How to stop ineffective and harmful medical practices

Ending Medical Reversal: Improving Outcomes, Saving Lives

What are medical reversals? We expect that medicine will progress in a generally orderly fashion, with good medical practices being replaced by better ones. We used to use cholestyramine—a horribly tolerated drug that had no effect on patients’ life expectancy—to lower cholesterol after heart attacks. Now we use atorvastatin, a well-tolerated drug backed by robust evidence that it saves lives. This is how medical practice should evolve. Reversal, however, is different. Reversal occurs when a currently accepted therapy is overturned, found to be no better than the therapy it replaced. This often occurs when a practice—a diagnostic tool, a medicine, a procedure, or a surgical technique—is adopted without a robust evidence base.

 Instead of the ideal, which is replacement of good medical practices by better ones, medical reversal occurs when a currently accepted therapy is overturned—found to be no better than the therapy it replaced. Now, you might argue that this is how science is supposed to proceed. In high school, we learned that the scientific method involves proposing a hypothesis and testing to see whether it is right. This is true. But what has happened in medicine is that the hypothesized treatment is often instituted in millions of people, and billions of dollars are spent, before adequate research is done. Not surprisingly, sometimes the research demonstrates that the hypothesis was incorrect and that the treatment, which is already being used, is ineffective or harmful.

So what?

Our medical system is too tolerant of unproven practices. Doctors are too comfortable recommending a practice without real knowledge of whether it is helping or hurting patients. People are too willing to accept practices that seem like they should help. When a medical reversal does occur, most physicians consider it an exception to the rule. 

We need a culture change in medicine. We need to recommit to evidence-based medicine and realize that it is the only rational way to provide care. In this book we have provided a few suggestions for ways we can improve. We do not advocate that these recommendations be immediately implemented but that they be carefully considered, alongside recommendations proposed by other thoughtful analysts, and tested in prospective trials. As we move forward, we must recognize that drastic and dramatic change can often be harmful. We acknowledge that there will be areas of medicine in which, for now, we must tolerate the status quo. As we go through the house of medicine and clean up each room, we have to prioritize.  

Well, let's say that the book focuses on the shadows of medicine, but this is only one part. Generalisations are inacurate. Anyway, good to review it. And medical education is not enough to solve the issue, incentives and culture play a crucial role.

The opportunity costs of excessive medical practice variations

 Atlas de utilización de procedimientos de dudoso valor. Actualización datos 2017

From the new report on practice variations:

La literatura científica abunda en estimaciones de la proporción de asistencia sanitaria cuyo valor para el paciente es cuando menos escaso. Este cuerpo de evidencia no ha hecho sino crecer en la última década, dando origen a varias iniciativas tanto académicas como gubernamentales para identificar y abordar lo que se considera uno de los principales problemas de los sistemas sanitarios modernos. Hay consenso: se trata de un fenómeno altamente prevalente que pone en cuestión el buen uso de los recursos sanitarios.
La actividad sanitaria de dudoso valor incluye tanto la utilización de procedimientos escasamente efectivos o para los que existen alternativas superiores, como el uso de intervenciones efectivas en indicaciones en las que los beneficios para el paciente son prácticamente nulos y en ocasiones incluso generan efectos negativos. Obviamente, para el sistema sanitario y la sociedad que destina los recursos necesarios, el coste oportunidad derivado de este tipo de actividad es sustancial.

So many years talking about it and nothing happens...

Great report, something should be done.
 Angulo-Pueyo E, Seral-Rodríguez M, Ridao-Lopez M, Estupiñán-Romero F, Martínez-Lizaga N, Comendeiro-Maaloe M, Ibañez-Beroiz B, Librero-López J, Millán-Ortuondo E, Peiró-Moreno S, Bernal-Delgado E, por el grupo Atlas VPM. Atlas de variaciones en la práctica médica en utilización de procedimientos de dudoso valor en el Sistema Nacional de Salud, 2017. Marzo 2020; Disponible en: www.atlasvpm.org/atlas/desinversion-2017

PS. Some books I'm waiting for.

