Now is the time for artificial intelligence in healthcare

Artificial intelligence in health care: within touching distance

Medical practice has so far been largely unchanged by the digital revolution that has disrupted so many other industries, but perhaps artificial intelligence (AI) will provide the improvements in medical care and research promised for so long.

A short editorial in Lancet highlights the importance of deep learning in healthcare.

In 2017, successful use of deep neural networks was reported for the analysis of skin cancer images with greater accuracy than a dermatologist and the diagnosis of diabetic retinopathy from retinal images. The inherent requirement for large-scale, high-quality, well structured data might ultimately limit the areas in which AI can bring benefits to health care.

 Jordi Parramon exhibition

The urgent need to define delivery models for genetic testing

Identification of Delivery Models for the Provision of Predictive Genetic testing in Europe: Protocol for a Multicentre Qualitative study and a systematic review of the literature

The increasing role of genomics in medical decision making requires a review on how services should be organised. Unless this effort is taken promptly, it will be much more difficult to adapt the messy organization to an efficient model for the delivery of services. This issues are explained in a recent article. The ten questions:

 The transfer of genomic technologies from research to clinical application is influenced not only by several factors inherent to research goals and delivery of healthcare but also by external and commercial interests that may cause the premature introduction of genetic tests in the public or private sector (i.e., introduction of a test despite insufficient evidence regarding its analytical validity, clinical validity, and utility). Furthermore, current genetic services are delivered without a standardized set of process and outcome measures, which are essential for the evaluation of healthcare services. It is important that only genetic/genomic applications with proven efficacy and effectiveness are delivered to populations, and particularly that technologies have favorable cost-effectiveness ratios

Why we must not let the tech and drug industry forge the future alone

On the tech industry by Martin Wolf in FT

Selected statements on 7 reasons

What are the economics of these extraordinary valuations? The answer must be monopoly. As of September 30, the book value of Apple’s equity was $134bn, while its market valuation was close to $900bn. The difference has to reflect the expectation of enduring “super-normal” profits. This may not be the product of malign behaviour, but of innovation and economies of scale and scope, including the network externalities that lock in customers. Yet only monopoly could deliver such super-normal profits

How should we think about competition policy for businesses that benefit from such powerful monopoly positions? A question is whether these positions are temporary — as the great Austrian economist, Joseph Schumpeter, with his idea of “creative destruction”, would argue — or lasting. This suggests a host of responses, but one at least seems straightforward. Schumpeter would argue that new entries are a necessary condition for eroding such temporary monopolies. If so, the technology giants should be strongly deterred from buying up their potential competitors. That must be anti-competitive

Yet these enormously profitable businesses are parasitic on the investments in collecting information made by others. At the limit, they will become highly efficient disseminators of non-information. This links to a further point: they can, as we now know, be used by people of ill will for the deliberate dissemination of dangerous falsehoods. These facts raise huge issues.

Finally, the activities in which the technology industry is now engaged — what Andrew McAfee and Erik Brynjolfsson call “machine, platform, crowd” — are going to have a huge impact on our labour markets and, if artificial intelligence continues to advance, on our very place in the world.

What are the implications? They are that our futures are too important to be left to the mercies of the technology industry alone. It has done magical things. Yet nobody elected it master of the universe. Policymakers must get an intellectual grip on what is happening. The time to begin such an effort is now

On a particular drug company, in Project Syndicate:  The Opiate of the Bosses

Business ethics are again making headlines. This time, the focus is on the rapidly escalating opioid crisis that is destroying lives across the United States. While there is plenty of blame to go around, the largest share of the guilt belongs squarely on the shoulders of the major drug companies – Big Pharma.

The cynicism with which pharmaceutical firms have encouraged opioid drug use is appalling. Providing far too little analysis and oversight, they distribute opiates widely, alongside misinformation about how addictive the drugs truly are. Then they entice doctors with inducements and giveaways – including trips, toys, fishing hats, and, in one case, a music CD called “Get in the Swing with OxyContin” (one of the most popular opioids) – to prescribe them.

In 2007, several executives of the parent company of Purdue Pharma, which markets OxyContin, pleaded guilty to misleading doctors, regulators, and patients about the risk of addiction associated with the drug. The company was hit with some $600 million in fines and penalties.

