Some of the most important engineering lessons were demonstrated on the tank battlefields of World War II when German Tigers faced off against Soviet T-34s. The Tiger tank was a technical masterpiece of for its time with many features that did not appear in allied tanks until after the war. Despite its much heavier armor it was able to match the speed of lighter enemy tanks and keep up with its own light tank scouts. The armor featured almost artisanally welded interlocking plates. The ammunition featured innovative eletric trigger primers and high penetration tungsten shells. The double differential steering system allowed the Tiger to rotate in place. A complex system of interleaving wheels distributed weight evenly, improved offroad mobility and even allowed mobility with damaged tracks. But while the Tiger was a star on the blueprints, it was a disaster on the Eastern front, not because of its combat performance but because it was a logistical and operational nightmare. The heavy armor made the tank a gas guzzler, which made tanks inoperable when supplies were low. The electric trigger primers would fail in cold weather. When rotating in place the gearbox would often break and German training manuals forbid the maneuver. The highly specialized internal mechanics made production slow and meant the tank often could not be repaired in the field but had to be sent back to Germany, and the great logistic costs meant that Tigers couldn’t drive to the front but had to be brought there by rail.

The Russian attitude toward the T-34, on the other hand, was that making a tank a work of art was a waste of time if its purpose was move a gun into position just before being destroyed. Systems were designed as simple as possible to improve mass production speed. Armor was often welded with gaps between plates that you could stick your fingers into. To prove the tank’s mobility, the T-34’s designer Mikhail Koshkin drove the tank through snowstorms from Kharkov to Moscow and back. The tank lasted the journey, but he died from pneumonia contracted during the trip. When German armor outclassed the tank’s 76 mm gun, instead of designing unique solutions the Russian simply stuck an 85mm anti-aircraft gun into the turret. The Germans built 1347 Tigers; the Russians built 84,070 T-34s.

The Soviet success provides many valuable lessons for those trying to design and implement solutions for the current coronavirus outbreak. It is critical to understand the problem that you’re trying to solve, the resources that you have available, and how those resources can be applied most directly towards solving that problem. Unnecessary complexity leads to increased failure and causes problems in operation through low reliability and throughout the supply chain by creating material bottlenecks. Production capacity is necessary for overpowering the problem and simple minimum-viable-product design improves total system throughput and allows for mobilization of non-technical personnel. Materiel, personnel and time are all limited resources that cannot be used on discreationary objectives. But most importantly, it is necessary to be radically honest because you need to design for the problem that you have, not the problem that you want.

Already we’re seeing these principles being violated in the US response to the growing COVID-19 pandemic. Time and initiative are some of the most limited resources available in critical situations like this. Napoleon said ‘space we can recover, lost time never’. Despite this, leadership has been absentminded at all levels of governance, from the local to the federal, even to the global. Most officials appear to be stuck in a daze unable to take action until natural events force their hand. The comparison to Stalin hiding in shame from the public during the first days of the German invasion is impossible to resist. Even on March 9th the WHO is still making absurd statements that the threat of a pandemic “has become very real“, a phrase that can only still make sense to a bureaucrat waiting for a case in the notorious Madagascar before declaring that the pandemic is in fact not in our imaginations. In the meantime the CDC has demonstrated inability to provide testing kits at the required capacity, furnishing only 2000 testing kits in the same time that South Korea mobilized over 100,000 tests. In fact, the CDC made matters worse by preventing hospitals from developing their own tests and requiring them to route all testing through the CDC, creating a capacity bottleneck, increasing system complexity, and adding the travel time as an additional lead time between testing and results. Furthermore, the CDC increased complexity of their tests by using three primers, causing the initial batch of the kits to fail validation because of varying performance between the primers. In the meantime, the President’s bizarre and false ‘everybody who needs a test gets one’ cheer campaign is deafening.

With nearly two months of time since the initial Wuhan outbreak wasted, the US is finally scrambling to develop a strategy that it can communicate to the public. Much of this scramble involves an academic debate over how deadly or how infectious this virus is, as if the pundits are trying to find some economic optimization point at which simply doing nothing would be justified. Lost in the debate is the fact that this point does not exist because we know from in vivo experience in China and Italy that this coronavirus is both highly infectious and highly deadly, far above the rates of the flu that it still keeps being compared to. The frailty of the modeling exercise is that it depends on information to model off, which we know is dramatically incomplete due to the testing bottleneck, a conclusion that is validated by the constant stream of surprise cases that are now cropping up across the country.

