Re-Engineering Healthcare – a Strategic Data Primer
Summary
American healthcare is not broken. Rather, it is operating in a manner consistent with its design. What does that mean? In the current healthcare “system”, the efficiency and effectiveness of patient care services is not measured. Indeed, the effectiveness of services rendered is entirely unrelated to the apparatus that determines the payment made to a provider. If we collectively have the goal of establishing a more efficient and effective healthcare “system”, this fundamental design flaw must be fixed. As a practical matter, this is a data (informatics) problem. We don’t gather the data we need to fix this lack of alignment between the provider services, patient outcomes, and the payment made for the services delivered. Said simply, we must stop paying for stuff that doesn’t work well (or at all). We must start paying for the most cost-effective solution for any patient with any given diagnostic code. There is only one way to do this – start collecting and quantifying the data on what works.
It is correct to say that the current healthcare “system” isn’t a system at all.
Systems invoke the notion of many activities deliberately designed to produce a specific result. In America, healthcare is NOT DESIGNED to pay for those services that promote the patient’s health in the most cost-effective manner possible. This lack of purposefulness is the design flaw that must be fixed.
To align healthcare services with high-quality patient outcomes we must re-engineer the healthcare payment apparatus to identify quality outcomes and pay for them. To pay for high-quality outcomes we must first capture the data that enables the quality assessment to be made. That data must be collected in a timely manner and more than once. A one-time snapshot is insufficient to establish a quantitative journey that moves the patient toward a healthier state. The data used to make the assessment must also reflect a wide range of patient care practices. WHY? If the data set only reflects standard-of-care (conventional) practices, then a comparative cost-effectiveness assessment can never be made. WHY?
Two reasons; first, standard-of-care practices were never established based on comparative cost-effectiveness criteria data. Secondly, a data set that only reflects conventional patient care practices will have already caused simple and inexpensive practices to be excluded from the pool of activities from which an assessment is made. For example, the subject matter domains of nutrition and biophysical therapies offer many simple and inexpensive solutions for patients suffering from a multitude of conditions, including chronic pain. However, such services do not generate large cash flows to a provider. Thus, they are not used in conventional patient care practices. If they are not used, no data on the quality of patient outcomes can be collected and no comparative assessment made.
In the final analysis, a re-engineered payment system must incentivize the use of those patient care practices that produce a healthier state for the patient. That healthier state must be quantified. Engaging in this informatics re-engineering strategy, as it becomes robust across multiple clinical domains, will generate savings to governments, the US military, large employers who self-insure coverage for their employees, families and individuals that will be measured in trillions of dollars.
The fiscal instability of the United States is a natural result of this lack of purposefulness embedded in the design of the healthcare payment apparatus. Economically speaking, paying for services, whether they work well for the patient or not, literally incentivizes the epidemic of chronic disease to be managed rather than fixed. It is time that this historical haphazard arrangement comes to an end.
Background
In the current healthcare infrastructure, data is illiquid. It is partitioned in silos controlled by multiple influencers and numerous owners, all operating under economic incentives designed and managed with diverse, unaligned self-interest. Thus, the functionality built into the current design generates significant friction when data seeks to flow through, or across, domains. Consequently, no thread of purposefulness in service of a clear patient-centric objective is sown into the data function across all operational domains of healthcare.
When examining this issue, contrasting the state of illiquid data with the illiquidity of capital is instructive. While it can serve a temporal purpose, the illiquidity of capital discourages 1) beneficial exchange, 2) the building of a more diversified asset base, 3) the use of leverage for other productive purposes, 4) long-term investment, and perhaps most importantly, 5) the creative forces that would otherwise be unleashed when greater liquidity pre-disposes such forces to rational economic action. The same observations could be made regarding the illiquidity of data. In short, data illiquidity discourages diversification. It therefore violates a core pillar of natural systems: diversification leads to greater strength and vitality in the entire ecosystem.
It is well known that the innovation cycle in clinical medicine is about 17 years. That is how long it takes for an innovation to reach the clinical doctor patient relationship. By contrast, one of the great benefits of enabling data liquidity across multiple subject matter domains is that it will enable the development of insights that are otherwise opaque. Positive feedback loops that aggregate data across multiple disciplines and feed it back out to the network will accelerate the learning process in the entire clinical network. The transition from illiquid state to free-flowing state will accelerate the time it takes for clinical medicine to learn something new. Patients, employers, the government and physicians would be inspired by the results of incentivizing such a learning network.
Imagine for a moment, that such an informatics design accelerated learning and those providers who engaged in that learning were economically rewarded. If they produce (and quantify) health in their patients – they’ll get paid for the innovation. Imagine that provider cash flows made for managing a patient’s disease(s) were subject to competition from providers who could document that their activities generated a healthier state. Imagine if the data that illustrated how to generate health were visible to patient, and whomever pays their bills. The disease management model would realize a natural deterioration. Why? That’s easy, people want to be healthy, not sick.
Imagine going to a cardiologist’s office in the future and outside his/her office door is a sign that says, “93% of our patients were documented to improve their cardiovascular health, not just their symptoms, after one year of our services. The average cardiologist is at 32%.” If you needed a cardiologist, who would you use?
In short, we need to turn our clinical advisors into professional athletes. The performance of the latter is measured every year. A few excel at their work and their numbers prove it. They are paid more because the quality of their results is superior to others. There is nothing wrong with that.
