation and little evidence is public. However, information available from expert interviews and from prior studies indicates that CED agreements have so far had a poor track record of reducing uncertainty around the performance of medicines. As a result, some countries have recently reformed CED schemes and some are discontinuing CED agreements altogether in favour of alternatives. The latter include restricted or conditional coverage without a MEA, whereby coverage is initially restricted to certain indications or patient groups and only broadened if and when additional evidence becomes available. Payment-by-result agreements continue to be used quite widely, but they do not always generate evidence
on product performance because data used for triggering payments are not always  aggregated and analysed.

Sophia Al-Maria at Tate

Trade-offs in algorithmic clinical decision making

On the ethics of algorithmic decision-making in healthcare

Great article.

Clinicians, or their respective healthcare institutions, are facing a dilemma: while there is plenty of evidence of machine learning algorithms outsmarting their human counterparts, their deployment comes at the costs of high degrees of uncertainty. On epistemic grounds, relevant uncertainty promotes risk-averse decision-making among clinicians, which then might lead to impoverished medical diagnosis. From an ethical perspective, deferring to machine learning algorithms blurs the attribution of accountability and imposes health risks to patients. Furthermore, the deployment of machine learning might also foster a shift of norms within healthcare. It needs to be pointed out, however, that none of the issues we discussed presents a knockout argument against deploying machine learning in medicine.

Book choice

Patient Care under Uncertainty

Medicine as a data science (3)

High-performance medicine: the convergence of human and artificial intelligence

If you want to know the current state of artificial intelligence in medicine, then Eric Topol review in Nature is the article you have to read. A highlighted statement:

There are differences between the prediction metric for a cohort and an individual prediction metric. If a model’s AUC is 0.95, which most would qualify as very accurate,
this reflects how good the model is for predicting an outcome, such as death, for the overall cohort. But most models are essentially classifiers and are not capable of precise prediction at the individual level, so there is still an important dimension of uncertainty.

And this is good summary:

Despite all the promises of AI technology, there are formidable obstacles and pitfalls. The state of AI hype has far exceeded the state of AI science, especially when it pertains to validation and readiness for implementation in patient care. A recent example is IBM Watson Health’s cancer AI algorithm (known as Watson for Oncology). Used by hundreds of hospitals around the world for recommending treatments for patients with cancer, the algorithm was based on a small number of synthetic, nonreal cases with very limited input (real data) of oncologists. Many of the actual output recommendations for treatment were shown to be erroneous, such as suggesting the use of bevacizumab in a patient with severe bleeding, which represents an explicit contraindication and ‘black box’ warning for the drug. This example also highlights the potential for major harm to patients, and thus for medical malpractice, by a flawed algorithm. Instead of a single doctor’s mistake hurting a patient, the potential for a machine algorithm inducing iatrogenic risk is vast. This is all the more reason that systematic debugging, audit, extensive simulation, and validation, along with prospective scrutiny, are required when an AI algorithm is unleashed in clinical practice. It also underscores the need to require more evidence and robust validation to exceed the recent downgrading of FDA regulatory requirements for medical algorithm approval

Therefore, take care when you look at tables like this one:

Prediction	n	AUC	Publication (Reference number)
In-hospital mortality, unplanned readmission, prolonged LOS, final discharge diagnosis	216,221	0.93* 0.75+0.85#	Rajkomar et al.96
All-cause 3–12 month mortality	221,284	0.93^	Avati et al.91
Readmission	1,068	0.78	Shameer et al.106
Sepsis	230,936	0.67	Horng et al.102
Septic shock	16,234	0.83	Henry et al.103
Severe sepsis	203,000	0.85@	Culliton et al.104
Clostridium difficile infection	256,732	0.82++	Oh et al.93
Developing diseases	704,587	range	Miotto et al.97
Diagnosis	18,590	0.96	Yang et al.90
Dementia	76,367	0.91	Cleret de Langavant et al.92
Alzheimer’s Disease ( + amyloid imaging)	273	0.91	Mathotaarachchi et al.98
Mortality after cancer chemotherapy	26,946	0.94	Elfiky et al.95
Disease onset for 133 conditions	298,000	range	Razavian et al.105
Suicide	5,543	0.84	Walsh et al.86
Delirium	18,223	0.68	Wong et al.100

1. LOS, length of stay; n, number of patients (training+ validation datasets). For AUC values:
2. *, in-hospital mortality; + , unplanned readmission; #, prolonged LOS; ^, all patients; @, structured + unstructured data; + +, for University of Michigan site.

