In a recent issue of JAMA, Daniel Morgan (U Maryland) and Deborah Korenstein (Icahn/Mt. Sinai) offered a viewpoint article that revisits the important evaluative conversation about the performance of our public health system during the COVID-19 pandemic and offers four useful principles for a “new approach to public health.” In this post, I will briefly summarize these principles and offer an important additional principle that I described in my comment published in JAMA this week.

Morgan and Korenstein noted that during COVID-19 there was an increase in distrust in public health institutions, with 65% of people having no trust or only partial trust in the CDC and 75% distrustful of their local public health departments. The authors admirably called for open discussion of “what went wrong” and “opportunities to build trust.” The authors implicitly kicked off that discussion by asserting that inconsistent, subjective and politically motivated guidance was among the causes of such distrust. They also implicated “unpopular aspects of public health” such as vaccine mandates, school and business closures and the “suppression of debate.” They proposed that trust-building opportunities should be based on four principles: (1) Do no harm, (2) Meet the public where they are, (3) Respect individual autonomy and a range of opinions, and (4) Be transparent and apolitical.

Before I propose a fifth principle, let me offer some commentary and quibbles about the authors’ first three.

During the final stretch in medical school, MD candidates in the U.S. reverently recite the Hippocratic Oath, which is commonly summarized as “first do no harm” (although the Latin phrase “primum non nocere” does not actually appear in the oath itself). Such a principle is of obvious importance in the context of ancient unethical healers who gained status and wealth offering theatrical treatments that were harmful but for which there was little or no scientific evidence of effectiveness — a problem that continues today with some “nutritional supplements” and, sadly, in parts of mainstream medical and surgical disciplines. But, in the context of the modern science-based practice of medicine, the “do no harm” principle is an unhelpful oversimplification. We obviously do not withhold chemotherapy from cancer patients just because it might have harmful side effects. And in the field of public health, we should not — as a matter of principle — forego quarantines and other mandatory policy options just because the expected outcomes include some harms to some individuals. The real principle should not be “do no harm.” It should be to balance benefits and harms. Or, even better, to carefully consider all relevant outcomes — including health benefits, health harms, economic benefits and economic harms.

In regard to the second principle, “meeting the public where they are,” I strongly agree with the authors’ admonition that public health officials should prioritize being physically present (rather than just remote) and that they should communicate through social media (rather than just TV). But the really important aspect of improved communication that the authors offer is in regard to the content of the communication. The authors correctly assert that public health officials must “present absolute risks and benefits” and “clarify the study type” to help the public understand the “level of trust to have in the data.” But, “risks” and “benefits” imply that only health outcomes are to be considered, while public health policies also influence economic outcomes. And, while communicating the study type is useful for experts, the public probably will not understand the implications of randomization, risk adjustment or other study methodology parameters. Therefore, what is needed to build trust and to support decisions by the public (and by policymakers authorized to represent the interests of the public) is information about the range of uncertainty around the outcome metrics presented. Therefore, I would prefer stating the principle as “explicitly estimate the magnitude and uncertainty of all relevant outcomes.”

Regarding the third principle, respecting “individual autonomy and range of opinions,” that seems like two separate things to me. The first is about dealing with the unavoidable reality that some of the outcomes from public health interventions accrue to individuals while other outcomes accrue to populations — the “public” in public health. For example, herd immunity is a herd-level outcome. Shutting down schools and businesses might affect the gross national product and unemployment rates as well as the crime rates and suicide rates that are causally downstream from economic desperation. Just as we should not say “do no harm” when we really mean “consider all outcomes,” we should not say “respect autonomy” when we really mean “consider both individual and population outcomes.”

Which brings me to my proposed fifth principle: science should stay in its lane.

Those who follow this blog might complain of repetition, since I’ve harped on this point frequently, such as in November, 2024 and November, 2021.

David Eddy, MD, PhD

In a famous series of articles in the early 1990s in JAMA, David Eddy described the “explicit method” for developing policies for clinical practices and public health. I summarize Eddy’s method as follows: All decisions inherently involve the selection among available alternatives. The process of supporting decision-making regarding public health policies and clinical practice guidelines rationally involves four steps: (1) identifying the alternatives, (2) identifying all the outcomes that are thought to be materially different across those alternatives, (3) estimating the magnitude and uncertainty of those outcomes for each alternative, and then (4) considering those estimates and applying values to determine which alternative offers the best set of outcomes.

Eddy’s explicit method called for doing those four steps explicitly — written down, providing transparency and accountability. When public health and medical scientists try to do all four steps implicitly inside our heads, using what Eddy disparaged as “global subjective judgment,” we step out of our science lane, applying our own values. We naturally want to make sure the epidemic does not get out of control and overwhelm the capacity of our health care facilities. We naturally prioritize deaths and other health outcomes directly connected to the outbreak, assigning lower value to health and economic outcomes indirectly associated with the disease or the public health policy responses. Morgan and Korenstein acknowledged this issue, noting:

“During the pandemic, initial guidance against community masking was intended in part to prevent mask shortages in health care, but the public was not informed of this rationale. A rapid switch to recommending masks eroded trust, as the public felt manipulated.”

This substitution of expert priorities for public priorities may be noticed by the public, contributing to a decline in trust in public health institutions, professionals and officials. This kernel of distrust can then be magnified by political actors that benefit from such distrust, leading to firings, budget cuts and derogatory messaging and the downward spiral of public health capability we observe today (as I described in my September 2025 post).

But there is an important difference between the way the authors frame this problem and the way I think it should be framed. The authors think the problem is just about the sharing of information by scientists when the scientists make public health decisions. Therefore, they propose that the solution is more “transparency” and “full honesty” and a commitment for the scientist decision-makers to be “apolitical.” But for more than thirty years, I’ve seen the problem and the solution through the lens of Eddy’s explicit method.

The first three of the steps in my summarized version of Eddy’s explicit method require scientific expertise in the fields of public health, medicine and perhaps health economics, since some of the relevant outcomes are economic outcomes. At the end of those three steps, there should be written-down documentation, including what Eddy calls a “balance sheet” with columns corresponding to the decision alternatives selected in step 1, rows corresponding to the outcomes considered in step 2, and the cells filled in with quantitative estimates of the magnitude and uncertainty of each outcome for each alternative, based on modeling done for step 3. I prefer to call that artifact a “decision outcomes table.” [Eddy’s PhD was in mathematics and not accounting, so he probably did not realize that “pro-forma income statement” is the more appropriate accounting artifact for summarizing relevant outcomes across decision alternatives.]

The forth step can then be described as the process of looking at the decision outcomes table that was produced in the first three steps and applying values and deciding which of the decision alternatives offers the best collection of outcomes. The key point is that the process of applying values and making a decision is more appropriately handled by people that are capable and authorized to represent the values of the public. In the case of consequential public health policies regarding such things as social isolation, vaccine mandates, and the closure of schools and businesses, it is local and state-level units of government and the associated elected officials that are the ultimate policymakers. So, when Morgan and Korenstein call for public health to be “apolitical,” I understand that to really mean that they want to avoid the entanglement of public health policymaking with partisan chaos. But consequential public health policies are unavoidably and appropriately political. When the authors assert that “public health officials must trust the public,” I would prefer to add “and the public officials they elect to represent them and make consequential decisions in their interest.” If those elected public officials cannot be trusted with those decisions, then perhaps we need to elect different public officials.

We will need to rebuild our public health system. When we do so, we can learn from the current evaluative conversation to try to build structures and processes that follow Eddy’s explicit method, where science builds up respect by staying in its lane.

This week, I had the privilege of participating as a reviewer of a master’s thesis (“major project”) for a candidate for a master’s degree in economics at the University of Windsor.  The candidate’s project was a decision-analytic model that explored a patient’s decision to seek care and to pursue second opinions and the physician’s decisions regarding diagnostic testing, implemented as a simulation written in Python.  I got a chance to attend the candidate’s defense of the work, which led to his successful attainment of his degree.  I am always encouraged when young people show interest in healthcare analytics and always inspired by the rituals of higher learning.

The experience got me thinking about our general challenges in the identification, recruitment and development of healthcare analytic talent, and the proven and promising strategies to meet those challenges.  This topic is important because I believe that one of the root causes of our slow progress in transformational improvements to our healthcare system is a lack of access by health care provider and payer organizations to analytic talent and the analytic outputs they can produce. 

