Medical Innovation and Reform in the US
When we debate health care reform in the US there is often considered to be a tradeoff between the pace of innovation and the healthcare system we create. Proponents of the current system emphasize that changing the US system towards a more universal or government sponsored one will dampen the pace of technological progress.
At the extreme this is probably true just as we know at the extreme that socialism leads to less prosperous societies than capitalism, but in the US we are often not talking about changes that extreme.
Instead we are talking about smaller scale adjustments in moving towards a model more similar to others in the developed world that is better able to provide quality care at an affordable cost.
Often the system we want to see in reality and the tensions it causes are a result of our values as opposed to what would be most effective from a policy standpoint. Atul Gawande makes exactly this point in an excellent essay titled: Is Health Care a Right?:
“…And we disagree profoundly about where we want to go. Do we want a single, nationwide payer of care (Medicare for all), each state to have its own payer of care (Medicaid for all), a nationwide marketplace where we all choose among a selection of health plans (Healthcare.gov for all), or personal accounts that we can use to pay directly for health care (Health Savings Accounts for all)? Any of these can work. Each has been made to work universally somewhere in the world. They all have their supporters and their opponents…”
To emphasize again there are many potential models that could work and almost all would be better than the current system on measurements of quality, cost, and access.
The ACA, for all its press as a major medical reform, was at the end of the day a law that built upon the existing healthcare system, nudging it closer to a system that expanded quality, lowered cost, and increased access. Its main impact was especially around increasing access, providing an insurance market to those left out of previous healthcare access points like employment, Medicare, and Medicaid.
Because the ACA mostly tried to tackle access to coverage and kept the existing system intact, we should not expect it to dramatically transform the healthcare landscape and affect things like the pace of technological innovation. Other things like the FDA, propensity of various health issues in the US population, and the intellectual property system have much more of an influence on healthcare innovation.
When considering how to maximize progress subject to engineering a healthcare system that attains high quality, affordable cost, and access to all, we should recognize that the system itself is not the only thing that matters and how the incentives that drive progress are operationalized often matter more. If we can understand that, then over time through experimenting with new policies, we can design a healthcare system that offers more of everything.
We can see exactly the above points by looking at both historical data on patent grants by country over time and recent research on encouraging innovation in healthcare.
While early the ACA has not dramatically affected patents granted in four healthcare related areas of biological materials, medical technology, biotechnology, and pharmaceuticals.
When we look at the share of patents granted to the US over time (solid lines in the plot below) and focus on the trend, we see that either it has remained relatively flat or has been declining for a sustained period starting prior to 2010 when the ACA was passed, meaning there is no real evidence of a difference post 2010.
In fact the decline in share of patents for the US seems to start when the share of patents granted to China in each area started to rise (dotted line), suggesting that the real drop in the US share of total patents is because of China’s rise as a medical innovator.
We can see this more clearly looking at the absolute number of patents granted in the US in these four areas over time.
We see that the pace of innovation is almost linear or in the case of medical technology and biotechnology appears to have clearly increased post 2010. Again it seems the ACA itself has not impacted the pace of innovation as defined by the number of patents granted.
Admittedly though, patents granted is a crude way to measure technological progress. Instead when considering technology in healthcare, we need to consider two commonly unasked questions:
(1) Do we have the right type of innovation?
(2) Does the existing innovation system set us up for further improvements?
On both of these questions, under the current system, the answer is often no, and the reason’s are both embedded in our system but also swayed by how we engineered other policies that can have as large an effect as the actual healthcare system.
Do we have the right type of innovation?
This idea is a bit subjective and hard to judge abstract consideration of our values, but we can get at it by comparing two extremes of investing 100% of our resources into developing a cure for something innocuous like baldness vs. 100% of our resources to curing cancer. It is clear that for most of us our values would prefer investing 100% of our resources into curing cancer because cancer is a more harmful condition than baldness and kills hundreds of thousands in the US every year, and we see medical innovation as a way to help those in need live better lives.
While the above exercise is silly given the resources we know are spent in cancer research, it actually hues much closer to a real example where we tend to see much more investment in curing baldness than in developing a vaccine for malaria, which causes hundreds of thousands of deaths a year around the world, concentrated among children.