Searching for principles to inform societal decisions

Health Economics from Theory to Practice: Optimally Informing Joint Decisions of Research,  Reimbursement and Regulation with Health System Budget Constraints and Community Objectives

Cost effectiveness analysis without a clear understanding of budget impact is a theoretical effort with limited consequences in practice. If you want a clear view of the whole process and go beyond cost-effectiveness, a new book tries to summarise the current state of the art.
I'm not so sure that the title "Health economics from theory to practice" really fits with content you expect, but think about a different title "cost-effectiveness from theory to practice" and this is precisely what you'll find.

This book provides a robust set of health economic principles and methods to inform societal decisions in relation to research, reimbursement and regulation (pricing and monitoring of performance in practice). We provide a theoretical and practical framework that navigates to avoid common biases and suboptimal outcomes observed in recent and current practice of health economic analysis, as opposed to claiming to be comprehensive in covering all methods. Our aim is to facilitate efficient health system decision making processes in research, reimbursement and regulation, which promote constrained optimisation of community outcomes from a societal perspective given resource constraints, available technology and processes of technology assessment. Importantly, this includes identifying an efficient process to maximize the potential that arises from research and pricing in relation to existing technology under uncertainty, given current evidence and associated opportunity costs of investment. Principles and methods are identified and illustrated across health promotion, prevention and palliative care settings as well as treatment settings. Health policy implications are also highlighted.

And the conclusion:

The framework and methods presented have been shown to enable optimising of joint research, reimbursement (adoption and financing) and regulatory (pricing and practice monitoring) processes and decision making. Jointly addressing these related decisions has been shown to be key in meeting current and future challenges of baby boomer ageing and more generally in identifying areas for policy reform to enable a pathway to budget-constrained optimisation of community net benefit. The bottom line for such reforms is that better use of existing programmes and technologies and associated research that reflect community preferences is required and particularly now in facing the challenge of budget-constrained successful ageing of the baby boomer cohort.

Cost-effectiveness of new (genomic) benefits, it's just the begining

HERC database of health economics and genomics studies 
Cost-effectiveness of cell-free DNA in maternal blood testing for prenatal detection of trisomy 21, 18 and 13: a systematic review

Just yesterday our government suddenly decided to introduce a new benefit in public insurance coverage: contingent DNA based non-invasive prenatal screening. And the question is: does someone know if this new benefit is cost-effective?
You can get the answer after reading this review article, and the summary is:

 In total, 12 studies were included, four of them performed in Europe. Three studies evaluated NIPT as a contingent test, three studies evaluated a universal NIPT, and six studies evaluated both. The results are heterogeneous, especially for the contingent NIPT where the results range from NIPT being dominant to a dominated strategy. Universal NIPT was found to be more effective but also costlier than the usual screening, with very high incremental cost-effectiveness ratios. One advantage of screening with NIPT is lower invasive procedure-related foetal losses than with usual screening. In conclusion, the cost-effectiveness of contingent NIPT is uncertain according to several studies, while the universal NIPT is not cost-effective currently.

If this is so, since uncertainty is the word that better reflects its current cost-effectiveness, why do the have introduced? Because they don't care about it. These are not the best days for a health economist (and for the society as a whole). Maybe it's just the begining of a new world without scarcity, and I can't figure out.

Manuel Anoro

The p53 nightmare

p53 and Me

This week you'll find a short piece in NEJM, a story written by a physician on how detecting a genetic p53 mutation changed her views. Key message:

Genetic knowledge is power only if both clinician and patient are equipped to move beyond a result and toward action, even if that merely means living well with what we know. I believe we need an expanded definition of genetic counseling; we require more data, yes, but also more sophisticated and sensitive ways of assimilating such data. And not just into databases we can mine to see what happens to people like me, but into programs for learning to live with uncertainty.

A must read