The engineer’s task of developing a solution to the problem begins with understanding the nature of that problem. First, accept that this is highly infectious and highly deadly virus and that the hyperparameter tuning of the modelers is effectively irrelevant at this point. Two Italian physician posted extensive descriptions of the situation in the northern Italian lockdown. Their reflections are mandatory reading and corroborate what we know about what happened in Wuhan weeks earlier. Now that transmission is indigenous to the US, proper risk management requires us to assume that without intervention the same will happen within the next few weeks. The exact CFR of COVID-19 will be irrelevant to the overflowing emergency rooms, the shortages of respiratory ventilators, and the insufficient clinician staff. The second thing to understand is that right now there is no vaccine or cure against the infection. Normally the healthcare system provides curative interventions that directly treat the infection, but in this case the only known countermeasure is our own immune systems, which means that the only thing the healthcare system can do is provide stabilizing interventions to critical cases until the immune system can do its job. Finally, it’s necessary to understand the unusual virulence of this infection, which devastates the elderly and chronically ill while leaving the youth relatively unaffeted.

Combining these parameters, we are able to make a clearer formulation of the problem. The primary mission before the nation is to save as many elderly and chronically ill as possible by preventing the healthcare system’s resources from being overwhelmed. When accounting for the ease with which this infection spreads, the recorded autochthonous cases and the timeline over which the infection has shown to operate, it can be assumed that containment is impractical, if not impossible, at this stage. We are already within the critical window. In the words of Martin Sheen’s Robert E. Lee, “there is no time for that“. The primary objective therefore becomes not limiting the total number of cases, but limiting the rate at which new patients enter the healthcare system, and most importantly the ICUs, such that this rate does not exceed the rate at which people recover and leave the system. If the intake rate outstrips the exit rate system performance begins to quickly degrade as patients are unable to enter the required facilities, materiel shortages prevent delivery of appropriate treatment, and exhausted clinical staff begin to make errors that put patients and themselves in danger. The healthcare system cannot increase the rate at which patients immune systems will function and it has limited ability to effect the intake rate. It can and should redirect resources from discretionary procedures towards COVID-19 treatment, but this only increases the treatment-in-progress capacity and has no effect on the intake rates. Furthermore, the capacity of the most important resources like ICU beds is highly inflexible because these are more complicated systems than simply shoving more stretchers into a space. Recovery for critical COVID-19 cases has been shown to be 3-6 weeks and 2 weeks for even mild cases.  This is much longer than the average 3.3 day ICU (+5 day non-ICU recovery) or 4.5 day regular hospital length of stay. From Little’s Law we know that the patient load is proportional to the intake rate and recovery time, which means that an increase in the overall ICU intake rate by only 25% would double ICU utilization. It’s easy to see how the system capacity quickly becomes exhausted. This will be especially devastating for highly optimized systems like the NHS where the ICUs operate at 90% utilization and only a 2% increase over the base intake rate will result in all current capacity being exhausted.

Without the ability to affect rates directly, this means that the control intervention has to extend beyond the healthcare system and be applied at the level of the entire national community. It must be again emphasized that at this point there is insufficient information to determine the spread of this disease and there is still insufficient testing capacity to improve this information. There is therefore a high non-zero probability that the US is ALREADY past the point where intake rates would outstrip current healthcare capacity. Furthermore, delaying significant action increases this probability with every day. Because of the speed with which this infection progresses it is therefore necessary to implement dramatic interventions to minimize further spread until testing capacity can catch up and information on the morbidity is more complete. In short, total mobilization of the society is necessary. Half-hearted measures such as recommendations to stay out of public spaces or to wash your hands are not sufficient. These solutions are in fact highly complex because it requires the cooperation of millions of independent actors. Such complexity only creates risk of failure in a time when insufficient information requires a restriction on the degrees of freedom to the problem. The only way to ensure the risk of further spread is minimized is by halting all public events, travel, commercial activity etc. Right now it is necessary to buy our clinicians as much time as possible. Mass quarantine is a simplistic, dumb, even brutal, tactic but it fits the requirements of the problem and desired outcome. These restrictions can be lifted as the healthcare system’s utilization by COVID-19 is validated as sustainable and testing capacity improves our information and enables effective containment and quarantine, but delaying them until the last moment implies a vulgar acceptance of risk to American lives for the sake of preserving economic activity.

Healthcare Twitter has been focused this past week on the closure of the Hahnemann University Hospital (HUH) trying to rationalize the root causes of such a large system failure. (although perhaps it’s more appropriate that medical Twitter is focused on this subject, as the economists have been surprisingly silent).

