This is not what Richard Thaler had in mind. Nor Daniel Kahneman. And certainly not Steven Levitt. The behavioural revolution was supposed to be impactful. It was definitely going to weave its way into our everyday lives. It would change the way we looked at and consumed food and how often we exercised and how we saved money. But it was not meant to save the world.

With the arrival of COVID-19, we have seen a slew of behavioural health ‘nudges’ emanating from hundreds of countries around the world that are, literally, a matter of life and death.

Stay at home. Wash your hands. Sneeze into your elbow. Practice social distancing. Shelter in place. Don’t shake hands. Don’t hoard food or medicine.

There is no question that these are the right principles for all countries to be adopting. The important question is how we make them ‘sticky’ without mandating citizens to do them. Without rolling tanks into the streets and imposing curfews and martial law. After all, when the government forces you to do something, this is no longer a behavioural nudge. And underpinning this large question around ‘stickiness’ are some of the smaller more nuanced questions that we need to consider: does expected utility theory still hold true in the case of a global pandemic? In other words, do decisions made under conditions of uncertainty about expected outcomes (becoming infected) and the hope that people behave rationally still apply when everyone has the same uncertainty? Or maybe everyone doesn’t have the same uncertainty about expected outcomes. Maybe some people know more than others. In which case, how does information asymmetry affect this pandemic and people’s behaviour?

You may recall that a few months ago (in PME, October 2019) I wrote a column about externalities and their importance in the context of public health. The notion that your behaviour has an impact on my health. In that column we talked about negative externalities (second-hand smoke) and positive externalities (vaccination or immunisation providing herd immunity). Today this concept, intertwined with the larger idea of ‘nudges’, is truer than it has ever been in modern times. Everything that everyone does has an impact on everyone’s health. Externalities are the new normal.

And what of another favourite topic of mine, health literacy? Telling people to ‘shelter in place’ is all fine and well. But do people know what this means? Does everyone have the same understanding of this concept? And, as a ‘nudge’, how do we actually accomplish this? Why is this important? Because research shows that people who are better informed about their health status are more likely to have better outcomes. In fact, let me be far more precise: differences in health literacy levels have been consistently associated with increased hospitalisations, greater emergency care use, lower use of mammography, lower receipt of influenza vaccine, poorer ability to demonstrate taking medications appropriately, poorer ability to interpret labels and health messages and, among seniors, poorer overall health status and higher mortality¹. In this current COVID-19 environment, with the exception of mammography use, all of these associations between health literacy and outcomes are strikingly important.

Another long-standing issue that I have championed for years is also critical at this time – the use of Dr Google to search for online health information. Many readers may be familiar with my views on the subject and, specifically, how the practice of seeking health information online can adversely affect access to care, treatment decisions and allocation of scarce resources. The COVID-19 pandemic has heightened my concerns in this regard with a veritable deluge of misinformation being spewed forth on a daily basis (particularly as it relates to treatment options for combatting this virus). How do we counter some of the dangerous and non-scientific data that is being disseminated? And, more importantly, how can we nudge tech companies to limit misinformation which is proving to have found an incredibly fertile host on the internet?

With millions expected to be infected and, unfortunately, tens of thousands expected to succumb to this terrible virus, the COVID-19 pandemic will change the way we conduct our lives forever. There is no doubt about this. The use of behavioural nudging and its incorporation into public policy is here. But, in its current format as a blunt instrument of direction to millions of people, are we using it the right way?

¹ Berkman ND, et al. Health Literacy Interventions and Outcomes: An Updated Systematic Review. Evidence Report/Technology Assessment No. 199. AHRQ Publication Number 11-E006. Rockville, MD. Agency for Healthcare Research and Quality. March 2011.

With COVID-19, society’s ability to adapt to behavioural measures is a matter of life and deaths

This article was originally published here.

Where to begin?

The mistakes started with the ignoring of the warnings that the virus was coming to our part of the world. And then there was the mask confusion. You don’t need one. Actually one might help. OK, they definitely help and should be mandatory.

The testing and the contact tracing was also a disaster. Not only in its conception, but also in its execution. Drive-through centres in every parking lot and phone-based apps that would inform us if we had been exposed which, in hindsight, seem like fairy tales from a more hopeful time.

The procurement and production of personal protective equipment and the lack of inventory in our collective national strategic stockpiles has laid bare the pure embarrassment of how we ignored warnings and did not invest in the most basic of resources. Still today we face shortages of items that were in shortage a year ago.

The rush to approve therapeutic agents without ample evidence. The travel bans and limitations on personal gatherings which were conveniently forgotten by the wealthy and the government officials who hosted maskless parties or who sunbathed on the beaches of faraway lands while we all hunkered down.