In this post, I will emphasize three proven tactics and three promising tactics.  The proven tactics include (1) developing what I call a “community of analysts,” (2) establishing analytic career pathways, and (3) collaborating with universities.  The three promising tactics include (1) the use of clinical decision-analysis projects to foster learning, (2) the provision of what I call “seed papers” to enable such projects to be on target and successful, and (3) the establishment of open-source modeling infrastructure to increase the value and effectiveness of such projects over time.

Sources of Analytic Talent

From the perspective of health care delivery and payer organizations, useful analytic outputs can come from external sources, such as academic institutions, government agencies and consultancies.  As a consultant, I benefit when organizations look outside for analytic talent. But the most valuable analytic outputs are the ones that come from internal staff that know the organization and can pursue lines of inquiry and tell actionable analytic stories over longer periods of time.  Internal sources of analytic talent include centralized, specialized analytics departments, such as the Center for Clinical Effectiveness that I led at Henry Ford Health System in the 1990s, the Biostatistics and Clinical Epidemiology Department that I established at Blue Cross Blue Shield of Michigan in the 2000s and the Medical Economics and Claims Analytics Department that I helped establish at Trinity Health in the 2010s.  Or they can come from decentralized analytic staff positioned within various clinical operations departments, regions, and facilities. As I’ll describe below, both have important roles to play.

Causes of deficiency of supply of analytic talent

There are a number of reasons why analytic talent is so scarce in healthcare organizations.  The most actionable cause of scarcity in the short term is defective knowledge of leaders of health care organizations.  In most organizations, leaders do understand the importance of analytics, and they are acutely aware of their lack of access to good analytic outputs.  I can’t remember the last time a leader said to me “I’m happy with our data and our analytics.”   But many leaders think that the problem is that they just do not have the right data reporting application.  I’ve seen many organizations go through cycles of investment in new reporting applications, each time leading to frustration because the reports do not make sense due to problems with the source data or because the reports are not designed to support the specific decisions they have to make.   Waves of new technologies are touted as the solution to these problems — including “data warehousing,” “business intelligence,” “big data,” and now the biggest of all, “AI.”   Hope springs eternal that the next shiny widget will auto-magically solve the problem.  I’ve lamented about this problem in many previous posts, such as this one back in 2011: We don’t need more data. We need better story-telling. – Reward Health. As a result, many health care leaders seem reluctant to recognize that useful analytic outputs require talented insightful human analysts.  As a result, some leaders fail to invest enough time and budget to the recruitment and development of analytic talent within their organizations. 

Note that I am talking about analytic talent.  I use that word to emphasize the idea that there are innate cognitive capabilities for analytic thinking.  Training and experience are necessary and important, but as with opera singers (and, to my dismay, swimmers), so is natural talent.  The last thing we need is to increase the supply of people lacking analytic talent that believe themselves to be analysts and that get hired or engaged into analytic roles, occupying spots where talented analysts could otherwise be.  Although there can be value in developing easy-to-use, intuitive analytic technology to enable more people to participate productively in analytic work, such tools sometimes have the disadvantage of empowering people lacking analytic talent to masquerade as talented analysts, clogging up the lanes in which valuable analytic outputs are intended to flow.

In this context, I have observed that the leaders of many otherwise sophisticated healthcare organizations lack of ability to recognize analytic talent and to distinguish real talent from masqueraders.  As an employee or consultant to many health care organizations, I have spent lots of time over many years assisting organizations in assessing and recruiting talented analysts. Such work has always seemed to pay dividends.

Developing analytic talent through a community of analysts

As noted above, analysts within healthcare organizations can be positioned in centralized or decentralized settings.  In my experience, the most productive and impactful analysts are those that reside in the decentralized settings.  By working in or near the operational front lines of an organization, analysts gain the context necessary to understand the challenges and questions that are the subject of analyses. And they gain access to facilitate the transfer of analytic outputs to the front-line leaders most directly able to put the insights to good use. 

However, such decentralized settings are not ideal for the development of analytic talent.  Decentralized settings do not have sufficient budget to hire a substantial group of analysts, nor do they have sufficient demand to keep such a group productively engaged.  Therefore, decentralized settings tend to have one or two analysts. They try their best to figure out how to do things on their own.  It can be lonely work.  Such analysts can feel frustrated and stuck. 

One important strategy to overcome this problem is to intentionally establish and nurture a “community of analysts,” allowing people to learn from one another.  The organization can bring decentralized and centralized analysts together for retreats and lunch-and-learn sessions. They can provide tools to foster collaboration and sharing of ideas and practices.   I’ve built such a community of analysts in a number of institutions and advocated for its development in many other institutions for which I have provided consulting support.  Such initiatives can promote professional development and sharing among analysts positioned in both centralized and decentralized organizational settings.  

Developing analytic talent through career pathways

To achieve deeper development of the analytic talent within a healthcare organization, it is necessary to create career development pathways. The idea is to initially hire new analysts into centralized analytic departments, where they can work directly with more experienced analysts. Then, as they mature, such analysts are intentionally moved to various decentralized areas with occasional stints back in the centralized analytic areas.  Such career development pathways make the analysts feel valued, improving retention and morale.  And, in my experience, such career pathways can provide a large benefit to an organization’s culture by reducing the silos separating departments and creating strong professional networks that cross the “whitespaces” in the organizational chart.

Opening the University pipeline for analytic talent

The strategies described above help to make the most of the analytic talent already existing within healthcare organizations.  But the deeper cause of analytic talent scarcity in healthcare organizations is a lack of supply.  There is just not a sufficient population of talented analysts available to be hired or engaged by health care organizations.  To address deficiencies in the supply of talented analysts, we need to go upstream to identify more people with analytic talents in graduate or undergraduate university settings and then attract them to the health care field.

Health care organizations can start this process by taking more proactive steps to build relationships with university departments that produce analysts.  Analytic talent most often begins development among students studying social science disciplines such as economics, epidemiology, biostatistics, and perhaps other areas such as business management, management engineering, organizational psychology, and the analytics and “data science” sections of information sciences departments.  When I was at Blue Cross Blue Shield of Michigan, we set up meetings with the leaders of the University of Michigan School of Public Health to raise awareness that we had interesting job openings in our Epidemiology and Biostatistics Department and asking for their help to hook us up with the best students.  We sponsored a reception offering a big charcuterie tray and refreshments to have a chance to show a few slides describing our opportunities and to meet with students.  Most importantly, we sponsored some summer internships to build deeper relationships with students and to be in a position to more thoroughly assess their capabilities.

Through this process, we discovered that the most talented analysts, particularly those that had had both talent and knowledge of advanced analytic methods, wanted to work in a place that offered an intellectually stimulating community and an opportunity to make a difference.  When we hired analysts with advanced degrees, we made candidates give a presentation of their work — typically their major project or thesis.  We invited a bunch of our existing analysts to attend and discuss the work.  This process took a lot of time, particularly because we were very picky, often interviewing ten candidates before we hired one.  But it allowed us to discern real talent as well as the professional characteristics that made for good analytic collaborators and effective communicators of analytic outputs.  It also allowed the candidate to see how serious we were and to allow them to meet people and confirm they would enjoy the stimulating environment.  It provided learning opportunities for the existing analysts.  And each time we hired an awesomely talented analyst, it made our team better and it made it easier and easier to attract more talent.  Over time, we developed a core analytic team that included 13 PhD-level analysts and 10 masters-level analysts, dramatically improving the organization’s analytic effectiveness and enabling us to pursue value-based contracting innovations and care management innovations that would otherwise have been impossible, or at least very risky.

Using analytic projects to improve analytic education

Of course, most health care organizations do not have the size and budget to build such large and sophisticated analytic core teams.  As a result, it is necessary to go beyond just improving the recruitment pipeline and the internal talent development process. In the longer term, it is valuable to go farther upstream to work directly with universities to improve the analytic educational experiences within undergraduate and graduate programs.   I believe that the most promising such initiative would be to establish and develop programs that give more students in analytic disciplines more opportunities to participate in interesting, applied healthcare analytic projects. 