The key driver of this disparity is the ability to pay that drives allocation of resources in free markets. Baldness is much more a direct concern in the developed world than malaria, so it offers a greater opportunity for profit than a malaria vaccine.
It is pretty easy to see then how free markets can incentivize the wrong kind of healthcare innovation that is discordant with our values as a society or even the needs of the majority of the population.
Another example is pharmaceutical companies focusing on expensive new cancer treatments that can prolong life modestly for another few months over more fundamental breakthroughs because of the expected costs and risks in those larger developments compared to minor optimizations that can lead to high priced cancer drugs that wealthy individuals can afford through quality insurance.
Hence, expanding access through governmental subsidies or having the government provide insurance for all can actually lead to more innovation that aligns closer to the needs of the majority of the population because we are empowering more consumers who may have previously lacked the ability to signal their demand through purchasing healthcare.
One point to consider along this dimension is that the provision of drug treatments is quite low in the US, which exacerbates the current Opioid epidemic. A lot of the people affected by this tend to have lower education levels, which means they also likely have a lower income. Since many might not be able to afford adequate healthcare, there is little demand for providing these services, but under a healthcare model where these people had the ability to purchase services, there might be more of a profit incentive for providers to produce these treatments.
There are also ways in which policies affect the type of innovation that we see because deciding whether to innovate for a healthcare company is a question of profit, which involves both the expected revenue, driven by demand concerns discussed above, and expected cost, which in turn is often affected by the types of policies we have in place.
In the pharmaceutical sector for example, a major source of costs in drug development are FDA trials. Therefore, the type of policy we have at the FDA can have a major influence on the type and amount of innovation we see in that area.
This article from 538, in discussing a paper by economist Andrew Lo, discusses how the requirements for drug approval are fixed across diseases and conditions and why that is inefficient. We can easily imagine if we were to adapt these FDA trial requirements based on condition severity, availability of quality, alternative options, cost of development, etc, then we would see a dramatically different set of drugs make it to market as we could lower costs for drugs where there are limited options available or that are expected to have a larger clinical effect to counteract the cost of development considerations.
With access where it is we clearly create an incentive to produce certain types of healthcare over others and policies like the FDA approval process also dramatically drive this as well.
Both the system and policies that operationalize innovation are key, meaning that well designed operational policies could drive innovation in ways that a particular system limits, lowering the actual tradeoff any healthcare system would create relative to the status quo.
Does the existing innovation system set us up for further improvements?
Technological progress in any field, but especially healthcare, is not a short run game. It plays out over the long run as previous discoveries serve as inputs to future discoveries. Open access to data and information is crucial in this process because researchers need to figure out what innovations to build upon and what worked and did not work to avoid wasting resources.
There has been recent empirical work by Heidi Williams at MIT with respect to the intellectual property system that highlights how it can dis-incentivize follow on innovation. Rather than the patent system itself appearing to have negative effects on future inventions (their causal estimates show little impact), this work and other papers in the area suggest that it is alternative, non-patent forms of IP that negatively affect the amount of follow on innovation.
Patents themselves though seem inferior to a policy around medical prizes. The idea here is that rather than reward a monopoly to companies creating new drugs or treatments, we provide a cash prize to cover the cost of development plus some return. The discovery is then made public to allow many companies to compete on producing and providing this product in the market. The cash prize incentivizes innovation and can be used to direct funding to underserved areas while the public information allows competitive market principles to take hold.
The kind of intellectual property system we have can and will determine the innovation growth path. Design a system reflects the best evidence we have on how to incentivize follow on innovation through balancing the profit incentive with access to information on discoveries is crucial.
We need to stop thinking of the healthcare system as solely determining the type and amount of innovation we get out. Universal healthcare or even government provided healthcare does not mean necessarily that we would see less innovation and conversely a completely free market system does not mean we would see both the right type and amount of innovation.
We need to work to first design a system that balances cost, quality, and access and then supplement that system with other policies around how we regulate innovation, grant intellectual property rights, and allocate funding that can work within it to maximize the amount and kinds of technological progress. The interaction of these policies with the system will ultimately determine whether we both (1) get the right kind of innovation and (2) set up the long term growth path of progress correctly.
Like many things in life, there are many options, each equally valid in their own way that would likely be a markable improvement over the current state of affairs.
Data Sources:
Data set on patent grants in US by year (originally from WIPO)