Dr. Gaffney in particular wrote a piece in The Nation arguing along predictable ideological lines that the source of the problem is the for-profit organization of the healthcare system and that reorginizing the system around non-profit providers funded by a centralized Medicare-for-All single payer plan would prevent such closures. Notably, and crucially, missing from his analysis is that the the HUH came under for-profit private management as part of a restructuring of Allegheny Health Education and Research Foundation (AHERF), a large NON-PROFIT healthcare system in Pennsylvania that went bankrupt. This is not an unusual occurence. The non-profit St. Vincent’s Catholic Medical Center in New York was the main hospital for the downtown Manhattan community. This was the main hospital where 9/11 first responders were brought to. As a personal note this was also the hospital where my stepfather died. In 2005 St. Vincent’s declared bankruptcy and went through reorganization, but in 2010 declared bankruptcy again, shut down the hospital and sold off its various properties and outpatient facilities to other healthcare systems. The site sat fallow for a few years until the hospital building was demolished and an apartment building was built there. I think part of the building is also some kind of religious center now. The experience of AHERF and St. Vincent’s shows that being a non-profit is not a silver bullet solution to preventing such hospital closures. The St. Vincent’s case is also notable because unlike HUH, which was a safety-net hospital in a poor community, St. Vincent’s serviced downtown Manhattan, the district with the highest per capita income in the country. If a non-profit in a high income community couldn’t make it, who would? This doesn’t appear like something that can be solved simply by a single payer, something else is going on.

Dr. Gaffney’s critique of running a a safety-net hospital as a for-profit business has a logic to it, but as I have argued above simply eliminating the for-profit nature of the hospital does not eliminate the hospital closures we’re trying to prevent. To counter the Jacobin argument, I propose instead a Jacobite critique of capitalism. The fundamental problem with capitalism is not that it encourages profit-seeking, which is inherent to human nature, but that it separates the conduct of business from the social responsibility of the aristocracy (later bourgeoisie) to the local community that existed under feudalism and monarchism. While it’s important not to over-romanticize this noblesse oblige, it is nonetheless necessary to emphasize that social organization in pre-capitalist societies was dependent on responsibilities of the aristocracy and peasantry to each other and that these responsibilities were enshrined in the law of the time. The fundamental problem that Marx was highlighting in his critique of capitalism was the erosion of the aristocracy’s legal obligations to the peasantry as they simulatenously accumulated capital through new technologies and legal “reform”.

The primary mechanism for this deterioration of noblesse oblige is the accumulation of capital to the point where the capital becomes effectively international. As capital accumulates it reaches a point where investment in other communities becomes not only desirable, but even necessary. As the business expands outside of the community that gave birth to it, the original community becomes increasingly less critical to the operation of the business and therefore vulnerable to outsourcing as the capital is shifted between communities as part of various arbitrage opportunities. Just as capitalism causes the worker to be alienated from their labor, the aristocrat is alienated from their noblesse oblige by accumulating capital. This can be seen distinctly in the case of HUH. The non-profit AHERF that operated HUH until 1998 was a Pennsylvania-wide system based in Pittsburgh, not Philadephia. Tenet Healthcare acquired HUH from AHERF and is based in Texas. Paladin Healthcare acquired HUH from Tenet and is based in California. None of these organizations had a responsibility to Philadelphia as part of their organizational mission. Not even AHERF, which did not need HUH to continue existing as the rebranded Allegheny Health Network. Naturally HUH kept getting undermined by decisions made somewhere far away, usually in conflict with the interests of HUH’s Philadelphia patients. This failure of noblesse obligate to the community is also highlighted in the situation that HUH’s residents have been put in by the closure, left to frantically look for new facilities to host them. The situation in St. Vincent’s is related to this systemic degeneration of the aristocracy in a subtle way. Although St. Vincent’s had various outpatient clinics in other boroughs, it was inherently a local institution that only served the New York City community. But not its staff. Even though St. Vincent’s mission was focused on serving the community, that’s not necessarily true of its personnel. This is especially true for physicians who will often be uprooted from their own community to go somewhere else to university, then somewhere else to medical school, somewhere else again for residency, and somewhere else yet again to practice. As a less extreme example, my understanding is Dr. Gaffney was from New York, but practices in Boston. Although physicians are a functional aristocracy through their privilege and wealth, can we expect aristocratic loyalty from such ronins?  60% of a hospital’s expenses are labor costs, and even after its bankruptcy reorganization St. Vincent’s was not able to reduce its payroll. The hospital could have been saved if staff had reduced their salaries, but why should they if they have no loyalty to the community? St. Vincent’s closed and the staff dispersed to high paying jobs at the many other NYC hospitals (who are also slowly going into bankruptcy as they accumulate debt to pay competitive wages). Similarly, HUH staff could have saved HUH by taking paycuts. I wouldn’t expect them to take paycuts to generate profits for a private equity firm, but the point remains that they could have done it. These physicians will also find highly lucrative jobs in other hospitals.

A basic critique of for-profit schemes (or even non-profits behaving as for-profits) is insufficient to resolve the problem of hospital closures. The root cause of the issue is a much bigger society-wide problem where the public in general, and the functional aristocracy especially, are increasingly alienated from their local community and the responsibilities that bind community members together. Single payer insurance could make the situation even worse as the financing of healthcare is further removed from the local community to Washington DC. In addition, the St. Vincent example provides a haunting demonstration that having a community oriented organization may be insufficient to provide stability if clinicians are not themselves committed to that mission and are allowed to exploit the resources of the community and then leave once the system collapses. Hospital stability cannot be ensured until there is a more thorough social reevaluation of the fundamental responsibilities of the bourgeois aristocracy to the local communities that function as the source of their status and wealth.