In this regard, we can learn from the talent development approaches that work in the physical and biological sciences.  Science fairs motivate very young students to participate in science projects.  Most science projects at elementary and secondary schools are nowhere near the cutting edge.   For example “does soapy water inhibit plant growth” or “demonstrating a volcanic eruption with vinegar, baking soda and red paint” do not generate new discoveries.  But such projects expose young students to the tools, jargon and conventions of science — developing a hypothesis, collecting and analyzing data, preparing a poster, and making a presentation.  Then, in science graduate programs, science students have an opportunity to work on actual cutting-edge projects.  Where do these project ideas come from?  Occasionally, student can think up their own original ideas.  But, more typically, good candidate projects are identified by the leaders of the labs in which the student-scientists work.  The leader of a laboratory typically has some core projects that they personally lead, and the students and post-docs do the dirty work, getting experience with the equipment and methods of the field.  The laboratory leader identifies various side projects, typically related to the main projects, that are suitable for students and post-docs.  Through some matchmaking, the laboratory leader helps the students and post-docs to select projects of interest and guides them through the process.

We can adapt some of those concepts to apply them to healthcare analytic talent development.  In social science and data science programs, students may be required to do papers at the undergraduate level, a major project at the master’s level, and a thesis in a doctoral program.  In my experience, faculty members in social sciences tend toward theoretical projects, since they lack exposure to real-world questions and challenges.  Such theoretical projects are helpful to build understanding of concepts and methods, but they tend to be less helpful in developing in students a passion for applied analytics and, more specifically into a passion for analytics focused on supporting healthcare process improvement.  Therefore, I think it would be beneficial for health care organizations to partner with social science departments to proactively define applied analytic projects that are “real” in the sense of relating to actual decision-making and design issues within healthcare organizations, but still suitable to serve as educational projects for students.

The ideal types of analytic student projects

Health care organizations have care processes and administrative processes, and the care processes can be divided into clinical decision-making processes (those that produce changes to care plans) and care delivery processes (those that execute care plans).   Analytic projects can also be classified as (a) retrospective data analysis and reporting, and (b) prospective modeling and simulation.    Putting these two dimensions together, we can define six types of analytic projects relevant to healthcare organizations, as shown below.

The table above shows some types of analyses that could serve as projects for students learning healthcare analytics.  Among mature, fully developed healthcare analysts, the most effective are those that are familiar with all of these types, as I noted in a post back in 2011: We need to cultivate more analysts that can do both rigorous science and practical decision-making – Reward Health. But that still leaves the question about which type of project is the optimal starting point for student analysts.

When most people think about analytic projects, their minds tend to conjure the top row: retrospective data analysis and reporting.  In such projects, a student is provided access to a database. They do queries and regression analysis and apply other retrospective analytic methods to characterize something or draw conclusions about potential cause-effect relationships.  Such projects can indeed by valuable.  But they also have some disadvantages in terms of student education.  Educational institutions often do not have access to proprietary and confidential health care data, so retrospective analytic projects tend to be done on generic public databases, such as datasets produced by government agencies.  Also, papers describing things and exploring cause-effect relationships do not always feel compelling to students.  They feel theoretical.  It is not obvious to the students how such projects make a difference. 

Therefore, I have a non-traditional proposal: the ideal student analytic projects are actually prospective modeling and simulation projects. 

Such projects  can be done under the control of the student, without substantial proprietary database access dependencies.  They directly relate to decisions that need to be made by actual healthcare organizations.  Among the three types of modeling and simulation projects, I propose that the ideal type is modeling and simulation projects to support clinical decision-making.  Such projects involve decision analysis, cost-effectiveness analysis, cost-benefit analysis, and population simulation.  They compellingly relate to the front-line decisions at the heart of the mission of a healthcare organization. 

The importance of seed papers

Just as elementary schools provide students with lists of ideas for science projects and faculty directors of research laboratories maintain lists of ancillary research projects suitable for graduate students and post-docs, it is very helpful to students learning analytic modelling to have access to a list of good candidate projects.  But, where do we find such projects, specifically for clinical decision-making modeling and simulation?  I propose that the best way to find such good candidate is to search for what I call “seed” papers.  A seed paper is a recently published paper that does not involve modeling and simulation, but that serves as a great starting point for a student modeling and simulation project.  The best seed papers for clinical decision-analytic projects are randomized clinical trials or retrospective observational studies (outcomes studies).  The fact that they were recently published suggests that they address a currently important topic.  Such seed papers provide a background section that frames the topic and the associated questions and that summarizes the current state of the literature on that topic.  Such a seed paper includes a great starting list of references for the student.  And, most importantly, such a seed paper includes results that will provide some of the inputs to a model or simulation.   They typically do not provide data on all the relevant outcomes needed to support the modeling of alternative candidate protocols for a particular clinical decision-making process.  But that is a good thing, because it gives the student an opportunity to get some experience in finding or collecting data or eliciting expert opinions to fill in the missing input values. 

The identification and matching of such seed papers can happen in the following ways:

1. The student can be provided with a list of journals and other search criteria to be used to cast a net for such papers, and directed to skim through many papers to find candidates of interest to them.
2. The faculty of the educational institution can keep a list of good seed papers that were noticed by faculty members or found through a purposeful search by faculty members or by other students.
3. The partner healthcare organization can keep a list of good seed papers based on topics that are relevant to the organization, and/or through a purposeful search.

The importance of mentoring during the conceptualization phase

Regardless of the source of the candidate topic and the associated seed paper, the next step is for the student to have an opportunity to have conversations with people with sufficient experience to help the student develop the analytic questions and frame the project.  Good guidance during the concept phase of any project is of obvious importance to avoid wasting time and to prevent the student from later becoming confused, frustrated and turned off by analytics. 

The investment of time with students during this phase is also valuable to assess the talent of the student.  Not everyone has adequate analytic talent and some students can be gently guided to apply their other talents. 

Spending time with students during this early phase also helps to shape the students’ analytic habits and provides them with the cognitive “frames” that can assist in clear thinking across diverse analytic topics and methods.  In my experience, the most important of such habits is to start with the end in mind and work back from there. Students must develop the discipline to always start by carefully defining the question or the decision being supported and the associated decision alternatives under consideration. Then, a student with good analytic habits works back from the decision to the analytic outputs needed, identifies the outcomes that are thought to materially differ across the alternatives, thinks through the basic logic needed to estimate those outcomes, and then works back to the model inputs needed to support such logic and the information sources available to provide those inputs.  The frames can take the form of patterns, templates or exemplar models that illustrate how to solve common modeling challenges.

Target state vision

When thinking through improvements to our process for developing analytic talent for the healthcare field, it is helpful to envision the eventual target state.  I envision a future with the following target attributes:

1. Students in social science departments and medical educational departments of local institutions of higher learning view the program to develop healthcare analytic talent as potentially valuable to their own career trajectory and as a way to make a difference.
2. The clinical and administrative leaders of local healthcare organizations view the program as both a pipeline for good analytic talent and as a source for useful and actionable analytic outputs.
3. The leaders of the educational institutions view the program as a valuable part of their educational mission and an important contributor to the reputation of the institution and its standing in the larger community.
4. There are defined faculty members at the educational institution and defined clinical and administrative leaders at the healthcare organizations who actively participate and collaborate in the program, attending meetings, working with students and providing other types of support to the program. They feel a sense of reward for what they have contributed and pride for what has been accomplished.
5. The students participating in the program go through a process that balances between being tailored vs. standardized.  It needs to be sufficiently tailored to serve students with different interests and a different mix of specific talents, while being sufficiently standardized so as to constitute a controlled process that can be systematically improved over time.  The program has defined activities and is supported by defined resources.  It generates defined analytic outputs, along with the structured data needed to assess the performance of the program.  Just as the program is intended to produce people capable of supporting the systematic improvement of healthcare organizations’ processes, the program itself should be a process subject to systematic improvement.
6. In addition to providing educational experiences and defined project-level analytic outputs, the program should accumulate intellectual assets that increase the effectiveness of the program over time.  For example, for model-building projects, rather than each student building each model from scratch, the program can build up templates, patterns, and reusable components in a modeling and simulation platform that can enable future student projects to build upon them. For example, a student model might include a component that does geographical analysis using GIS data for the local region. That component can be retained, packaged and documented for use by other students who can use it in their own projects, possibly making their own improvements in the component.  In this way, the program can operate as a type of open-source project, teaching students how to collaborate to build big, complex things.
7. As alumni of the program enter the healthcare organization workforce, they continue to benefit from their affiliation with the program through opportunities for continual education, mentoring and access to the intellectual assets of the program.  And they continue to contribute to the program, helping to identify good projects and seed studies, providing access to datasets for student projects, and mentoring both students and other alumni of the program.
8. Over time, the program, initially focused only on healthcare analytics, inspires others to develop similar programs for other types of talent development and contributes to a larger reimagining of the relationship between educational institutions and other organizations within communities.