The decision to close schools but keep pubs open. The decision to build and allocate more critical care beds for COVID-19 patients but not to staff them with critical care nurses and critical care doctors which, even to the untrained eye, seemed completely obvious.

After all, without the trained personnel they are just beds. It’s the people managing the beds that make them critical care beds. Not the beds themselves. The late recognition and admission of airborne transmission of the virus which fundamentally altered our view of indoor congregation and may have helped accelerate the case load.

And there were other mistakes too. But one is exceptionally egregious: the failure to plan for the vaccine roll out.

Let’s be clear about the fact that we have known since January 2020 that a vaccine was in development. We have known that the two frontrunners would require mild-to-extreme refrigeration.

And we have known that there was going to be some vaccine hesitancy. We have also known that the necessary redundancy required in our global supply chain would be stress tested as we needed billions of needles, stoppers, glass vials, cartons and swabs.

We have known that distribution of vaccines to big cities and geographically remote areas would require public and private sector involvement. We predicted the need for a sophisticated vaccine tracking system so that we knew who got their first dose but not their second. Or both. Or none.

We needed this tracking system, not to satisfy airlines and hotels who insisted upon it as the minimum foundation for a ‘return to normal’, but to satisfy the most basic public health requirements, including the need for robust pharmacovigilance.

And as we sit here in the terrible, apocalyptic winter of 2021 that every epidemiologist predicted, with increasing case counts, hospitalisations, deaths and more virulent mutations circumnavigating the globe, we basically have none of what we knew we needed twelve months ago.

This is our great global tragedy. That with tens of millions of doses sitting in fridges in North America, only 40% of available doses have been used. The onslaught of the initial virus caught us all flat-footed. But there is no excuse for this.

We have to ramp up inoculations, prioritise people in the right order and notify them when it’s their turn and get them vaccinated. We must open stadiums and shopping centres and use every available resource at our disposal. We must give people paid time off to line up and get vaccinated. We must train more people to help deliver the vaccines.

The apologists will point to this as being the greatest logistical challenge in over a century. They will say: ‘But look at how much we’ve done already.’

They will defend the government and point to the fact that managing a herculean problem while planning for another herculean problem is too much to ask. But, of course, they will be wrong. For this is exactly what governments do and why they exist.

It is the ability of the government to manage the problems of today while keeping their eye on what is needed tomorrow that is their raison d’être.

Nobody is asking for miracles. Nobody is suggesting that there is a single magic bullet that can pull us out of this situation overnight.

All we are asking is that the people in charge do the things that have to be done, today and tomorrow, that they knew they would have to do 12 months ago.

The mistakes are plentiful, but one stands out as most egregious

This article was originally published here.

It has been six long months. Excruciating and depressing. Horrific and unfathomable. The Americas (North, South and Central) have been utterly gutted by COVID-19.

As of the middle of August, six of the top ten countries in total COVID-19 infections are in The Americas and they account for over 50% of worldwide infections as a group, despite representing less than 10% of the world’s population.

The easy answer would be to blame populist leaders like Trump, Bolsinaro and López Obrador for the mess that The Americas are in. But that would belie larger problems that are not of their making. The more complicated answer would be to examine the way out of this pandemic through regional cooperation and a Pan-American approach of scientific and economic collaboration with public health imperatives to help frame decisions.

Instead, as we are in the early stages of what appears to be a protracted and drawn-out fight with this virus, it might be more useful to look at what we have learned (in no particular order) with almost virtual certainty through the accumulation of scientific evidence.