I came across a revealing  post by Dr. Hans Duvefelt on KevinMD discussing the game theory problem created by health insurance where hidden cross-subsidies incentivize individuals to make healthcare decisions that benefit the individual tremendously but are inefficient for the group as a hole. The traditional tragedy of the commons problem. In particular Dr. Duvefelt focuses on an example presented by Dr. Siddhartha Mukherjee of ‘Emperor of All Maladies’ fame in an earlier NYT article, the CV drugs Brilinta and Plavix.

Brilinta, at $6.50 per pill, twice a day, reduces cardiovascular events more than generic Plavix (clopidogrel), which costs 50 cents per pill, once a day. But only by a little: a 20% relative or 2% absolute risk reduction. The event risk was 10% with the more expensive drug and 12% with the one that costs 82% less.

Put differently, if 100 patients were treated with Brilinta for a year, at a cost of $4,680 for each patient, 10 patients would still have an event. With clopidogrel, 100 patients, each one at a cost of $180, 12 events would occur. That means two fewer events would happen per 100 patients on Brilinta at an extra cost of $450,000, or $225,000 per avoided cardiovascular emergency (number needed to treat, NNT=50).

There’s an important error in this back of the napkin calculation because it forgets to factor in the savings created by the avoidance of a cardiovascular event. One study reports that the average initial and follow costs of a cardiovascular event in the US are $16,981 per case, ranging from $6,669 to $56,024. This does not include any non-medical costs that are incurred after a cardiovascular event, including lost productivity, personal utility, etc. Some may say that this is a pedantic criticism of back-of-the-napkin math, but it is necessary to emphasize that there is more to all of this than initially meet’s the eye.

Drs. Duvefelt and Mukherjee thus try to reconcile the ethical conflict between patient welfare and social welfare. What is to be done? Which drug should the physician prescribe? Mukherjee points out the argument that the physician has a responsibility to provide the best care available to the patient. It’s arguable that cost containment is in conflict with the Hippocratic Oath. For Mukherjee the physician can’t do anything, “the solutions are abstract and political”. Dr. Duvefelt points out that although Americans may dislike socialized medicine, ultimately all healthcare where somebody else pays for it, and therefore all health insurance, is by definition “social” and therefore all medicine is social medicine. In his opinion physicians have a responsibility to be stewards and spend resources wisely. But as I said earlier, there’s more to this than meets the eye. Notice anything missing in this entire discussion?

The patient. Both gentlemen are so focused on fixing society that they forget about the patient that they’re supposed to be helping and begin treating him or her as just another input into the medical industrial complex. Gauze, saline, Plavix, patient. The patient no longer has needs, preferences, goals or capacity, they are now merely the vehicle of a disease to be solved efficiently by the system. Social medicine therefore infantilizes the patient in two key ways. The first is that  it strips the patient of agency. Like children that have to do what their parents tell them, the patient now has to follow the decision made between the physician and insurer. What if my insurance covers the more expensive Brilinta, but I would like to make the socially responsible choice and go with Plavix? There are strange people like that out there after all; some even voluntarily pay more in taxes than they have to. What if my insurance only covers the cheaper Plavix, but I want to pay for Brilinta out of pocket? What do I work for after all? Some people work for a bigger house, others for a car, and perhaps I work for better care, even if it’s only 2% better. I don’t get to express my preferences because the system has decided for me. The second way that social medicine infantilizes the patient is by assuming that they are incapable of taking care of themselves. Children don’t have jobs and need to be financially supported by their parents. Similarly social medicine assumes that patients are financially helpless and need to be taken care of by the insurer. Perhaps that’s true for the $4,680 per year Brilanta, but why is that an assumption for the $180 per year Plavix? It’s insulting that the automatic assumption of the healthcare system is that the patient needs insurance to afford this generic. Frankly most people don’t need insurance for a $180 per year prescription, they need it for that $16,891 cardiovascular event we mentioned earlier. Mukherjee mentions the Ashish Jha study that I critiqued in my previous post to discuss the causes of high healthcare expenditure in the US: administrative waste, drug prices, procedure prices. Maybe expenditure wouldn’t be so high if we stopped using health insurance as a healthcare subscription and stopped hiding every single cost, including $180 per year generics, in the premiums?

Drs. Mukherjee’s and Duvefelt’s ask “what is a doctor supposed to do?” How about stop treating their patients like helpless children. Don’t withhold information from them, don’t make decisions for them, don’t assume their resources and circumstances. The patient doesn’t need you to be their parent and they certainly don’t need you to be an efficient steward of economic resources. There is no profound ethical dilemma to be resolved. You work neither for the government nor the insurer, as a physician you work solely for the patient. Explain to them all of their options, give them your recommendation, and let them decide what to do. That’s what a doctor is supposed to do; nothing more, nothing less.