Key Indicators for Health

Leading Health Indicators 2030: Advancing Health, Equity, and Well-Being

Hyper-personalized medicine is just starting

If DNA is like software, can we just fix the code?

 10 Breakthrough Technologies 2020

From technology Review:

Here is our annual list of technological advances that we believe will make a real difference in solving important problems. How do we pick? We avoid the one-off tricks, the overhyped new gadgets. Instead we look for those breakthroughs that will truly change how we live and work.

• Unhackable internet

• Hyper-personalized medicine

• Digital money

• Anti-aging drugs

• AI-discovered molecules

• Satellite mega-constellations

• Quantum supremacy

• Tiny AI

• Differential privacy

• Climate change attribution

What hyper-personalized medicine stands for?

Here’s a definition of a hopeless case: a child with a fatal disease so exceedingly rare that not only is there no treatment, there’s not even anyone in a lab coat studying it. “Too rare to care,” goes the saying.
That’s about to change, thanks to new classes of drugs that can be tailored to a person’s genes. If an extremely rare disease is caused by a specific DNA mistake—as several thousand are—there’s now at least a fighting chance for a genetic fix.
One such case is that of Mila Makovec, a little girl suffering from a devastating illness caused by a unique genetic mutation, who got a drug manufactured just for her. Her case made the New England Journal of Medicine in October, after doctors moved from a readout of her genetic error to a treatment in just a year. They called the drug milasen, after her.
The treatment hasn’t cured Mila. But it seems to have stabilized her condition: it has reduced her seizures, and she has begun to stand and walk with assistance.
Mila’s treatment was possible because creating a gene medicine has never been faster or had a better chance of working. The new medicines might take the form of gene replacement, gene editing, or antisense (the type Mila received), a sort of molecular eraser, which erases or fixes erroneous genetic messages. What the treatments have in common is that they can be programmed, in digital fashion and with digital speed, to correct or compensate for inherited diseases, letter for DNA letter.
How many stories like Mila’s are there? So far, just a handful.
But more are on the way. Where researchers would have once seen obstacles and said “I’m sorry,” they now see solutions in DNA and think maybe they can help.
The real challenge for “n-of-1” treatments (a reference to the number of people who get the drug) is that they defy just about every accepted notion of how pharmaceuticals should be developed, tested, and sold. Who will pay for these drugs when they help one person, but still take large teams to design and manufacture?
—Antonio Regalado

Allocating drugs by lottery

Health ministers condemn Novartis lottery for Zolgensma, the world’s most expensive drug

Novartis has held the first draw to choose four babies who will receive its one-shot treatment for the genetic disease spinal muscular atrophy, Zolgensma (onasemnogene abeparvovec), amid criticism of its lottery programme from patient groups and EU health ministers.
Priced in the United States at $2.1m (£1.6m; €1.9m), the most expensive drug course of treatment ever, Zolgensma is not yet approved elsewhere. In December the company announced a plan to give away 50 treatments in other countries over the next six months, the recipients to be chosen randomly from among applicants every two weeks.
Recipients must be under 2, the upper age limit for which the drug is approved in the US. Most of the children in the Zolgensma draw were registered by their doctors. About one child in every 8000 live births is born with spinal muscular atrophy. The most severe type, called type 1 or Werdnig-Hoffmann disease, usually causes death during early childhood if untreated.

Does this makes any sense? In my opinion is a perfect strategy (for Novartis) to create artificial  scarcity. It is a well known approach to increase willingness to access/ willingness to pay. It was described by Adam Brandenburger in a book long time ago: Coopetition.
I hope it will not succeed (at least in Europe).