  1. Social distancing works. Period. Full stop. As Solomon Hsiang who leads the Global Policy Laboratory at Berkeley said about social distancing in a recent article, “I don’t think any human endeavor has ever saved so many lives in such a short period of time.”
  2. Masks protect you from getting the disease and spreading it. We have not fully quantified all the nuances around viral loads that are shed from the mouth and nose, and absolute vs relative risk reductions in mask-wearing
    vs non-mask-wearing populations, but the evidence is clear. Everyone should wear masks when they leave their houses.
  3. The virus does not go away during ‘warmer’ weather. Ask people in Brazil and India and other equatorial nations. It just does not work that way.
  4. A vaccine may be ready sooner than we think. But it will not be for you and me. If we get a few hundred million doses by early 2021, they will likely need to go to first responders, the elderly, children and the immunocompromised along with those who are at high(est) risk. You and I will get vaccinated in 2022.
  5. This is not just ‘an old person disease’. The data is clear and unambiguous. Younger people get infected. Younger people get very sick. Younger people die.
  6. It will ‘come back’ again in the winter in Northern Hemisphere countries. Presuming it ‘left’ in the first place. Actually, this talk about a ‘second wave’ is framed incorrectly. The virus does not go away and come back. It is not a human being with access to public transport or a car. The better way to describe this phenomenon is to say that the virus will be worse in the winter months than it is in the summer months.
  7. Based on what we know, the chances of transmitting or getting this virus from surfaces is very low. If it makes you feel better to wipe down every surface and every item with a disinfectant wipe, go for it. In reality, it is an infinitesimally small mode of transmission and infection.
  8. Hydroxychloroquine does not work. Ingestion of disinfectants, hand sanitiser and light into the body do not work either. The only therapeutics that we have seen with some level of efficacy are remdesivir and dexamethasone. There are anecdotal reports of convalescent plasma and other treatment approaches using monoclonal antibodies which are promising too.
  9. Misinformation may be as dangerous as the virus itself. Countless groups have sprung up promising cures and antidotes to the virus. Other groups have propagated baseless theories about the virus not being that bad – just a flu. Or even worse: the invention of a cabal of left-leaning elites. And still others have taken up the cause of insisting that a vaccine, when ready, should not be taken by the masses for various senseless reasons. This has sown confusion and mistrust.
  10. We are tired. The virus is not.

The reader should note some obvious things about this list. Firstly, it is not exhaustive by any means. We have learned a lot more than I could possibly squeeze into my monthly column. That this is not only a respiratory illness. That we need to pay attention to ventilation.

That one single vaccine cannot possibly inoculate everyone and that we will likely need multiple vaccines. That asymptomatic spread is our worst nightmare. That this virus now appears to be airborne or aerosolised. Secondly, this list is not meant to be a retrospective analysis of what we did as a society to combat COVID-19. One day I hope to produce such a list. But for now, for this moment, these pearls of wisdom are about learning to adopt practices and attitudes that will get us through the next 12 months.

The Americas have been ravaged by COVID-19 – here are some important lessons we have learned

This article was originally published here.


We have always rewarded speed. The fastest car. The fastest man in the world. The fastest animal. The fastest person to swimthe English Channel. Everywhere you look, we reward speed. We have become addicted to ‘fast’. In fact, almost all public health officials will concede that in a pandemic situation it is crucial to act quickly.

But, sadly, with COVID-19, speed has killed. There is no better example than the United States (US), where the rush to ‘reopen’ the country has resulted in a new wave of infections that is threatening to blow the numbers from the initial months of this pandemic in the US out of the water.

The US has seen one record day after another, with new infections peaking at 45,000 per day recently. “I would not be surprised if we go up to 100,000 a day if this does not turn around,” Dr Anthony Fauci, Director of the National Institute of Allergy and Infectious Diseases, told a Senate hearing on the pandemic in late June. And some of those infections will turn into hospitalisations and deaths.

Another example comes to us from the world of peer-reviewed medical journals. On 22 May, The Lancet published a research paper about hydroxychloroquine, the antimalarial drug, as a potential treatment for COVID-19. In a wide-ranging interview with The Lancet’s editor, Richard Horton, the New Yorker notes that “unlike other studies, which had merely questioned the drug’s effectiveness, The Lancet article claimed that the use of hydroxychloroquine carried a greater risk of heart arrhythmia and death.

The paper’s stark conclusions and huge sample size – it purported to use data from 96,032 patients on six continents – halted hydroxychloroquine trials around the world. But, within days, reporters and public health experts noticed anomalies in the study’s data set, which was provided by Surgisphere, a small tech company outside Chicago. Thirteen days after the paper was published, The Lancet retracted it.

An hour later, The New England Journal of Medicine, the world’s other pre-eminent medical journal, also retracted a COVID-19 study that relied on Surgisphere data.” Did the journal rush to publish data without proper peer-review? According to Horton, stuff happens. He cited bad actors and fraud as the main culprits in this unfortunate scenario with the retracted paper.

In another instance where The Lancet published a review examining the potential effectiveness of masks and social distancing which has been denounced by some for its statistical ‘looseness’, Horton pushed back at the idea that speed is to blame. In today’s environment it is difficult to have conversations “about the nuances of work” when the world is waiting for information, he said.

The speed factor also comes into play in the procurement of ventilators and personal protective equipment needed by hospitals. Throwing all caution to the wind as many hospitals and healthcare providers competed with each other, the rush to procure equipment without properly understanding who these medical supply vendors were led to actual fraud in which money changed hands, with amounts ranging from hundreds to millions of dollars.