Ashish Jha et al published a paper in this month’s JAMA comparing the US healthcare system to other countries. It’s a fairly robust effort comparing several system features, so the paper made its rounds through medical twitter, although I wouldn’t say that it’s any different from the Commonwealth Fund reports that get published in Health Affairs every year and from which the Jha paper draws heavily (more on this later). I’m not sure how familiar the medical community is with Health Affairs literature, since that’s a journal designed for administrative types that are the mortal nemesis of clinicians, so there may be some utility in publishing this in JAMA. The paper’s basic argument is that the US does not appear to be meaningfully different from the mean of a basket of comparable OECD countries on measures of utilization or social spending and therefore the cause of the high healthcare spending in the US is prices. This is an analysis that we have seen before, notably Uwe Reinhardt’s 2003 paper which reached the same conclusion. Although I do not disagree with the Jha conclusion that the U.S. prices for goods and services are elevated and need reform, the proposition that the U.S. isn’t structurally different from other systems is incorrect and needs to be put into context.

When I started my degree in health policy, it immediately struck me how the U.S. healthcare system does not function as a free market. Without prices set by competitive interaction between patients and providers the system suffers from the chronic information problem Hayek highlighted in The Fatal Conceit where any innovations are stuck in the hospital where they were developed and diffuse through slow channels like academic journals and conferences rather than the much faster market pressures. Even if a large system like Kaiser Permanente is doing something very unique and effective, the rest of the healthcare system may never find out about it because of the lack of signal mechanisms. My intuition was therefore to focus on comparative healthcare research and to look for international perspectives of how a healthcare system could be organized and coordinated. Perhaps we can identify best practices by figuring out what some countries are doing drastically differently to achieve drastically different results? The good news is that there is indeed a rich area to explore, especially in the developing world where healthcare systems can’t afford to waste money like we do (my quick look into Iran’s community health worker system is just one example). The bad news is that that comparative healthcare research is a very new area and the research methodology and data are often questionable. Although comparisons of the US healthcare system to other countries have been made many times in the past, such as this talk from the 1980s by Milton Friedman at the Mayo Clinic in which he compares us to the NHS, the first major thrust for comparative healthcare system research was made by the WHO with their 2000 health system report  in which they attempted to rank 192 countries in a variety of areas, ultimately nominating France as the best healthcare system. Although this pleased the French, the report was deeply criticzed by academic and politicians.

Data Problems

The major problem with the WHO report, and much of comparative healthcare literature in general, is that the data was unreliable. The report provided an 80% uncertainty interval for each country, which obviously leaves a lot of room for error that would be unacceptable in other healthcare literature. These intervals made it difficult to state with a large degree of certainty that any system really was better than another and resulted in large potential swings, with the US potentially ranking anywhere between 7th and 24th. In order to improve the data available and expand our understanding, the Commonwealth Fund began compiling comparative reports on the performance of several English-speaking countries, eventually expanding to include several other Western European and Nordic countries. The Jha paper relies on the Commonwealth Fund surveys for perception and access metrics. Does the Commonwealth data achieve their goal of providing high quality data? I respect the Commonwealth Fund tremendously and appreciate the difficulty of their task, but having worked with their survey I don’t believe that this goal is achieved for the US. The Commonwealth surveys have sample sizes of 1000-2000 per country, with the exception of Canada which generally has samples of 4000-5000 because the Canadian Ministry of Health contributed funds to expand their survey sample size. A sample of size of 1000 may be sufficient for small countries with homogeneous systems like the UK, Switzerland or Sweden, but a basic sanity check tells us that it would be inadequate for competently describing a diverse federal system like the United States, especially when you consider that New York is not Alaska and the 2000 sample size is spread out evenly across the 50 states. 40 responders don’t tell me much about healthcare in NY, especially considering how the differences between NYC and upstate NY are like night and day. In addition to the Commonwealth surveys, the Jha paper also relies on the 2016 OECD Health panel for utilization and spending data. I haven’t personally worked in depth with the OECD data, but my general attitude can be summarized by a gaffe from a former Ukrainian state economist, “the Ministry hasn’t published the report yet because we haven’t decided yet what we want the numbers to say.” The Jha paper itself admits that there’s problems with the OECD data: “When OECD data were not available for a given country or more accurate country-level estimates were available, country-specific data sources were used.” I propose that it’s irresponsible to make authoritative claims like “utilization rates in the United States were largely similar to those in other nations” when the data is compiled from a diverse set of sources of clearly varying reliability.

Analysis Problems

Jha et al suggest that the US system is unremarkable because US utilization for various services is similar to the average utilization of the eleven countries analyzed. For example, US discharges for mental and behavioral conditions is 679 per 100,000 population so the US is unremarkable because it’s close to the average discharge rate of 736. This analytical strategy is fundamentally flawed for two reasons.