David Hockney

Predictive modeling in health care (2)

Data-Driven Approaches for Health Care Machine Learning for Identifying High Utilizers

Predicting health outcomes using data modeling approaches is an emerging field that can reveal important insights into disproportionate spending patterns. This book presents data driven methods, especially machine learning, for understanding and approaching the high utilizers problem, using the example of a large public insurance program.

Five years ago I explained in this blog our experience on predictive modeling. This a key reference book.

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.

Repairing DNA: a review

The promise and challenge of therapeutic genome editing

Jenifer Doudna publishes a must read review article on genome editing in Nature this week. 

Current clinical trials using the CRISPR platform aim to improve chimeric antigen receptor (CAR) T cell effectiveness, treat sickle cell disease and other inherited blood disorders, and stop or reverse eye disease. In addition, clinical trials to use genome editing for degenerative diseases including for patients with muscular dystrophy are on the horizon.

 Notably, all of the genome-editing therapeutics under development aim to treat patients through somatic cell modification. These treatments are designed to affect only the individual who receives the treatment, reflecting the traditional approach to disease mitigation. However, genome editing offers the potential to correct disease causing mutations in the germline, which would introduce genetic changes that would be passed on to future generations.

 At the time of writing, international commissions convened by the World Health Organization (WHO) and by the US National Academy of Sciences and National Academy of Medicine, together with the Royal Society, are drafting detailed requirements for any potential future clinical use.

Meanwhile, CRISPR is closer than you think.

Fig. 1: Ex vivo and in vivo genome editing to treat human disease.

Fig. 2: The genome editing toolbox.

Fig. 3: Emerging tools.

Fig. 4: Editing the human germline.

Germline genome editing under scrutiny

Societal and Ethical Impacts of Germline Genome Editing: How Can We Secure Human Rights?

Geneva Statement on Heritable Human Genome Editing: The Need for Course Correction

A CRISPR Moratorium Isn't Enough: We Need a Boycott

The Human Right to Science and the Regulation of Human Germline Engineering

The last frontier in genome editing (if it exists) is germline. The special issue of The Crispr journal on bioethics contains an article of special interest and proposes a third process for evaluating individual and societal harms: a Human Rights Impact Assessment.

Human germline alteration is possible, due in part to democratization of genetic tools required for genome editing, and international scientific and legislative bodies are developing frameworks to manage the ramifications of this technology. Common among these frameworks are two pillars: public engagement and foundational principles. These components are necessary for respecting the autonomy of individuals and for fair processes and respecting diverse values.

However, they are not sufficient for protecting the most vulnerable members of society who may not even be in a position to participate in democratic processes. We propose implementing a HRIA, which captures concerns of public health and offers an opportunity to evaluate and anticipate the societal impact of GGE iteratively as the technology advances, public sentiments evolve, and cultural contexts shift. We recognize that this will raise new challenges of how such assessments are shared and implemented and how they can be enforced. We urge regulatory bodies and policy makers to consider this assessment approach in helping to establish robust regulatory frameworks necessary for the global protection of human rights.

And the Geneva Statement on Heritable Human Genome Editing says:

No decision about whether to pursue heritable human genome modification can be legitimate without broadly inclusive and substantively meaningful public engagement and empowerment. Such deliberations may be challenging and messy. They will take time and organizing them will necessitate creativity, hard work, and significant human and financial resources. The course correction proposed here is essential to these efforts.
We must in the meantime respect the predominant policy position against pursuing heritable human genome modification, if we are to prevent individual scientists or small committees from making this momentous decision for us all. This will preserve time to cultivate an informed and engaged public that can consider and discuss the societal consequences of altering the genes of future generations and make wise, democratic decisions about the shared future we aspire to build. 

I agree.

PS. CRISPR in 2020  Two major reports on germline editing, from the National Academies/Royal Society and the World Health Organization, will be released in 2020. We hope the reports will coordinate, with all the voices of CRISPR being heard, so we can build consensual and broadly acceptable frameworks to ensure we use CRISPR responsibly, especially regarding usage in human embryos for germline editing. The public has asked for it, and the community has been working on it. The science versus society gap will be bridged.