But, more importantly, by avoiding traditional procurement channels in which background checks and the normal steps of quality assurance are conventional, lives were lost.

Diagnostic and antibody testing has not been excluded from this obsession with velocity either. In a rush to get COVID-19 tests out to hospitals, the CDC ended up botching the situation which set the US back by weeks, if not months, in its ability to ramp-up testing. As Neev Patel outlined in MIT Technology Review: “On 5 February the CDC began to send out coronavirus test kits, but many of the kits were soon found to have faulty negative controls (what shows up when coronavirus is absent), caused by contaminated reagents. This was probably a side effect of a rushed job to put the kits together.”

And then there is the rush to develop therapeutics and vaccines to treat COVID-19. Experts have warned that the acceleration of vaccine development at its current pace is unheard of in the history of our planet. It took 28 years to develop both the Varicella and FluMist vaccines. And 15 years each for the Human papillomavirus and Rotavirus vaccines. Scientists are trying to develop a COVID-19vaccine in 18 months.

But as Robert van Exan, a cell biologist who has worked in the vaccine industry for decades, stated in a recent interview: “There’s a cost to moving so quickly, however. The potential COVID-19 vaccines now in the pipeline might be more likely to fail because of the swift march through the research phase.”

It is a tough balancing act indeed. This is all new to everyone. But COVID-19 has shown us that slowing down might be the fastest way to get back to normal.

From testing to reopening to vetting scientific articles, we have lost track of time

This article was originally published here.


I have written extensively and spoken about the slew of behavioural health ‘nudges’ emanating from hundreds of countries around the world that are, literally, a matter of life and death as the COVID-19 pandemic started descending upon the world.

Chief among these ‘nudges’ was the holy triad of quarantine, shelter-in-place and stayat-home. Take your pick, really. They all had the same basic intent and end goal in mind: do not go out and needlessly expose yourself or others to the virus. Unless you are a frontline worker, you should stay at home as much as possible. We must flatten the curve and keep the R0 (pronounced R naught) number, otherwise known as the reproduction number, from spiralling out of control. In the absence of a vaccine or larger population-level herd immunity, we need that R0 to be below one or, at the very least, at one if we have any chance of tamping down the spread of COVID-19.

And now, as we stare at the numbers of infected and dead scroll across our television screens on a nightly basis, we realise that there is a massive healthcare implication, outside the obvious direct impact of COVID-19, that comes with this stay-at-home nudge. People are afraid to leave their houses. And with good reason, to be sure. The disease is still ‘active’, is incredibly infectious and has a devastating effect on the most vulnerable in society, such as the elderly and immunocompromised.

But people are skipping much-needed medical appointments. Postponing previously cancelled elective procedures that have been rescheduled. Families are not vaccinating their children, as the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) have warned might happen. And the New York Times reports that ‘in a new study released by the Centers for Disease Control and Prevention, the vaccination rates in May for children under two years old in Michigan fell to alarming rates, including fewer than half of infants five months or younger.’ So, what the CDC and WHO said might happen, is actually happening.

As counterintuitive as it might sound, you might be risking your health by not risking your health.

The cataclysmic manifestation of the problem may lie in the effects that we see with respect to chronic disease management. The diabetic and hypertensive patients who are skipping check-ups and routine follow-up visits because  the mortality profile of this disease suggests that one of the groups that is most susceptible are those with underlying co-morbidities like metabolic syndrome and cardiovascular disease. Of course, the respiratory patients – those with chronic obstructive pulmonary disease, bronchitis and asthma – are in the same boat as the diabetics and hypertensives. The impact of this disease on the lungs is well-documented. The mental health impact of COVID-19 is clear as the pandemic weighs heavily on front-line healthcare workers. But there were millions and millions of people who had mental health conditions that required treatment and follow-up before COVID-19 hit. And then we have a whole ‘other’ group – patients with psoriasis, rheumatoid arthritis, Crohn’s disease, and also oncology patients.

And there is some evidence, albeit small, that points to the potential for a real problem in the next 12-18 months. A study published following the 2002-2004 SARS outbreak showed that chronic-care hospitalisations for diabetes dropped precipitously during the crisis but rebounded afterwards and, to no surprise, public health and health policy experts are worried that similar problems could crop up as a result of the COVID-19 pandemic.

This tsunami of chronic disease management patients who are missing regular and routine  care, the reticence to rebook previously cancelled elective surgeries and the alarming reductions in important childhood vaccinations are all exacerbated by two factors. First, we do not have a vaccine available and we probably will not for at least another 12 months when manufacturing and distribution times are factored into the equation. Second, there is widespread agreement that there will be a potentially calamitous return of this virus in the late fall and early winter in northern hemisphere countries.