Recent efforts by HHS Secretary Kennedy to politicize CDC policymaking leave us without the science-driven process the nation requires. Therefore, while we wait impatiently for the storm to pass, we will need a temporary arrangement where provider and payor organizations, professional societies and states cooperate to establish, fund and delegate some authority to a substitute policymaking process to keep the flame alive and meet urgent needs.

The sudden need

In a Medscape post this week, Dr. Kenny Lin, a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, posed the question “Who Should PCPs Look to for Vaccine Recommendations?“

In our free market system, individual physicians and other private, licensed health care professionals have the primary responsibility for decision-making about medical orders. In our constitutional structure, states bear the primary responsibility for public health. However, in the last century, the federal government has played an important role as a public payor (Medicare and Medicaid), as a funder and executor of basic biomedical and health services research (NIH and AHRQ), and as a source of deep expertise and centralized policymaking and policy advisory services (CDC and FDA). I have previously been critical of the CDC, arguing that it has sometimes strayed from its “lane” as a source of science-based advisory services into the realm of political policymaking. But such criticism was fine-tuning and optimization. As with the vast majority of physicians, I have always had a deep respect for the CDC. The CDC’s Morbidity and Mortality Weekly Report (MMWR) and immunization policy-making processes were formerly recognized as the world’s standard for scientific competence and integrity.

Formerly, sad to say.

This week, HHS Secretary Kennedy fired the CCD Director, Susan Monarez, after she refused to endorse Kennedy’s insistence that vaccine policy should be based on his personal conspiracy theories and misinformation, and not based on science, as interpreted by actual scientific experts. Her firing was accompanied by the resignations this week of four other senior CDC leaders:

• CDC Deputy Director Dr. Debra Houry,
• Head of the CDC National Center for Emerging and Zoonotic Infectious Diseases, Dr. Daniel Jernigan;
• Head of the CDC National Center for Immunization and Respiratory Diseases, Dr. Demetre Daskalakis, and
• Director of the CDC Office of Public Health Data, Surveillance, and Technology, Dr. Jennifer Layden.

These resignations follow a period of crippling budget cuts, firings, limits to CDC’s communications, withholding of previously public sources of data, cancelation of research contracts for mRNA technology, and relentless misinformation and attacks by Secretary Kennedy and other federal officials. Secretary Kennedy recently fired the entire membership of the Advisory Committee on Immunization Practices (ACIP) and replaced them with a group that includes several notorious vaccine “skeptics” — a term that sane-washes their positions as if they might have good faith differences of expert opinion about the interpretation of study results.

Deputy Director Houry wrote “For the good of the nation and the world, the science at CDC should never be censored or subject to political pauses or interpretations.” Dr. Daskalakis wrote that Secretary Kennedy put “people of dubious intent and more dubious scientific rigor in charge of recommending vaccine policy” and “I am unable to serve in an environment that treats CDC as a tool to generate policies and materials that do not reflect scientific reality.” These were principled, conscientious resignations. Our profession owes these people our gratitude.

What to do about it

It was right for those that recognized the attack on the longstanding science-driven CDC policymaking process to first try to resist it and then resort to conscientious resignation. But that leaves the question: what are members of the medical-scientific community to do now that those two lines of defense have clearly failed?

Americans still need to keep the flame of science burning—especially when it comes to healthcare. They need reassurance that science is good, and that truth, evidence and expertise are still important. But on a more practical level, the US population still needs a source of trusted medical-scientific expertise and guidance. Such guidance is sometimes used by individual citizens. More commonly, it is used by physicians, medical leaders of private healthcare provider organizations and health plans, and by state and local public health officials. It would be impractical for each physician, each organization and each state to duplicate the work of gathering and analyzing evidence to support policymaking.

Furthermore, there is value to standardization. We do want healthcare provider organizations to innovate. That implies the existence of intentional, useful variation across organizations. But, for some types of clinical practices, variation is unintentional and not useful. For some clinical topics, the direction of good is toward standardization. Such topics include basic immunization, infection control, public health, preventive services and coordinated management of epidemics. Therefore, for those topics, we need a centralized source of scientific evidence development and science-driven policy development and consensus-building. We need what the CDC was formerly qualified to do.

Therefore, I propose the establishment of a temporary entity to provide the required science-based policymaking process regarding the prevention and control of infectious and chronic diseases, until such time as the CDC can be trusted to do so.

Not just states or regions

Presumably thinking along the same lines, Dr. Robbie Goldstein, Commissioner of the Massachusetts Department of Public Health, organized a recent two-day meeting of a group of public health leaders across eight states in the Northeast to discuss collaboration on public health guidance for vaccinations and implementation of disease-tracking and emergency response services independently of the CDC. The meeting included heads of public health departments from every New England state except New Hampshire, as well as New Jersey, New York and Pennsylvania. Notably, the participants included both Democrats and Republicans. Such an effort is a good way to get a conversation started, but I do not think the target state, even in the interim, should be a process limited to state-level public health officials. And it should not be limited to a particular region of the country. Many of the relevant services are covered by private health plans that are national. Individual physicians look to national professional societies for guidance. Dr. Goldstein implicitly acknowledged this point, saying that states could look to major medical associations to shape their vaccine policy, instead of the CDC.

Not permanent privatization

It is also important to point out that this proposal is not conceding permanent privatization of the CDCs functions. There can, of course, be legitimate discussions about which societal functions should be assigned to the government and which should be handled by the private sector. And there are legitimate discussions about which government functions should be assigned to the federal government vs. states, counties or other local jurisdictions. As a life-long advocate for moderation, I see the allocation of power for specific functions across public vs. private and centralized vs. decentralized dimensions as an interesting optimization problem, not a matter of political doctrine. In other words, we should look at each function and figure out where it should be carried out.

With respect to the general functions of care process design and clinical protocol and guideline development, I would adjust the dial a few clicks toward privatization and decentralization — with more investment by health care provider and payor organizations. But on the more specific topics of infectious disease control and preventive services, I think the optimal place for that to take place is in the federal government — in the CDC. So, I want to make it clear that calling for the establishment of a policymaking process outside the federal government is in no way conceding that their specific functions should be permanently privatized. It is an interim solution. The goal is to serve the continuing need for real science-driven processes while we wait impatiently for the anti-science episode at the CDC to conclude. I hope that happens quickly, but it is unrealistic to expect that to happen within the next few years. That length of time is too long for the nation to go with an anti-science CDC policymaking process, so the private sector and the states need to make alternative arrangements.

Forming an interim privatized policymaking process requires three things: organization, authorization, and funding.

Organization

The new interim process will be a creature of civil society. It will be an “initiative,” a “collaborative,” an “alliance,” a “coalition” or a “committee.” In our healthcare field, we have lots of experience and precedent to draw upon. For example, one of my former clients is the National Comprehensive Cancer Network (NCCN), a not-for-profit “alliance” of 33 of the most respected comprehensive cancer centers in the US. The NCCN brings together oncologists and scientists from the member institutions and other experts and stakeholders to develop what many consider to be the world’s gold standard for detailed expert consensus-based protocols for 67 different types of cancer.

Another exemplar is the Primary Care Collaborative, a “coalition” that includes primary care practices, primary care specialty societies, payer organizations, employers, purchasing coalitions, patient and consumer groups, and technology and pharmaceutical companies. They “convene and unite stakeholders around research, care delivery and payment models and policies.”

There are plenty of other examples, including the Interdisciplinary Association for Population Health Science, the Health Transformation Alliance, the Healthcare Leadership Council, Better Solutions for Healthcare, the Future of Value initiative, the Population Association of America, and too many more to mention.