Estimating individual life expectancy for alzheimer patients

Personalized predictive modeling for patients with Alzheimer's disease using an extension of Sullivan’s life table model

Alzheimer's disease is the most common type of dementia. Ageing is boosting its spread over populations. Eric Stallard et al. asked wether it was posible to estimate the residual total life expectancy (TLE) and its decomposition into disability-free life expectancy (DFLE) and disabled life
expectancy (DLE) for individual patients. It sounds really of interest, though it may seem unattainable.
Fortunately you may find succesful results in this article, it says:

Methods: We estimated a new SLT/L-GoM model of the natural history of AD over 10 years in the Predictors 2 Study cohort: N = 229 with 6 fixed and 73 time-varying covariates over 21 examinations covering 11 measurement domains including cognitive, functional, behavioral, psychiatric, and other symptoms/signs. Total remaining life
expectancy was censored at 10 years. Disability was defined as need for full-time care (FTC), the outcome most strongly associated with AD progression. All parameters were estimated via weighted maximum likelihood using data-dependent weights designed to ensure that the estimates of the prognostic subtypes were of high quality.
Goodness of fit was tested/confirmed for survival and FTC disability for five relatively homogeneous subgroups defined to cover the range of patient outcomes over the 21 examinations.

Results: The substantial heterogeneity in initial patient presentation and AD progression was captured using three clinically meaningful prognostic subtypes and one terminal subtype exhibiting highly differentiated symptom severity on 7 of the 11 measurement domains. Comparisons of the observed and estimated survival and FTC disability probabilities demonstrated that the estimates were accurate for all five subgroups, supporting their use in AD life expectancy calculations. Mean 10-year TLE differed widely across subgroups: range 3.6–8.0 years, average 6.1 years. Mean 10-year DFLE differed relatively even more widely across subgroups: range 1.2–6.5 years, average 4.0 years. Mean 10-year DLE was relatively much closer: range 1.5–2.3 years, average 2.1 years.

Excellent, good job from Duke University, where I did part of my PhD, using the same methodology Grade of Membership.

PS. My speech at the Economist's day.

Anders Zorn au Petit Palais



Improving health in OECD countries

Health at a Glance 2017

Let me highlight toady one figure: 54% of adults in OECD countries today are overweight, including 19% who are obese. Obesity rates are higher than 30% in Hungary, New Zealand, Mexico and the United States.
Many indicators about current state of health in OECD and partner countries can be found in the report. And the public health message is:

While smoking rates continue to decline, there has been little success in tackling obesity and harmful alcohol use, and air pollution is often neglected

These are the new epidemics and prescriptions are not easy to find to curb the current trend.
The report shows many positive messages and this should be a reason for trust in our future, however the uncertainties regarding new risks and how to tackle remain.

Gramophone All Stars Jazz Band. Maraca Soul album. Iko Iko

Uncertainty and regret in medicine

The Power of Regret

While reading NEJM I find an article on regret:

As physicians, we are acutely aware of the element of uncertainty in medicine, but we less often recognize its close companion, regret. Regret in all its forms can be a powerful undercurrent, moving patients to act in ways that may baffle us.

Kahneman and Tversky said  that bad outcomes from recent action are more regretted than similar outcomes from inertia. There two types of bias that affect regret. Omission bias is the tendency toward inaction or inertia — reflects anticipated regret. Commission bias is the tendency to believe that action is better than inaction, and can result in regret arriving later when a bad outcome occurs.

When we’re in pain or acutely anxious, we are “hot” and apt to make choices that we imagine will rapidly remedy our condition, which predisposes us to commission bias. In a hot state, patients may discount too deeply the risks posed by a treatment and overestimate its likelihood for success, paving the way for later regret if the outcome is poor. Patients who choose elective procedures while in a hot
state and end up with a bad outcome may be at particular risk for regret due to commission bias.

Georges Seurat. Two Sailboats at Grandcamp (Deux voiliers à Grandcamp), c. 1885.
Oil on panel, BF1153. Public Domain. Barnes Foundation

The apocalypse and our true fate, who knows?