All of these factors combined make those who are staying at home fearful of venturing anywhere near a healthcare facility (and, in some cases, other places too). We need to find a way, through the use of telehealth, the designation of specific hospitals as non-COVID-19 facilities, and urgent health communication strategies, to get those who are able and who are at low risk to continue with regular and routine care.

Otherwise, it seems, COVID-19 is not going to be our only problem.

You thought getting people to stay at home was hard – getting them to leave home might be harder

This article was originally published here.


It’s hard to take something away once you’ve given it to society. It’s even harder when you have no back-up plan. This is the undeniable lesson from the Obamacare debate that we’re witnessing.

At a macro level this is true of the entire Affordable Care Act unto itself as Republicans are finding out. It’s not easy to just make an entire piece of legislation that accounts for 20% of total gross domestic product (GDP) go away. But it’s also true of the individual elements and clauses within the act itself. Pre-existing conditions: don’t touch that! Children covered on their parents’ plan until the age of twenty-six: need it. Federally established essential health benefits and services: an absolute must. If Republicans are to succeed (or, for that matter, if any political party in any jurisdiction is to succeed in this type of bold attempt to overturn a previous administration’s legislation), the focus needs to be on ‘building upon and improving’ instead of ‘repealing and replacing’.

This is as much a lesson in communication and the subtlety of language and word choice as it is about astute health policy, navigating the corridors of power and working in a bipartisan manner. By articulating words like ‘repeal’ and ‘replace’, there is a palpable sense of a void and a hole. And with it a connotation that is inescapable: that repealing and replacing implies I must give something up. To be clear, the use of different words probably would have made a very minor impact in the whole debacle that was the Republican strategy. There still needed to be an actual ‘plan’ that made sense. And there wasn’t.

The behavioural sciences have taught us for decades that ‘taking things away’ elicits a very strong reaction in people. Loss aversion is what we call it. And empirical studies have shown that it’s about twice as powerful psychologically as the prospect of gaining something. And those same behavioural sciences have also taught us that language and words can be critically important drivers of behaviour. It’s bizarre that a president who has used psychology and communication as effectively, if not better, than anyone else who has occupied his seat missed this one.

Before Obamacare became the law of the land, we knew that loss aversion in healthcare was a political hot potato. A survey experiment in 2009, in the midst of the debate over healthcare reform, provides an illustration of loss aversion within the context of health insurance. Respondents to this survey were randomly assigned to one of two different scenarios and then asked to make a hypothetical choice between two healthcare plans (see above).

In both cases, the absence of a lifetime limit on health insurance benefits will cost the respondent $1,000 per year. In the first scenario, the cost will come by foregoing the savings of a plan with a lifetime limit. In the second scenario, the cost is directly tied to the lifetime limit. However, despite the equivalence, the different framing of the options (one emphasising ‘savings’ with the other focusing on ‘cost’) is critical. This example is illustrative of what we have always known (except for a few Republicans in the US apparently) which is that the framing of choices and the presentation of such choices as ‘costs’ vs ‘savings’ can greatly impact perceptions and behaviour in healthcare.

And so, in the end, we continue to regard our neighbours to the south as being caught in an echo chamber of sound bites emanating from two political parties whose views on healthcare are so fundamentally different. The lessons we learn do not stop with loss aversion. Social and public health policy does not afford us the luxury of thinking about the heterogeneity of healthcare. We cannot think of the ‘ones and twos’ in society. We need to take something heterogeneous and treat it as though it were homogeneous. This is hard. Every nation struggles with this harsh reality. And, to be fair, no single nation has managed to get it right. The outside view, however, is that our American neighbours struggle more deeply than others.       Perhaps it is due to partisan politics and political myopia.

Or perhaps it is due to ignoring the most basic lesson of them all: if you have your health, nothing else matters. If you don’t have your health, nothing else matters.

Healthcare’s most important behavioural lesson

This article was originally published here.

Happiness, or subjective well-being as some refer to it, has long been associated, on some level, with beneficial health outcomes at the individual level. 

There are a multitude of studies that have shown a relationship between an individual’s subjective well-being and some measure of improvement in overall morbidity or mortality. Simply put, we know that happier people have been reported to be healthier people.

But, then, why is everyone in healthcare so miserable?

I mean, patients are unhappy with the whole thing – prices, access, quality of care. I cannot recall a single chronic patient that I have ever met who has raved about the health system and how efficiently it operates. Occasionally, one will meet a patient who went through an acute scenario and was pleasantly surprised at how it turned out.