The first is that the average behavior of a basket of very different systems effectively has no meaning because you cannot infer any conclusions on the behavior of any single system based on its relationship to the average. If you take the mental health discharges example, if I wanted to design an average healthcare system from scratch, the average of 736 per 100,000 population tells me nothing about how much staff I should hire, what facilities I should construct and which patients I should admit. Furthermore it would be incorrect to describe the US protocols as “typical” because these systems are clearly very different. The mental health discharge rate for Netherlands is 119 per 100,000 population, while the rate for Germany is 1719 per 100,000. These are not systems that are doing the same thing and marginally varying around a mean, there’s a difference by a factor of 14! The US can’t be a “typical” system because a “typical” system clearly can’t exist with such large variability. Either the Germans or the Dutch are doing something very wrong, because I highly doubt that they’re both right. And if they are somehow both right then the entire concept of the average becomes irrelevant because countries deliver care that is appropriate for their population and any nation’s position with respect to the average is irrelevant since the average is not representative of their proper patient population. Another example of this problem is hospital discharges per 1000, where Japan and US have similar rates of ~125 hospital discharges per 1000. Can we safely conclude that Japan and US have similar admission and discharge practices? That would be an impossible conclusion if you knew that Japanese hospitals are simply closed on evenings and weekends. These two systems will clearly behave very differently as a result of this hospital schedule even though they may appear to be similar when only looking at them on a graph.

The second problem with the use of the average of eleven countries as a benchmark to compare US performance to is that the average can be manipulated to serve your needs by adding or removing countries. Why does Jha use these specific countries? They state that it’s to use the wealthiest countries, but that’s a dubious proposition because the US is MUCH wealthier than the rest of the world. If Germany was a US state it would be the 40th poorest state in the union, France would be the 48th poorest. Are these really comparable? We have to therefore admit that we’re taking some liberties in how strict we are in our selection, and if we are going to do that then maybe the analysis should be expanded to the full 34 countries of the OECD? Alternatively, maybe we should be more strict with the selection criteria. Culture obviously has a large impact on population health, and Japan is one included country that’s clearly very different from the rest as reflected by Japan being the lowest on just about every utilization metric. But if we’re removing Japan, are the Germanic countries really that similar to the US? Perhaps it would be best to focus on the four English-speaking countries which will be the most similar culturally? But doing that changes all the averages. It shifts the COPD discharges per 100,000 population from 206 to 252 and suddenly the US utilization rate of 230 goes from the middle of the pack to the lowest utilizer! The point is that there is no such thing as objective data, all data has an opinion, and all comparative analysis of many countries is vulnerable to manipulation by adding or removing countries until you make the US look the way you want it to look. Using an average as a benchmark does not fix this critical flaw.

Knowledge Problems

Finally, a major challenge that I encountered while working on comparative system research data is simply figuring out the truth of what I’m looking at and how the data maps onto reality. Nassim Taleb calls it the pseudo expert problem, where you may have a lot of knowledge about a topic but still have no idea what’s really going on because you don’t have direct contact with the situation. In comparative system research the problem comes from trying to reduce complex systems to simple metrics that can be analyzed and compared, which unfortunately strips out critical qualitative information.

For example, Jha puts forward data suggesting that social spending in the US is similar to other countries, but doesn’t comment that this is only true after adding up public and private contributions to the programs that constitute “social spending”. That’s not exactly what most people would consider as “social” spending and after taking away private contributions the US quickly falls behind. Is it appropriate to combine private and public social spending because Americans are generally wealthier than Europeans and have more disposable income? I don’t know, but the issue is too complicated to be casually glanced over. One more example is hospital beds per 1000, which Jha reports as 2.8 in the US and 2.7 in Canada. These figures may lead a policy analyst to conclude that these systems have similar levels of hospital bed supply and similar problems, but this would be incorrect. Canada has a chronic shortage of hospital beds that does not exist in the US because Canadian law requires patients to be discharged directly to appropriate post-acute care. However, there is a shortage of post-acute care facilities in Canada, resulting in as many as 13% of Canadian hospital beds at any time being occupied by patients who no longer need acute care and are simply waiting to be discharged, sometimes for months. These beds are effectively virtual, resulting in blockages in other areas, such as EDs where it is not uncommon for Canadians to wait over 24 hours just to be admitted. Another example is primary care physician consultations for which Jha reports 4 visits per person per year in the US compared to an average of 6.6 visits per person per year. An analyst may even be impressed with Germany which reports just over 10 visits per person per year. This analysis, however, makes the error of assuming that all consultations are born equal and ignores the fact that American consultations are on average ~22 minutes long compared to only ~8 minutes in Germany. To make matters worse, we don’t even know if the American consultations are of higher quality, because the American doctor may be fiddling with the EHR for 20 minutes with only 2 minutes dedicated to the patient, while the German doctor may be able to dedicate the full 8 minutes to the patient. Baicker and Chandhra raise similar concerns regarding intensity, quality and measurement  in their response to Jha. We almost never know what the reality is!