How do such entities get organized? It starts with individual leaders, thinkers and influencers having conversations with each other. They recognize unmet needs, common themes and opportunities. Such conversations happen in board rooms, in the hallways of conventions, in discussion panels, in podcasts, in viewpoint articles in scientific and trade publications, in LinkedIn posts, etc. Hopefully, this humble post might provide a spark or at least contribute another pixel in a picture emerging in the brains of many people trying to wrap their heads around what to do now that science is being evicted from the CDC and other federal health agencies, such as the FDA, AHRQ and CMMI.

Once there is informal recognition of the need, the formalization of an organizational entity can proceed quickly and easily. It starts with a short slide presentation or whitepaper that gets vetted by leaders of a few well-known organizations. Maybe the American Medical Association (AMA), the American Medical Group Association (AMGA), America’s Health Insurance Plans (AHIP), the National Committee on Quality Assurance (NCQA), Mayo Clinic, Kaiser Permanente, Henry Ford, Geisinger, Blue Cross Blue Shield Association, Aetna, United Healthcare, American Epidemiological Society (AES), the Infectious Disease Society of America (IDSA), etc. I’ve been involved at the birth of a few collaboratives over the years, and my experience is that once a few of the cool kids say they are coming to the party, everyone wants an invitation. Fear of missing out (FOMA) is the starting capacitor for many big motors.

Then, someone writes a charter and some bylaws and files some not-for-profit entity papers and opens a checking account, and someone gets creative with a logo. Copilot AI generated the one above in about 20 seconds.

Sadly, finding a qualified and available set of leaders will be quite easy, since many have recently lost their day jobs.

Authorization

A key step will be for various organizations that tie their own policies to those of the CDC to make changes to their own policies and contracts to unplug the CDC and plug into the interim process. Many health care payers tie their coverage and the associated medical and benefit policies to CDC policies, particularly the policies of CDC’s Advisory Committee on Immunization Practices (ACIP). For example, under the Affordable Care Act (ACA), non-grandfathered plans must cover ACIP-recommended vaccines without cost-sharing. Medicaid expansion plans are required to cover all preventive services specified in the ACA, which includes coverage for all ACIP-recommended vaccines. State-run traditional Medicaid programs are incentivized to cover all ACIP-recommended vaccines without cost-sharing.

Beyond these mandatory public payor requirements, many private commercial health plans voluntarily aligned their medical and benefit policies with CDC guidance. They did so either as a matter of standard practice or simply by routinely citing CDC guidelines in their internal policy documents. This alignment went beyond just ACIP immunization guidance. They also aligned with CDC guidance regarding cancer screening, testing for sexually transmitted diseases, tracking and management of chronic diseases, and more. Private insurers were motivated to align with CDC guidance partly because they had a trusted reputation and doing so reduced their burden of doing all of their own literature reviews and consensus building. They were also motivated to adopt CDC guidance because there was a perception that legal liability would be higher if they adopted unconventional local policies.

Note that the current anti-science corruption of the CDC’s policymaking processes is largely unidirectional — stopping recommendations for things for which there is good scientific support, rather than recommending things for which evidence is weak or absent. Therefore, when private organizations unplug from CDC and plug-in to the interim process, they will generally be deciding to cover things beyond the mandatory minimum, rather than failing to comply with requirements. Of course, if private health plans are too greedy, there could be a race to the bottom, dropping coverage as permitted. But my sense is that the cost differential for the types of services likely to be affected is not sufficient to motivate health plans to move toward the dark side. I’ve been wrong before when betting on health plan morality, but I do think getting the plans to commit to covering the services recommended by the interim process could be doable – as long as it steers clear of costly GLP-1 receptor agonists!

Funding

If large organizations join the coalition supporting the interim process, the cost sharing dues to the member/sponsor organizations could be affordable. A core team of experts would need to be hired. Expert consensus panels would be convened. Literature reviews, meta-analysis, document preparation and publication processes would need to be funded. In this regard, I consider the NCCN to be the existence proof of the economic feasibility of a trustworthy participant-funded organization producing detailed, useful, science-based consensus guidelines. The guideline development processes of many professional societies demonstrate what can be done on an even smaller budget. But in my opinion, all but a few are below the level of quality and rigor that I think is warranted for the interim process I am proposing.

Of course, the level of funding that could be assembled from a coalition of the willing will never be close to replacing the former CDC’s taxpayer-supported budget. The CDC does a lot more than just support policymaking. The coalition cannot be expected to replace the primary data collection, training support and, most importantly the field operations that are part of the crown jewels that are in the process of being squandered.

Making it palatable

The current federal administration proudly acknowledges “retribution” as a campaign promise and tactic. Therefore, the biggest barrier to establishing an interim policymaking entity might be the fear of retribution. There is, unfortunately, an emerging pattern of cowering among leaders of large organizations across civil society. Leaders of some major universities, law firms, technology manufacturers and others have chosen to crouch, flatter and comply. Organizations that sponsor and fund an interim policymaking entity might feel like they have popped their heads out of a foxhole, attracting fire from federal officials. Therefore, to actually get this done, it will be necessary to overcome such fear and make the idea politically palatable. It will be necessary to win at least some bipartisan support, which is not implausible since support for medical science has not traditionally been a partisan issue.

This support can be accomplished by two main methods. First, we will need to communicate the invitation in a way that appeals to the deep, ancient sense of duty and responsibility as professionals who took an oath — as physicians and scientists. Second, we will need to communicate with finesse and a sense of pragmatism. It should take care to minimize the use of partisan rhetoric and inflammatory terminology. In this spirit, the name for an interim entity should avoid being provocative and vulnerable to misinterpretation. Perhaps we should call it something like the “Disease Control Science Committee” — something peaceful, apolitical and humble, but nevertheless staunch in defense of truth and scientific integrity. I asked Copilot AI to draft a seal designed to convey those ideas, and it quickly came up with the one at the top of this post. I think it looks pretty good. I can imagine it on lapel pins given to the members of the committee, who will be proud to wear them because they were selected for their scientific credentials and integrity.

Groups matter

When working as a consultant, leader, analyst or designer,  the work requires a mix of individual contributions and group contributions.  Group processes are required to (1) generate ideas, (2) vet ideas, and (3) achieve consensus, which is often a pre-requisite to buy-in and approval of initiatives to improve or build something.

Group processes only succeed if no member of the group feels fear that their membership in the group is under question or that sharing their ideas might lead them to experience humiliation or disrespect.  Even a single member with such feelings often poisons the group dynamic.  I wrote about this in a blog post in 2011: What makes teams effective? (2) Achieving the Piranha Club state – Reward Health.  The post emphasized the use of humor and intentional efforts to build a group culture and motivate group behaviors to participate non-defensively and provide only “loving” feedback.  Although I did not use the terms “emotional intelligence” and “group dynamics” in that post, I think those concepts are applicable.

Two talents matter

In group processes, different members of the group have different talents and capabilities in two key areas: (1) conceptual clarity, and (2) capacity for complexity.  

Conceptual clarity is about cognitive neatness and precision, the ability to understand parallelism when authoring a list, and the ability to intuitively pay attention to semantics and terminology and taxonomy, even when they don’t use any of those words.  It is the talent and habit of mind that drives people to put the right silverware in the right slots in the drawer, or to organize their toolbox or dresser drawers, or to put labels on things, or to organize their electronic files in a folder hierarchy, or to use a standardized format for naming files.  If a group is engaged in brainstorming a list of fruits, and the first person says “apples,” the second person says “bananas,” and the third person says “carrots” and the fourth person says “apple pie” — that is an example of failure of conceptual clarity.  Similarly, if a group is developing a process flow diagram, and some of the boxes are process steps and other boxes are types of information and other boxes are outcomes, and other boxes are the names of people, and some of the arrows describe sequence and other arrows represent information flows and other arrows show cause-effect relationships and other arrows show reporting relationships — that is also an example of a failure of conceptual clarity.