Caregivers are exhausted and equally disconsolate with the never-ending red tape of insurance paperwork and lengthy wait times as they shuttle their loved ones to and from appointments. Doctors are burnt out and despondent with a system that is nothing like they imagined. Ditto for nurses and pharmacists. Politicians are exasperated with spiralling costs and voter backlash at every turn.

Insurance companies and pharmacy benefits managers are crestfallen at the venom directed their way as they are blamed for the egregious costs being imposed on the system. Patient advocacy groups, while doing incredibly important work, often struggle for funding and to find a clear voice that resonates and can move the needle on meaningful policy change on behalf of their membership.

And manufacturers – both medical device and pharmaceutical – are no less miserable as they face a barrage of questions from lawmakers about their costs and pricing practices, and a slew of disapproving looks from the general public about their perceived greed and callous approach to patients’ lives.

In 2018, See and Yen published a paper showing that happier nations have better health system performance as measured by efficiency. They used the ‘happiness index’, which is a comprehensive indicator that includes
several important components, such as caring, freedom, generosity, honesty, health, income and good governance. And they measured efficiency as a function of life expectancy and inverse mortality rates.

The findings show that happiness is one of the factors that contributes to the efficiency of a country’s health system. Others have also published similar results on the topic of happiness and health.

And then there’s the United Nations 2017 World Happiness Report (see chart), which shows that among the top 20 happiest nations on the planet, a healthy life expectancy matters in the overall ‘happiness equation’ but accounts for less of a nation’s overall happiness than, perhaps, we think.

This doesn’t necessarily contradict any of the peer-reviewed work out there, as the empirical data is very clear that happiness is only one of the factors associated with health and health system efficiency. What the UN World Happiness Report reinforces is that the impact of happiness might be even less than we think.

Of course, there are methodological issues with all these papers and reports. How do we measure happiness? And how do we measure both health and health system efficiency? Is there an inherent selection bias in our data? Or other biases that we are not aware of?

But it is all confusing, isn’t it? If happiness is important at an individual level and a societal

level in contributing to population-level health as well as overall health system efficiency, why are so many people so unhappy with the state of healthcare. And if so many people are unhappy with the state of healthcare, how is it that we are producing this efficiency in certain countries?

Maybe, the answer is that other variables related to happiness (like income, education, housing, good government, security, religious freedom, etc) mask or overwhelm the low numbers associated with happiness from healthcare.

Or maybe there is some reverse causality happening – it’s not entirely that happiness leads to a healthier life and a more efficient health system but that a healthier life and a more efficient health system lead to happiness.

Regardless of these important nuances, I have not been able to find any stakeholder group that is truly ‘happy’ with the state of healthcare today. I’m betting you haven’t either.

How come I can’t find anyone who is ‘happy’ with healthcare?

This article was originally published here.

Suppose there are screening tests available for every single disease or ailment under the sun. Now suppose, for a moment, that you are screened for a broken arm. And let’s say that our screening test gives us a positive result. In other words, the test predicts that you have a broken arm. Now, let’s suppose that I proceed to order an x-ray and it confirms the positive result of our screening test and that you, indeed, do have a broken arm. This, in the parlance of epidemiology and, specifically, decision science theory, is what we call a ‘true positive’. You screened positive for a disease or ailment and you truly do have the disease or ailment.

Let’s look at the flip side. Suppose now that you are screened for a broken foot. And let’s say that our screening test gives us a negative result. In other words, the test predicts that you will not have a broken foot. Now, let’s suppose that I proceed to order an x-ray and it confirms the negative result of our screening test and that you, indeed, do not have a broken foot. This is what we call a ‘true negative’. You screened negative for a disease or ailment and you truly do not have the disease or ailment.

In both instances, there is something obvious that has occurred: the screening test was correct. In one scenario it said you had an ailment and you did. In another, it said you didn’t have an ailment and you didn’t.

And so, in the ‘arm’ scenario, since the test was correct, painkillers are prescribed and, perhaps, a cast or splint is placed on your arm and you go home. The costs incurred are in-line with the resources delivered. And in the ‘foot’ scenario, since the test was also correct, nothing is prescribed, no casts or splints are placed and the costs incurred are also in-line with the resources delivered.

But what happens when the test is wrong?

What happens when the test delivers a false positive? In other words, what happens when the test says you have a broken arm and you don’t. Well, let’s think about this for a second. If I’m treated for a broken arm and I don’t really have one, then the system incurs costs. I see the doctor for follow-up visits. I take time off work. I get a cast and painkillers when I don’t need them. Maybe I suffer from some adverse events due to the painkillers that I didn’t need. And maybe I suffer from some stress and anxiety.