THE FIVE HORSEMEN OF THE MODERN WORLD: Climate, Food, Water, Disease, and Obesity

In the book of Revelation or Apocalypse of John, you'll find the seven bowls. Seven angels are thus given seven bowls of God's wrath, each consisting of judgements full of the wrath of God poured onto Earth:

First Bowl: A "foul and malignant sore" afflicts the followers of the Beast. (16:1–2)
Second Bowl: The Sea turns to blood and everything within it dies. (16:3)
Third Bowl: All fresh water turns to blood. (16:4–7)
Fourth Bowl: The Sun scorches the Earth with intense heat and even burns some people with fire. (16:8–9)
Fifth Bowl: There is total darkness and great pain in the Beast's kingdom. (16:10–11)
Sixth Bowl: The Great River Euphrates is dried up and preparations are made for the kings of the East and the final battle at Armageddon between the forces of good and evil. (16:12–16)
Seventh Bowl: A great earthquake and heavy hailstorm: "every island fled away and the mountains were not found." (16:17–21)

As you may notice Apocalypse is just that, a book. Daniel Callahan set a title of five horsemen of the modern world as a metaphor of current evils. Global warming, food shortages, water shortages and quality, chronic illness, and obesity could be the key ingredients of our fate?.
At the end, Daniel Callahan calls for a diplomatic model:

to persuade the research, academic, and policy communities to accept what I will call the diplomatic model of relationships, typically now seen between and among nations, and to open a serious dialogue with the business community

Agree.

Obesity: a multifaceted approach

The Current State of Obesity Solutions in the United States

We all agree that we need to face the obesity epidemic. But, when talking about solutions, difficulties and uncertainties  arise. US National Academies held a worksop on 2014 that described interventions designed to prevent and treat obesity in seven settings:
• early care and education,
• schools,
• worksites,
• health care institutions,
• communities and states,
• the federal government, and
• businesses and industry
The book is only a first approach to these experiences, though more evidence is needed in my opinion. They say in the book:

"Much of what needs to be done is clear, he said. The challenge now is to figure out how to do what needs to be done."

 DocField 2017

Now in Barcelona

How to change individual letters of your DNA?

Gene editing has made another step forward. And maybe a complementary to the former one, the CRISPR-Cas9,  that was proved viable by Jennifer Doudna and I explained some months ago in this post. No it is indeed more interesting. Two different approaches, base editing and CRISPR-Cas13, have been described in Science and Nature. Adenine base editing allows to correct mutations, it doesn't cut the gene to insert a new one. It is a sharp pencil rather than scisors. With CRISP-Cas13 it is possible to edit RNA, which converts genetic information into proteins. An exciting approach, you correct a book with temporary ink that disappears, rather than making a permanent mark (like in CRISPR-Cas9).
These are exciting times for genetic research, though we'll have to wait for specific clinical applications

Modigliani, now at Tate Gallery

Voluntary health insurance: fulfilling expectations

Memòria entitats d'assegurança lliure 2015
Regulació de l'assegurança voluntària de salut

Let's take one country that has a mandatory social security system for the whole population, though its funding comes from taxes (?). If 25% of the population in this country voluntarily buy  duplicate coverage for the roughly the same benefits, what would you say?. The potential answer is that the public system is not fulfilling people expectations and has a big problem. Unfortunately, politicians don't recognise the situation. Imagine that in the capital more than one third of the population hold private insurance, you would say indeed that the problem is larger. This is the case of Barcelona.
Somebody should review the situation. Both public and private systems have their drawbacks. If public mandatory funding is not providing an efficient system, than a prescription is needed. If voluntary health insurance solves the unfulfilled expectations, then a close relationship should be established, and this is not an option by now.
I wrote a paper some time ago on the required new regulation for voluntary health insurance. My impression is that nobody read it. Maybe now it's the time.

PS. Right now 735.997 patients are waiting for a surgery, a visit or a diagnostic procedure.