The capacity for complexity is the ability to hold many things in one’s brain at the same time.  Such capacity is a pre-requisite for reading and understanding a diagram with many boxes and lines or reading and understanding a block of text with many words and some long sentences.  Although there is value in simplicity, there is also great harm in oversimplification, as I noted in blog post in 2016 (Three ways to keep it simple — one of which is bad – Reward Health). The distinction between simplicity and oversimplification is based on the notion that complexity is something that exists in the world and people may succeed or fail in acknowledging it.  When people lack the capacity for complexity, they are vulnerable to oversimplification.   If a group is making a process flow diagram, and they assume only the “happy path” and fail to acknowledge the need for process steps to deal with exceptional circumstances, that is an example of failure to acknowledge complexity.  If a group is creating an entity relationship diagram, and someone wants to “keep it simple” by assuming that every market area is associated with one and only one state, even though some market areas actually overlap state boundaries, that is another example of a failure or unwillingness to recognize the existence of complexity. 

The harm is huge

It is my experience that a majority of the resource waste and quality defects in large, complex health care delivery and health insurance organizations can ultimately be traced to failures in conceptual clarity and the acknowledgement of complexity.  Bad data causes problems that require costly remediation.  Processes that break when facing exceptions cause defects that harm customers and brand reputation and also require costly remediation that cuts into profits.  In health care organizations, broken care processes can kill people. Organizational structures with ambiguous responsibility boundaries cause turf conflicts when boundaries overlap and failures of accountability when there are gaps between boundaries.  All these problems have common root causes in failures of conceptual clarity and failures in the acknowledgement of complexity.

Group dynamics challenges

When a person receives a chunk of information, they naturally make a quick assessment of the complexity of the information and a quick determination about whether that level of complexity exceeds their capacity or if it would require a level of cognitive burden that exceeds the level they are motivated to accept.  They reject the information if it is too complex or burdensome.  My daughter taught me the text messaging shorthand “TLDR” — an acronym for “too long, didn’t read.”   (As an aside, people also are prone to reject information that does not support their prior beliefs or perceived self-interests.) 

If they are in a group context, they face some social pressure to process the information and contribute to the discussion.  But they also face an additional level of fear that if they try and fail to process the information and then try to contribute, they might be humiliated, driving them to not only reject the information but to preemptively discredit the provider of the information and reject the expectation for active participation in the meeting.

Similarly, when a person considers whether to participate in a group process of authoring a new chunk of information, they naturally make an assessment of the level of expected rigor and discipline, and if that level exceeds their capacity for conceptual clarity, they withdraw from participation and preemptively discredit the process and the associated expectations as being “too academic” or “too nerdy” or “too picky.”  They roll their eyes and cross their arms and lean back in their chairs.  After the meeting, they will badmouth and disown the group process and connive ways to thwart it.

These challenges are not rare.  In my decades of experience working with large and small organizations, I would estimate that at least a third of all group processes were substantially harmed due to challenges related to resistance to information for which the conceptual rigor and complexity exceeded the comfort level of some of the participants.  (And another third failed entirely because the information failed to conform to the prior beliefs and self-interests of some of the participants.)

What works for groups

Knowing about these principles of human psychology and group dynamics, a leader of a team that is to engage in group processes involving analysis or design must select group process methods that do not exceed the talents and capacity of any of the people involved.

In my experience, the following methods ended up being most useful:

• Brainstorming — making a list of concepts or ideas with no expectation that all the contributions have to be of the same type, thereby minimizing requirements for conceptual clarity.  Each person contributes only a single concept or idea at a time, with no expectation of relating to other items, thereby minimizing requirements for capacity for complexity.  The group is explicitly instructed to abstain from making any negative comments about the contributions of other people, thereby reducing the fear that contributors will feel humiliation or disrespect.
• Mind Mapping — like brainstorming, but instead of a simple list, the artifact is a diagram with concepts inside shapes and with lines connecting the shapes.  As with brainstorming, there is no expectation that the concepts or lines are of the same type, minimizing the requirement for conceptual clarity.  Each participant contributes only a single concept and line at a time, and each concept is connected to only a single other concept in a simple hierarchy, thereby minimizing the required capacity for complexity.    Also like brainstorming, the group is typically reminded to refrain from providing any negative feedback, reducing fear.
• Commenting on a draft.  Based on the results of brainstorming and mind mapping exercises done during previous group sessions, one individual or a small subgroup of a few individuals with greater talents for conceptual clarity and capacity for complexity can draft clearer and more complex artifacts, such as textual documents, process flow diagrams, concept maps, cause-effect diagrams, entity-relationship diagrams, organizational structure diagrams, GANTT charts, or system architecture diagrams.  Then, these materials can be provided back to the group, preferably with lead time for review by the subset of group members that are willing and able to review the draft and formulate feedback comments.  Finally, in a group meeting, a brief presentation of the draft can be provided, allowing some participation by those that were unwilling or unable to review the material in advance.  Comments are gathered in a manner that resembles brainstorming, with the expectation that the subgroup will use the comments to consider edits to the draft after the group meeting has ended.  A wise group leader intentionally avoids getting into “group wordsmithing” mode, other than for very specific scenarios such as when drafting a mission statement, product name or tagline.

No guaranteed success

Even if the group leader does an optimal job in selecting group methods and managing group dynamics, success is not assured.  The likelihood of success is a function of the subject matter and the talents of the members of the group.  If the group is conducting an analysis in order to make a decision, the analysis might involve complexity that exceeds the group’s capacity.  If the group is participating in a design process, the integrity and coherency of the design might exceed the group’s capacity for conceptual clarity and the optimal design might involve a number of components that exceeds the group’s capacity for complexity.  The requirements for success and the capacity of the group are like a Venn diagram for which there may or may not be overlap.

A dozen ways to improve current group processes

Given this uncertainty about success, all a leader can hope to do is take steps that increase the chances of success for current and future group processes. The following 12 strategies can help with current group processes.

• Choose the right participants.  When choosing group participants, stack the deck with people with greater capacity.
• Use spoon feeding.  Focus on the least capable group members and try to spoon feed them safe opportunities to provide input, careful to avoid any appearance of condescension.  One way to accomplish this is to schedule time to pre-review draft materials with the least capable or most difficult participants outside of the group meeting to provide them with a safer opportunity to provide feedback, then incorporate the subset of those ideas that make sense into the draft.  Then, at the group meeting, make a big deal out of those contributions.  Again, even traces of condescension are poisonous.  I’ve seen group leaders and otherwise sophisticated consultants try to use this technique with a kindergarden-teacher “uptalk” voice (with rising pitch at the end of sentences), giving the verbal impression of patting the audience member on the head.  The result was the immediate cessation of group process effectiveness.
• Use naming.  Try to get people to name things, which has a strong effect on their sense of ownership, being careful to avoid setting the expectation that they should name important things for which they might provide a conceptually unclear name.
• Use choosing.  Try to provide opportunities to choose between multiple “good” versions of the design, giving participants the necessary sense of ownership and control with reduced risk of selection of conceptually unclear or oversimplified designs.
• Use “bracketing” to overcome objections to complexity.  This technique involves responding to people that reject complexity by providing multiple versions of the design artifact with different levels of complexity, anchoring the spectrum with one version that is clearly oversimplified and another version that includes gratuitous, unnecessary complexity, thereby steering the participant away from the “simpler is always better” mindset and toward more thoughtful designs that are conceptually clear and that acknowledge the right level of complexity.
• Use “build up” approach to presenting artifacts, such as by using a series of PowerPoint slides or using slide animations to present a more complex diagram by starting with one box, then adding an arrow connecting to a second box, then adding another arrow connect to a third box, and so on.  When a group of people see a complex diagram all at once, some of them will immediately tune out.  But by initially presenting something very simple, it is possible to lull some of those people into continuing to pay attention, receiving bite-sized increments of complexity until they are eventually comfortably viewing the whole diagram they would otherwise have tuned out.
• Use intuitive graphical language.  I learned long ago that diagrams can be much more intuitive if attention is paid to layout, particularly to avoid crossed or tortuous lines. Colors and font sizes are important to convey grouping and relative importance. Some visual symbols are more intuitive than others.  For example, I learned long ago that people accept concept maps (with labels on the lines) more easily than entity-relationship diagrams (with symbols on the lines). Among entity-relationship diagrams, people find “crow’s feet” symbols far more intuitive than the Chen or UML notations sometimes favored by IT architects. 
• Use casualness.  In my consulting practice, I often use hand-drawn diagrams toward the beginning of engagements to convey the idea that the group participants are involved in the early stages of idea development, making them feel a greater sense of ownership.  Even when I have developed a more polished version of the information for a previous client, such that I could easily adapt it for use with the current client, I nevertheless use a hand-drawn version.
• Use humor.  Particularly when people are anxious, they appreciate comic relief.  Therefore, when a group process involves conflicting interests or opinions or fear of being disrespected, the leader can insert some humor and encourage the group members to keep it light.  Self-deprecating humor, in moderation, can be particularly effective in smoothing over conflict and setting an example to increase humility and decrease aggression.
• Use polish and beauty.  People will pay attention to things that are visually interesting and beautiful, and they will put more trust in things that are polished and that otherwise reflect high production values.  So, especially toward the end of a group process, polish and beauty are good.
• Use story telling.  Humans are naturally attentive to stories. They want to experience movement and change. They are curious to see what comes next and how it ends.    Therefore, the more analysis and design artifacts can be presented in some order that sets the scene, presents a problem, reaches some climax, and offers some resolution that is actionable, the information will be far more effective and impactful.
• Use tailoring.  The artifacts can be provided in multiple forms, providing different versions of the same information tailored to people with different backgrounds and interests.  For example, in addition to providing a complete detailed version of a diagram, a simple “high level” version of a diagram can be provided for the people unwilling or unable to deal with the detailed version. Also, different versions can express overlapping information using terminology that is more familiar to different participants, such as using clinical terminology for clinicians, financial terminology for finance people, IT terminology for IT people, and general common terminology for lay audiences.  Some participants may benefit from access to a video explanation that they can play and replay as needed to help them absorb and understand, giving them the confidence to participate in the group meetings.