Or a false negative? What happens when the test says that you don’t have a broken foot, but you really do? In this case, you may not incur splint and painkiller costs at the point of diagnosis, but you incur the cost of ongoing pain and suffering, the cost of this untreated morbidity, lost productivity, time off work potentially and, maybe, some downstream effect of not being properly diagnosed in the first place (ie, more advanced disease requiring now more expensive intervention that would not have been necessary had your disease been caught in the first place).

This is where healthcare gets really, really tricky. We tend to focus on the costs we incur in the patients who have disease. And we disregard costs for patients in whom we find no disease. But, the false positives and false negatives in healthcare drive costs that nobody accounts for. Ever. When a hospital gets its annual budget, the government doesn’t give it a few extra million dollars to treat patients who don’t really have disease. And nobody puts together a budget that has a buffer for taking care of patients who have disease that we didn’t realise was there in the first place. It just doesn’t work that way.

And, yes, not every disease or ailment falls into this neat little ‘false positive’ or ‘false negative’ category. It is definitely hard to quantify the extent of these costs—it is conceivable that they are a drop in the bucket. And perhaps, most germanely, there is no obvious way to eliminate the problem of false positives and false negatives.

Medicine is not perfect. It makes mistakes. And that’s ok. But it is worth pointing out that even these honest mistakes cost money.

Decision theory sheds light on the things we don’t often pay attention to in healthcare

This article was originally published here.


Suppose you put $100 in a savings account that earns 10% interest each year. After five years how much will you have?

That was a question posed in a multiple-choice quiz (completed by 150,000 people in 144 countries) by Standard & Poor’s, a rating agency. The answers proffered were ‘less than $150’, ‘exactly $150’ and ‘more than $150’. The intention was to test whether respondents understood compound interest, in addition to basic mathematics. According to The Economist‘s data team, the results were not that impressive with just one-third of respondents answering three out of five similar multiple-choice questions correctly.

The main thrust of this scenario is that it has been shown that it can be difficult to drum in financial know-how at a young age. Instead, it is gained through experience. Let’s think about that. Financial know-how which, without question, aids in financial literacy is gained through experience.

Health literacy is defined as ‘the degree to which individuals have the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions’. Now how do we suppose that people gain ‘health know-how’ which would, presumably, drive health literacy? I mean, if we rely on experience as one of the main drivers of literacy, you’d have to get sick in order to accumulate ‘health know-how’ wouldn’t you? You’d have to have experience with seeing doctors and taking therapies and wading through complicated private insurance plans that are massively confusing and looking up health information on the internet, etc. And since nobody knows when they’re going to get sick, the first time they do, they’re literally health illiterate. I mean aside from the annual physical and the odd needle here or x-ray there, people largely are in the dark about healthcare.


People who are better informed have better outcomes

Go ahead and ask people the questions in Figure 1 and see what kind of answers you get. Some might argue that these questions are not that easy. Maybe. Maybe not. Some would argue that these questions aren’t a standardised or even-close-to-widely-accepted measure of health literacy. Very true. But that’s not the point. The point is this: healthcare cannot be lumped into the basket of ‘other literacies’ because it’s fundamentally different from ‘other literacies’. It’s different because the idea that we can develop literacy by amassing know-how is difficult in healthcare. It’s different because compound interest of $100 after five years at 10% is always the same answer and cancer, diabetes and cardiovascular disease are not. It’s different because we have been trained to defer our property rights to clinicians and let them tell us what is right, important and required with respect to our own health.


If you have had a persistent fever, nausea and respiratory difficulty for more than 24 hours, you should: A) Call an ambulance
B) See your primary care doctor immediately
C) Drive yourself to the emergency department 

There is strong scientific evidence that routine childhood vaccinations may cause autism

True or False?
What is considered to be the ‘normal’ value for body temperature? A) Between 34 and 36 degrees celsius
B) Over 38 degrees celsiusC) Between 36  and 38 degrees celsius 

Figure 1. Sample health literacy questions

Why is all this important? Because research shows that people who are better informed about their health status are more likely to have better outcomes. In fact, let me be far more precise: differences in health literacy levels have been consistently associated with increased hospitalisations, greater emergency care use, lower use of mammography, lower receipt of influenza vaccine, poorer ability to demonstrate taking medications appropriately, poorer ability to interpret labels and health messages, and, among seniors, poorer overall health status and higher mortality. And it’s important because in today’s day and age of ‘information at my fingertips’, there are an increasing number of people who self-diagnose and self-treat based on information they’ve found online.