Companies and governments that pay for healthcare services would rather pay for effective services than not. Economic rewards will accrue to those who align their capital allocations with those providers that produce health rather than those providers who continue to manage disease. Aligning the incentives to produce health will move capital allocation toward those most effective in operationalizing and documenting this skill.
Foundation for this Initiative
It is well known that the innovation cycle time of clinical medicine is 17 years. The declaration of a goal to cut this time frame in half would unleash a broad-based economic revitalization across many segments of the economy. This is a policy statement everyone could understand. Such a political target would encourage the development of data-gathering infrastructure, operations and analytics that can reduce the friction of moving information across the entire healthcare domain. This is especially true if patients are legally empowered to own their own data. Patient ownership of their own data will unleash tremendous economic activity that incentivizes the production of health in the population. In short, achieving the goal will transform the static data character of current operations to a dynamic one. The importance of this change cannot be emphasized enough.
Dynamic systems are empowered to learn. In addition to capturing significant savings and generating health in the population exposed to it, the economic vitality generated will garner significant support from the business and finance communities. This support will make the entire transformation process more politically tenable. The precedence for this expectation already exists in the private sector where LEAN and Economic Value Added (EVA) are well-established methodologies for creating dynamic data environments that generate value for all stakeholders. That value is created by the operational integrity of feedback loops that enable learning.
EVA was developed as a contrasting mechanism to the financial metrics created from the use of GAAP (Generally Accepted Accounting Principles) accounting, the regulatory framework used to standardize financial metrics. While GAAP was originally developed to support the interests of capital lenders, EVA was developed as a tool for identifying value creation for shareholders. This was done by taking GAAP accounting data and adjusting that data for items that camouflaged value creation. Similarly, applying a coherent systems approach (grounded on creating health) to the entire domain of clinical medicine will enable the quantification of value creation for all stakeholders. It will be a kind of advanced EVA-like lens that will guide the efficient allocation of healthcare capital to produce the greatest health benefit for the dollar spent.
In engineering terms, data functionality in healthcare must be re-designed to produce a higher signal-to-noise ratio. Instead of operationalizing this principle by re-writing the massive regulatory framework, establishing Demonstration Projects that document the above principles in a scalable, open architecture, and which serve the best interests of patients, would catalyze the strategic objectives addressed herein. Indeed, once operationalized, a project designed to weave a coherent thread across all subject matter domains would produce significant economic results within two years. The well-being of individuals, families, and communities will all be enhanced by operationalizing and rigorously documenting such an endeavor.
Said a bit differently…
Since the passage of the Obamacare legislation in 2010 ~$30+ trillion has been spent on chronic healthcare services – yet the transformation needle has barely moved. So long as the production of “health” is not defined and not measured, payment operations will be disconnected to the results produced from the expenditures made.
The data function in healthcare reflects a collection of workflow processes that are fundamentally misaligned as to purpose, operating with mismatched incentives and undocumented utility. This reflects the distinct self-interest of multiple parties operating within their respective domains. This characterization isn’t new. It was explicitly addressed in the Institute of Medicine’s 2001 tome on fixing American healthcare: Crossing the Quality Chasm. One of the primary conclusions of the Chasm report was that the existing system could not be tweaked. A purposeful patient-centric redesign was required.
Once you measure the right things, you can manage the right things.
In engineering, when a solution to a vexing problem hasn’t been identified, it’s typically because the problem isn’t properly understood. In effect, the failure to understand the problem camouflages all possible avenues for fixing it. In engineering, the persistence of this condition compels an examination of what is often called First Principles. These are the unspoken and oft-forgotten assumptions embedded within the current understanding of the problem. While engineers are trained to tackle this kind of examination, there’s no such corollary in clinical medicine. In the domain of engineering, understanding information is always contextual. For various reasons, however, biological sciences and clinical medicine have never adopted a contextual framework for processing information.
In the last century, some of the most profound breakthroughs in scientific understanding have come when the prevailing contextual framework shifted.
Einstein fundamentally altered the sensibility of “time”’ by proving it was not a static phenomenon. (Looking at a watch might have made you think it was). Rather, it was dynamic and related (amongst other things) to the speed at which the observer measuring it was traveling. When that door to a dynamic contextual framework opened, our understanding of how the universe worked changed. We are still in the early stages of exploring a similar example from the domain of cell biology.
When it was first identified the human genome was thought to be a kind of static blueprint for life. The prevailing notion at the time was that once we articulated the architecture of the genome in detail, we would be closer to understanding the cause of all diseases. The shift from a static to a dynamic blueprint generated a new understanding of how biological systems are designed to identify signals from the environment and change their biochemistry to optimize responsiveness to that signal. Soon enough, a recontextualization of genomics occurred and the discipline of epigenetics was born. Again, our understanding shifted from a static framework to a dynamic one.
In summary, both scientific history and economic vitality are on the side of seeking contextual shifts from static frameworks to dynamic ones. In the context of healthcare, this can be done by applying a ‘whole health/systems’ lens to patient care, documenting the quality of patient outcomes produced and contrasting those outcomes to current clinical standards of care. The resulting lexicon of patient care practices will be very desirable for patients, for physicians (those who began with a deep yearning to help others heal), and for governments that budget limited resources for healthcare services.
The future favors the bold.
Author: Wayne H. Miller (512 787 5028)