Five ways to succeed in future group processes

• Over time, hire more talented people, with attention to the talents of conceptual clarity and capacity for complexity.
• Engage consultants with those talents to augment the internal team, and maintain long term relationships with those consultants so they can function as trusted members of the team, rather than being seen as threatening “outsiders.”
• Build a “piranha club” culture over time — rewarding people for the bravery of participating and “gracefully” (non-defensively) accepting negative feedback and for providing only “loving” (constructive and gentle) negative feedback to the contributions of others.
• Build familiarity and skill with tools, methods and forms of information that facilitate thinking and communication that is conceptually clear and that acknowledges complexity.  The capabilities of people for conceptually clarity and capacity for complexity are largely matters of natural talent, but are partly a function of familiarity and training.  For example, if people have a lot of experience reading process flow diagrams, they will over time have a greater capacity for reading more complex diagrams, just like a person has an easier time playing more difficult pieces on the piano after spending lots of time learning their scales and playing simpler pieces.
• Cleanliness begets cleanliness.  Over time, an organization that insists on conceptual clarity and appropriate acknowledgement of complexity will build a strong base that sets an example and facilitates continuation of those good organizational habits.  When data is a mess, cleaning it properly seems hopeless. But when data is clean, errors that are introduced over time seem fixable and doing so is of obvious importance.  Similarly, when organization structures and job descriptions and process boundaries are clear, it is easier to keep it that way.  When information systems have a strong, clear architecture and bug-free implementation, they can be far more easily maintained and improved over time.  Therefore, conceptual clarity and acknowledgement of complexity are self-reinforcing annuities that pay dividends into the future.

Slashing and Dissolving

As reported by the Kaiser Family Foundation’s Health News site blog, last week the Trump administration “dissolved” the Agency for Healthcare Research and Quality (AHRQ), the agency that has for decades been at the center of the US Federal Government’s support for health services research in the service of improvement of our healthcare system. The agency had about 300 employees and a budget of $369M (less than 0.04% of HHS spending), allocated mostly to grants to university-based health services researchers. This month, the administration slashed more than half of the employees, draining essential PhD-level expertise. They proposed to dump the remainder into the NIH for now, where it is unlikely that the prior focus will persist. KFF reported that when members of Elon Musk’s “DOGE” team first met with AHRQ leaders last month, they said they planned to cut the staff by 80-90% even though they acknowledged they did not know what the agency did. In the US House of Representatives, the Republican 2025 budget proposal for AHRQ is zero.

It felt like home

Since my career started in 1990, I’ve been involved in various interrelated fields, with the field of health services research near the center, with nearby adjacent fields being quality and outcomes measurement, clinical effectiveness, practice guidelines, quality improvement, and population health management. The U.S. federal government’s involvement in health services research touches a number of different agencies, including NIH, FDA, CDC, CMS, CMMI, and ASPE. But the part of the federal government that was most focused on supporting the health services research community was the Agency for Healthcare Policy and Research (AHCPR), which was renamed the Agency for Healthcare Research and Quality (AHRQ) in 1999. Among all the government agencies, that felt most like home.

1990s: Guidelines and Cost-Effectiveness Era

When I first got to know AHCPR in the early 1990s, it still had “policy” in its name. Policy included broad ideas about how healthcare should be organized and financed. But, in that era, the policy emphasis was on the type of policies that applied to the actual decision-making of clinicians at the point of care: clinical practice guidelines (and their cousins: protocols, algorithms, care maps, appropriateness criteria and quality metrics). This was the era when science was revered. Among physicians, particularly in specialties such as cardiology and oncology, clinical trials were worshiped. The term “evidence-based medicine” was on everyone’s lips. But the reverence for science was not limited to biomedical science. The fields of epidemiology and economics were also respected, along with other social science disciplines. This co-existing respect for biomedical science and economics led to intellectual integration. We understood that healthcare resources were not unlimited, and that we could use rational thought, logic and data to inform and optimize the allocation of resources. We understood that we had to focus not only on efficacy (the outcomes achieved in rigorously controlled clinical trials), but also on effectiveness (the outcomes achieved in real-world conditions). We understood the need to avoid waste and improve efficiency. And we understood that we needed to consider not only the “hard” outcomes such as deaths, but also more subjectively measured patient experience, including pain and physical function. To enable the simultaneous consideration of multiple health outcomes, we understood the need for summary measures — most notably the concept of the “quality-adjusted life year” (QALY). To simultaneously consider both health and economic outcomes, we understood that we needed to create summary measures that combined both — most notably the concept of the cost-effectiveness ratio. I consider this era to be the golden era, when things fit together into a coherent, internally consistent whole and there was a sense of optimism that our thinking and methods were only going to improve, for the benefit of the whole population.

2000s: Safety and Quality Era

But then, toward the end of the 1990s, there was a backlash. While the idealistic health services research community was flourishing, the HMOs were being taken over by more opportunistic people who realized that clinical process improvement was tedious and that there was money to be made by focusing instead on two main alternative strategies: (1) avoid sick people, and (2) deny payments. These strategies harmed the HMO members, and they also were particularly harmful to specialist physicians for whom denials were pay cuts. So, the specialists got their patients riled up, and “managed care” was transformed from an idealistic euphemism to a nasty insult, as I described in a blog post in 2011. During that era, the AHCPR was busy developing guidelines, openly and innocently considering cost-effectiveness based on cost per QALY metrics, declaring certain procedures to be cost-ineffective, even if they were profitable to the specialists. The mustache-twisting HMO administrators were happy to apply those guidelines to justify their denials, and the specialists (starting with the urologists) were incensed that the government (AHCPR) was complicit in the evil called “managed care.” A political furor erupted, leading to regulations prohibiting the US government from uttering the word QALY. AHCPR needed to be cleansed, which was accomplished by removing the word “policy” from the sign outside their headquarters and renaming it the Agency for Healthcare Quality and Research (AHRQ). No more government meddling in “guidelines.” Avoid controversy. Stick to the boring topics like safety and quality. I was disappointed in the defanging of the agency, and more broadly in the fact that our noble field had been discredited by HMO CFOs and that we were banished from real influence for what I call the lost decade. Nevertheless, quality and safety are actually important. We were able to remember to order needed mammograms, pap smears and immunizations more often. We were able to reduce hospital-acquired infections. We developed lots of HEDIS metrics and produced a lot of “report cards.” It kept us busy and occupied and not talking about QALYs or cost-effectiveness.