And because some estimates suggest that upwards of 40% of all the online health information is wrong, incomplete or purposefully deceitful. And because private health insurance plans are becoming increasingly complex as insurers look to shift spiralling costs to patients through the use of co-pays, pre-existing condition clauses and annual or lifetime caps on drug spending, which (you guessed it) require a high degree of literacy in order to ensure that the right plan is being selected. And because we hope and need ‘generational’ health literacy to take hold. We need older adults to start teaching younger people about the importance of being health literate.

Teach health earlier in school. Teach it more often. Incentivise patients. Incentivise providers. Enact laws around health literacy. Change the way search engines display search results on health queries so that paid results don’t jump to the top of the page. While it may be hard to quantify the effect of health literacy in terms of life expectancy, it is not hard to agree that we need to do things differently. Plain and simple.

The importance of health literacy

This article was originally published here

The pharmaceutical industry faces many headwinds as we enter the third decade of this century, the greatest of which may be the continual loss of branded medications due to expiring patents.

As companies face this ‘patent cliff’, there is increased pressure to develop drugs faster and more cheaply in order to continue to replace products that contribute to top-line revenue.

It is well known that the average drug costs somewhere between $2bn and $5.5bn to develop and that the total time to bring a new molecular entity (NME) to market can take upwards of 13 years with an exceptionally high failure rate.

With these facts in hand, the motivation for machine learning (ML) and artificial intelligence (AI) to address the issues of drug discovery speed, cost and lead candidate identification holds tremendous promise. But, in today’s world of ML and AI, there exist three fundamental ‘blind spots’ that should be addressed.

Firstly, many of today’s ML/AI algorithms are designed for simple, linear tasks. In a linear system, the relationship between the input and the output are proportional and rather easy to predict. If you smoke ten packs of cigarettes per day, you will likely suffer from emphysema or COPD or some other respiratory illness.

But the world is full of non-linear complexity. In these systems there exists no proportionality and no simple causality between the magnitude of responses and the strength of their stimuli: small changes can have striking and unanticipated effects, whereas large stimuli will not always lead to drastic changes in a system’s behaviour.

Non-linear systems often appear to be chaotic, unpredictable or counter-intuitive, and solving them requires tremendous application. To further illustrate this point, in mathematics and physical sciences, a non-linear system is a system in which the change of the output is not proportional to the change of the input. For example, the stock market. Or weather patterns. Or disease. A single word from the Governor of the Bank of England about fiscal policy can send the stock market into a spiral. Or a half-degree temperature change somewhere on the planet can have immeasurable effect for centuries to come. Or a single mutation, inversion or translation of a gene can result in devastating effects for human health. These are all examples of non-linear complexity.

And all examples why ML/AI need to be able to solve this problem. A second challenge with today’s ML/AI platforms is that they are dependent on ‘big data’. The cost and time involved with aggregating huge volumes of data and then ‘cleaning’ the data are not insignificant and may delay attempts at optimising drug discovery.

There are, however, platforms that can learn from as little as a few hundred data points with equivalent or superior insights to alternate ML/ AI predictive models and help pharmaceutical companies reduce failures significantly across the drug development cycle with their solutions.

These solutions uniquely predict the root cause for failure across the drug development cycle, including identification of specific patient subpopulation and placebo response. Finally, the issue of opacity is a troubling problem with some ML/AI solutions.

Known as ‘black box’ syndrome, this problem is exacerbated in industries, like healthcare, that are heavily regulated and where submitting a drug for regulatory approval requires a precise ability to show how a drug acts in a particular way to affect a particular disease pathway. It
is the ability to understand what the system is doing and the basis for its recommendations that are proving to be crucial.

Last year a team at Google used data on eye scans from over 125,000 patients to build an algorithm that could detect retinopathy, the number one cause of blindness in some parts of the world, with over 90% accuracy, on a par with board-certified ophthalmologists. These results had constraints; humans could not always fully comprehend why the models made the decisions they made.

Other such examples are also readily available. And, to be truthful, some are resisting these methods, calling for a complete ban on using ‘non-explainable algorithms’ in high-impact areas such as health because they may lead to forced, faulty or ‘unethical’ logic. Earlier this year, France’s minister of state for the digital sector flatly stated that any algorithm that cannot be explained should not be used.

So, there you have it. A promising technology, like many others before it, that has friction points that may prevent wider adoption. This isn’t the first time we’ve seen this, and it won’t be the last. But, as with all other such examples through the course of history, knowing what the pitfalls are and having our eyes wide open will help us unleash the true power of ML/AI in healthcare in the years to come.

What are the hurdles for machine learning and artificial intelligence in pharma?

This article was originally published here.