Digital health next to you

Bringing health care to the patient: An overview of the use of telemedicine in OECD countries 

Benchmarking deployment of eHealth among general practitioners

EHEALTH TREND BAROMETER: ANNUAL EUROPEAN EHEALTH SURVEY 2019

Several reports have been recently released. I would like to highlight the first one by the OECD, it reviews the current state of telemedicine and explains what works. In my opinion we do need an assessment of cost effectiveness of telemedicine, otherwise technology driven change is not enough.

Telemedicine services have the potential to improve effectiveness, efficiency and equity in health care, helping policy makers respond to increasing patient demands and needs. However, telemedicine interventions can also introduce new risks and amplify existing inequalities. In order for countries to maximise the benefits and limit the risks, telemedicine services need to improve the quality of care and provide clear benefits for patients. Telemedicine programmes that do not have benefits for patients are not worth pursuing and detract attention from other more effective interventions.

Josep Segú - Central Park

Health services research as a data science

Health Services Evaluation
Health Services Information: Key Concepts and Considerations in Building Episodes of Care from
Administrative Data
Assessing health systems

The provision of relevant, accurate, and timely performance information can play a pivotal role in ensuring the health system is able to deliver effective and efficient health services. Through its capacity to secure accountability in the health system, to determine appropriate treatment paths for patients, and to plan for future service patterns and structures, information can be used to identify and implement potential improvements in service delivery

AI in clinical decision making

International evaluation of an AI system for breast cancer screening

Artificial intelligence is capable of surpassing human experts in breast cancer prediction. You can check it in Nature:

 To assess its performance in the clinical setting, we curated a large representative dataset from the UK and a large enriched dataset from the USA. We show an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. We provide evidence of the ability of the system to generalize from the UK to the USA. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. 

If AI outperforms radiologists, than there is no argument to delay its implementation.
Powered by Google.

How to distort priority setting?

Biases distorting priority setting

Why priority setting in health care has so poor outcomes while relevant systems are well developed and readily available? You just have to read this article to understand it.

It starts to identify some rational and structural explanations for the discrepancy between theoretical efforts and practical outcomes in priority setting. However, even if these issues are addressed, practical priority setting may still not obtain its goals. This is because a wide range of irrational effects is hampering priority setting: biases.

Table 1. Overview of various biases and potential implications for priority setting and at which level they may be most prominent.

Bias	Potential Implications	Level
Identifiability and Singularity effect	Undermining principles	Micro
Rejection Dislike	Hamper disinvestment	Micro, meso
Failure Embarrassment effect	Overuse	Micro (primarily)
Prominence Effect (Opportunity Cost Neglect)	Non-warranted use	All
Status Quo Bias	Overuse, Underuse	All
Endowment Effect	Overuse, Underuse	Micro (primarily)
Loss Aversion	Hamper disinvestment	All
Aversion to Risk / Ambiguity	Overuse	Micro, meso
Availability Heuristics	Overuse	Micro, meso
Sacred Values and Taboo trade-offs	Undermining principles	All
Progress bias	Overuse	All
Adoption Addiction	Overuse	Micro, macro
Complexity bias	Undermining principles, overuse	Micro, meso, macro
Extension bias	Overuse	All
Asymmetry of risks and benefits	Overuse	Micro (primarily)
Positive cognitive feedback loops	Undermining principles, overuse	Micro, meso, macro
Prestige bias	Undermining principles, overuse	Micro (primarily)
Imperative of Action	Undermining principles, overuse	Micro (primarily)
Technology Placebo Effect	Undermining principles, overuse	Micro (primarily)
Imperative of Knowledge	Overuse	Micro
Competency Effect	Overuse	Micro
Multiple Replacements	Overuse	Micro
White Elephants	Overinvestment	Micro
Boys and Toys Effect	Unwarranted use	Micro