 

 

Richard Estes. Galeria Marlborough

The end of marginal revolution

Richard Thaler was awarded with the Nobel Prize some days ago. If you follow this blog you'll know his works on behavioral economics and nudging. Since many years I've been interested in this perspective, though it has still more to deliver.
Today I would suggest you to read JM Colomer blog. He has written an excellent post on him and its impact on economic science. Selected statements:

Marginalist microeconomics held that we could understand collective outcomes by assuming that they derive from free interactions among homines economici.
A first big counter-revolution was the reintroduction of institutions in the basic analysis, especially since the 1980 and 1990s (including by Nobel laureates related to the social choice and public choice schools such as Kenneth Arrow, James Buchanan, Ronald Coase, Douglass North, Amartya Sen, Thomas Schelling, Leonid Hurwicz, Roger Myerson, political scientist Elinor Ostrom, Oliver Williamson, and others).

The second is the reintroduction of realistic observations about people’s motivations and behavior, including emotions. This has been based on psychology, on the background of huge progress in neuroscience (while pioneers include political scientist Herbert Simon and psychologist Daniel Kahneman). That Richard Thaler professes at the University of Chicago, once the temple of the neoclassical school, shows the depth of the change.

Now we know again that the three pillars of social analysis are, together with people’s calculated self-interested choices, emotions and institutions, as Hume and Smith masterfully had already established.

And this is the return to the roots of economics with a new toolkit.

Parov Stelar

Understanding Generic Drug Markets

Comparing Generic Drug Markets in Europe and the United States: Prices, Volumes, and Spending

The development of generic drug markets depends widely on an active regulator. This is the main reason of differences in consumption among countries. A new article highlights these differences and allows to understand better such market.

Substituting generic medicines for more expensive brand-name versions is likely among the most cost-effective interventions in health care systems.

There remain large differences in the usage and prices of generics in Europe and the United States. The barriers to market entry for generic companies vary between countries, as do pricing and reimbursement policies. Beyond such features of the market,
there are differences in whether, and to what extent, patients and health care professionals perceive generic and branded medicines to be bioequivalent.

Healthcare Quality Lessons

Caring for Quality in Health

An OECD report provides the lessons on caring for quality, quite general but of interest to dive into each one:

Lesson 1. High-performing health care systems offer primary care as a specialist service
that provides comprehensive care to patients with complex needs
Lesson 2. Patient-centred care requires more effective primary and secondary prevention
in primary care
Lesson 3. High-quality mental health care systems require strong health information systems
and mental health training in primary care
Lesson 4. New models of shared care are required to promote co-ordination across health
and social care systems
Lesson 5. A strong patient voice is a priority to keep health care systems focussed
on quality when financial pressures are acute
Lesson 6. Measuring what matters to people delivers the outcomes that patients expect
Lesson 7. Health literacy helps drive high-value care
Lesson 8. Continuous professional development and evolving practice maximise
the contribution of health professionals
Lesson 9. High-performing health care systems have strong information infrastructures
that are linked to quality-improvement tools
Lesson 10. Linking patient data is a pre-requisite for improving quality across pathways
of care
Lesson 11. External evaluation of health care organisation needs to be fed into continuous
quality-improvement cycles
Lesson 12. Improving patient safety requires greater effort to collect, analyse and learn
from adverse events 

It is like a check list, have you done your homework?

Searching for a book to read

Manel Castro

Precision medicine initivatives around the world

Human genomics projects and precision medicine

Governments and research funders in developed world have decided to support precision medicine with different initiatives. Its scope and strenght it is quite diverse. It is good to know what's going on, and this is explained in an article in Nature. A data driven medicine is raising with next generation sequencing (NGS) tools:

The tremendous amount of data that NGS technologies are producing and the difficulties to manage and analyze such quantity of data require the implementation of powerful data centers for storage and analysis. Nevertheless, recent improvements in cloud computing allow managing and analyzing these huge data amounts remotely. With this goal in mind, the main internet companies have taken positions to compete in this area of NGS (data storage and analysis).
As three main examples, Google Genomics, Microsoft Genomics and Amazon Web Services (AWS) Genomics In The Cloud allow researchers to store, process, explore, and share large and complex data sets. The idea behind is to provide userfriendly tools to the researchers.

But finally it is no only for researchers, there will be one day that will be applied by clinicians. The whole article worths to be read.

Lita Cabellut. Barcelona exhibition