Conclusion

This rant isn’t meant to discredit the work done by Jha et al, or the effort of the organizations responsible for the data they use in this paper. A lot of work obviously went into this report and it has definitely expanded people’s knowledge of the area, as evidenced by the fanfare it continues to receive on social media. I’m only seeking to express my frustration over the significant analytical errors that are present in most literature comparing healthcare systems. Healthcare is 18% of the economy and any errors have massive ramifications. My own view, to echo Hayek, is that these systems are far too complex to be properly analyzed en masse and instead it is best to focus on very specific topics ie rural nursing home care in Japan and United States. Anything more broad than that is almost certainly going to be wrong and I encourage researchers to hesitate before making firm statements like “it’s the prices stupid”. If we don’t respect the complexities of these systems, how could we possibly expect the New York Times to understand them?

That must be how the healthcare story of the week must have started. Amidst the drama about the State of the Union address and the FISA memo release came a tidal wave of coverage about the announcement made by the leaders of Amazon, Berkshire Hathaway and JP Morgan Chase that they will be collaborating on a new healthcare venture seeking to reduce the burden of healthcare costs. The subsequent media coverage was torn between promises of major disruption and foreboding of incontinent failure. What struck the most was not the announcement itself, but the absolute conviction with which these predictions were made despite there being so much unknown. Here’s what I mean.

The first thing that struck me is the simple question “why these guys?” Nobody seems to be asking this, but it’s critically important. It’s especially strange to have a venture between three different partners. Usually it’s a single company entering into the market, such as with Oscar insurance which also promised disruption. Otherwise it’s two firms cooperating, such as Apple’s venture into the data space while partnering with various healthcare provider networks around the country or the CVS/Aetna merger. But three firms getting together? That’s bizarre. Why not four? It would make perfect sense to throw in a pharmaceutical company into the mix. Why not two? It’s not immediately clear what critical skillset JP Morgan is bringing to the mix. Three? When was the last time you heard of a joint venture between three leading partners? It’s so strange that one has to wonder if they came up with the idea while drinking in a bar or playing golf. It’s probably the golf course, Warren Buffett is on record saying that he doesn’t drink.

The bizarre partner mix brings us to the strategy, which also doesn’t make much sense. Despite the deep conviction of the media analysis, the actual announcement didn’t include ANY specifics. The official statement says that the mission was “to address healthcare for their U.S. employees, with the aim of improving employee satisfaction and reducing costs.” This roughly means nothing, so all we’re left is looking at the core competencies of the companies involved. Much of the subsequent coverage focused on Amazon’s technological innovation and ability to disrupt markets. Is this even true? Does Amazon disrupt? Their Fire OS phones were an abysmal failure, which died off so fast that the media seems to have forgotten they even existed. Similarly their Anime Strike service was shuttered after only a year, while less notorious businesses like Crunchyroll and Funimation are operating just fine. Would people even pay for the Amazon Video service if it was unbundled from Amazon Prime? Amazon acquired Whole Foods but the results so far are mixed. The only business that Amazon ever truly disrupted was the retail market which was done through aggressive supply chain management and competitive pricing. But healthcare isn’t a retail market, it’s a services industry which is an entirely different game (read Baumol’s cost disease). The other business that Amazon has been successful with is Amazon Web Services, which provides an on-demand cloud computing service to individuals and businesses. Amazon could try to leverage this IT expertise, along with their AI projects, to better assess patient data, but this runs into the dual problem of legal restrictions on the usage of patient data and the fact that most patient data is junk (there’s a very good reason why MedPAC is rolling back on MIPS). Amazon also has a relationship with the Express Scripts PBM, but it’s hard to call that a core competence. Berkshire Hathaway is a conglomerate that runs a variety of businesses of which there are too many to describe individually, but my understanding is that they’re also involved with healthcare only tangentially. They don’t run any hospital networks, nor any biotech or pharmaceutical companies. Warren Buffett basically said fairly recently that pharma investment is too difficult to discern and Berkshire Hathaway at the time accordingly only had exposure to positions in vanilla players like Johnson & Johnson and GlaxoSmithKlein. They do provide health re-insurance through the General Re business, but that’s a somewhat different expertise compared to what a health insurer does. Finally we have JP Morgan Chase, which is a bank, they do bank things. Could they be replaced with any other large bank? Probably. So why JP Morgan Chase? Who knows, but I cite my golf course theory. Presumably they’re involved in the project to capitalize this new venture, but the announcement stated that this is going to be a non-profit, which is a hard sell to both shareholders and potential investors.

This brings me to the final thing that I noticed, which was the assumption by the media that this was some benevolent project that seeks to save American healthcare from persistent cost inflation. The joint statement clearly said that the goal was to provide solutions for “their employees”, which indicates that they would start locally before even considering expanding to America as a whole. This isn’t a charitable action, despite calling it a non-profit, but intended to benefit these companies and their shareholders. The question is what that benefit is. The thing that struck me most was the statement of intent to focus on providing “simplified, high-quality and transparent healthcare at a reasonable cost.” That could be a plausible benefit, but is this really something that people at these companies don’t have access to right now and are complaining about? All of these are big brands, one would assume that employees there have access to cadillac plans with low deductibles and broad access. This bares out in the Glassdoor pages, where Amazon’s insurance benefits’ rating of 4.1/5 is supported by comments like “Amazon has the best health insurance benefits I have ever seen.” The same is true for JP Morgan and Berkshire Hathaway. Employees at these companies seem satisfied with their benefits, so something is not right about this announcement. If the employees are not going to benefit, maybe the companies can? It could make sense to establish a non-profit insurer and then transfer the savings that would normally be the insurers markup to the shareholders of the three companies, therefore effectively transferring wealth from the insurer shareholders to parent company shareholders (note, not a charitable action), but this assumes that the new non-profit can operate more efficiently than the existing businesses. Contrary to popular belief insurers aren’t making enormous profits, publicly traded health insurers have a profit margin of only 3% on their revenue. Can Amazon et al negotiate more aggressively with pharmaceuticals and hospitals than their current insurer? It’s doubtful, they don’t have much market power. Although much of the media tried to position the 1 million US employees who work for these companies as a competitive advantage, this isn’t actually a large number. To put things into context the largest US insurer, United, has 70 million subscribers and the 5th largest, Cigna, has 15 million. Amazon+Berkshire+JP Morgan Health just doesn’t bring much weight to the table.

So what’s going on and what’s going to happen? The problem I have with this announcement is the amount that we don’t know and the amount that doesn’t make sense. There’s no company name. There’s no business plan. There’s no competitive advantages. There’s little relevant expertise. The bizarre partner mix makes no sense. All we have is a vague mission statement about improving quality and reducing cost for their employees, but even this doesn’t square out with what employees are saying on Glassdoor. If this announcement was brought in front of a venture capitalist by no-name Jeff, Jim, and John instead of Bezos, Buffett and Dimon they would have been thrown out of the room for wasting people’s time. By their own admission the business is in “early stages” but if you’re just starting out, why even make an an announcement? This is the one question I can’t reasonably answer, why even make an announcement? When nothing makes sense, your assumptions are wrong. This leads me to believe that this isn’t about making a business but all about sending a message. Obviously there’s no proof, just as none of the other theories have proof, but it seems the least incoherent answer. They’re signaling to somebody, most likely insurers but also possibly the government or phama, that this is something they consider to be a problem and are willing to exercise a nuclear option on (ie starting a dubious business) if the rest of the industry doesn’t shape up. It’s about having a strategic bargaining chip on the negotiating table, however credible or real it may (not) be.

After Jimmy Kimmel’s impassioned critique of the Graham-Cassidy repeal bill and the canonization of the Jimmy Kimmel Test for evaluating health policy,  celebrities are becoming increasingly active in the nation’s policy debates. I really wish they would stop, because they don’t know what they’re talking about, especially when it comes to healthcare, and the attention that they receive with their celebrity status makes it incredibly irresponsible for them to misinform the public. The newest entrant into the policy arena is Julia Louis-Dreyfus, famous for her roles on Seinfeld and Veep, who announced on Twitter that she has been diagnosed with breast cancer (get well, Julia).

The problem is that she then took the opportunity to plug universal health care as the solution to cancer mortality. Why is this a problem? Let’s take a quick look at how poorly the US compares to other OECD countries in breast cancer survival rates:

Oh. Well. That’s awkward. Despite not having universal health coverage, the United States is the best healthcare system for actually treating women with breast cancer and keeping them alive. That’s because contrary to popular belief, we do actually receive higher quality care than other countries. It’s important to point out that the US is a lot better than some key countries. The Commonwealth Fund’s darling “Best Healthcare System” in the United Kingdom is almost 9% behind the US in 5 year breast cancer survival rates. This is actually a massive problem, because as Julia Louis-Dreyfus correctly points out, 1 in 8 women will be diagnosed with breast cancer. That’s a full 1% of all British women who die because the NHS is failing them on this single important disease. Julia Louis-Dreyfus should study the facts before pushing such a system on American women as well.

By the way, since we’re sharing personal stories, my mother was diagnosed with breast cancer and as an uninsured poor immigrant went through chemo paying for treatment with loans. She lived. It’s certainly tough financially, but it’s not impossible, and the idea that being sick while uninsured is an automatic death sentence is yet another irresponsible falsehood.