2010s: Health Disparities Era

Even back in the 1990s, it was a common subject of health services research to measure and comment on the lower utilization of healthcare services and worse health outcomes among “disadvantaged” or “vulnerable” populations, defined by race, ethnicity and soci0economic status. In the 90s, I was a coinvestigator in an AHCPR-funded “Medical Treatment Effectiveness Program” (MEDTEP) center at Henry Ford Health System in Detroit, addressing such disparities among the African American population through efforts to improve the “cultural sensitivity” of our care processes and by engaging with churches and other community organizations. So, we’ve been on the case for a long time. But, during the 2010s and particularly in the first half of the 2020s, the emphasis on health disparities by AHRQ and other government agencies continually increased, and the goals of health disparities efforts changed from the types of things we were doing in the MEDTEP center to a more “affirmative” and “reparative” approach. I can fully understand why people that have been subject to injustices and who have long been frustrated with slow and seemingly insincere efforts to resolve those injustices would demand stronger and more exclusive focus. They are rightfully angry and they want the problem solved yesterday. However, it is also true that the health disparities focus of the AHRQ and other agencies increased to the point of crowding out efforts to make fundamental improvements in healthcare structures and processes intended to help everyone. It has sometimes seemed as if the AHRQ team (and cousins at CMS CMMI and in other agencies) were saying “first things first, we reallocate resources to the disadvantaged and underprivileged, and then maybe later we can talk about improving the overall healthcare system.”

During this era, the federal government continued efforts to push further into the background the 1990s interest in cost-effectiveness. When leaders of the private healthcare and health insurance industries talked about health care “value” and advocated for “value-based contracts” — the federal government agencies continued to treat the topic as an untouchable third rail. As I’ve noted in a blog post in 2024, the CMS CMMI played Animal Farm-style linguistic games to define “value-based care” so that it had nothing to do with cost, efficiency, waste, economics, population or finances — it is just about quality of care, provider performance, and whatever experience individual patients value. They’ve been continuing the health economics denialism of the late 1990s.

2025 to ?: Anti-Science Era

Now, we have entered a new era. In the current era, the federal government prioritizes loyalty over evidence. The federal government is no longer just focused on directing the priorities of health services research. From the perspective of the current administration, scientists, and particularly social scientists, cannot be trusted because the scientists say their loyalty lies with the evidence, and to the degree they have shown bias, it has been in favor of fighting injustice or saving the environment. In the specific context of health services research, the administration sees the track record during the health disparities era, and concludes that the agencies and the community of health services researchers they fund are unlikely to become loyal to the administration’s commitment to end “woke” and “DEI.” Therefore, burning down the agencies and starving the researchers is the best option. Unfortunately, there is an asymmetry between the tediousness of building up a scientific community and the expediency of destroying such a community. The brain drain can occur quickly. Health services researchers will reluctantly but certainly move on to other domains of research, or perhaps take up gardening.

Rebirth Era

Although I am not feeling optimistic, I’d rather not end with gloom. At some point, in four months, or four years, or forty years, the anti-science era will wane. At some point there will be an opportunity to start again, to reestablish the scientific and philosophical foundations of the field of health services research, and of the healthcare field itself. My hope and expectation is that a fresh start will look more like the 1990s, before health economics denialism set in, before the field was co-opted to any particular narrow focus, and before loyalty was prioritized over evidence. I hope to live to see the day when health services researchers operate again at the intersection of biological sciences and health economics, willing to bravely:

• Acknowledge limited resources and make trade-offs even when others prefer to blame the messenger,
• Get into details and take on complexity even when the audience demands simplicity,
• Stay on mission even when others focus on serving narrower interests and signalling their own virtue, and
• Follow facts and logic to conclusions even if they are opposed by those profiting from the status quo.

I want to call your attention to a very thoughtful blog published by the semi-retired health care journalist, Merrill Goozner: the GoozNews Substack. Goozner’s most recent post, “It’s mostly overutilization, stupid” provides a delightful trip down memory lane, summarizing the decades-old debate about the causes of geographic variation in health care cost.

Goozner started by reminiscing about work done at Dartmouth starting in the 1970s by John Wennberg, Elliott Fisher and others analyzing “small-area variations” in the utilization of services for Medicare patients. That work, beautifully reported in the Dartmouth Atlas of Health Care, pointed the finger at excessive utilization, including what was called “supply-sensitive care” where doctors and hospitals keep themselves busy.

Goozner also reminisced about the contrary view that “it’s the prices, stupid” espoused in 2003 by Gerard Anderson at Hopkins and health economics legend Uwe Reinhardt at Princeton based on their analysis of country-level variations using OECD data. This conclusion was reiterated in 2015 by Zack Cooper of Yale, Martin Gaynor of Carnegie Mellon and in a series of Rand Corporation studies. Goozner noted that this price-oriented perspective informed the Biden administration, which directed its DOJ and FTC to scrutinize mergers thought to harm competition and enable price increases, while the administration repurposed CMMI away from what they considered to be disappointing “value-based care” efficiency incentives to instead emphasize health equity.

Then Goozer summarized an important new paper published in JAMA this month by Joseph Dieleman and colleagues from the University of Washington Institute for Health Metrics and Evaluation, exploring the Drivers of Variation in Health Care Spending Across US Counties. In the mother of all claims data analyses, the U of W team assembled 2019 claims data from diverse public and private sources to conduct the first analysis of county-level and payer-specific drivers of health care cost, controlling for disease prevalence, age and other factors. Their main regression analysis findings:

They provided subanalyses for six different types of services, four different payer categories, nine different disease categories, and ten different states, all showing that while the specific percentages differ, the overall conclusion that services utilization is the main driver of variation was consistent.

In the discussion, Dielemen implicitly recognized the longstanding utilization vs. price debate and offered an insightful explanation. Price increases were responsible for 50% of the growth in national health spending over time between 1996 and 2019, but as of 2019, prices explained only 24.1% of the variation in county-level expenditures. In other words, both are partially true. The utilization variation is large and important and has been there all along. But half of the overall growth over time (in both low and high utilizing counties) is due to price inflation that exceeds the general inflation rate — possibly due to reduced competition from provider consolidation.

Why do we care about geographic variation?

As I explained in a comment in Goozner’s blog: all along, the main purpose of studying cost variation by geographical areas has been to use lower-cost areas as the existence proof that achieving lower cost is possible. The idea was to find out what works in those areas, and try to do the same elsewhere — or the reverse: finding out what does not work in the high cost areas and trying to stop doing those things.

This is a specific application of what we used to call “benchmarking” — where you find performance outliers and then learn from them to identify improvement opportunities that can be shared and implemented elsewhere. Over the years, the original meaning of benchmarking has been largely lost. People have been focusing on the performance comparisons, but forgetting the part about learning and improving. Too many people think benchmarking is about rewarding excellence with praise and shaming the “bad apples” — a problem I’ve described in 2011 and again in 2021.

Don’t get distracted from innovation.

But, it is always useful to step back and see the bigger picture. What we really care about is improving our health care system. That includes improving both health and economic outcomes — in other words, improving “value.” When you focus on cost variation, there are two problems. First, you are focusing only on cost, not on value. Second you are implicitly limiting the scope of your search for cost saving opportunities to the structures and processes that already exist in some geographic areas. Cost variation studies do not identify opportunities to reduce cost (or, more importantly, to improve value) through innovation and optimization.

Innovation can take two main forms. First, we can improve clinical decision-making processes — the processes used to determine what needs to go on each patient’s plan of care. This type of innovation involves the development of superior practice guidelines and protocols. This development requires literature review, modeling, cost-effectiveness analysis, and consensus-building to design guidelines and protocols expected to deliver more value. Then, we can confirm that those expected outcomes were true through observational outcomes studies and clinical trials. The second type of innovation seeks to improve care delivery processes – the processes used to efficiently execute the items on the plan of care, in conformance with quality standards and without errors, waste or delays. The improvement of care delivery processes can be accomplished using methods such as CQI, TQM, 6 sigma, etc.

As a traveler in the health care improvement field, I lament the decline in interest in the admittedly tedious work of improvement, optimization and innovation. And I also lament the distraction from that tedious work caused by the continued focus on debating the relative strength of prices or utilization in explaining existing geographic variation in cost of care.

Can we learn from the work of Dielemen et al? Yes, absolutely. But, the real opportunity is not to emulate low cost counties. It is to innovate to create care processes that offer greater health care value than is currently achieved in any county.

The recent nominations of science-disrespecting people to lead HHS and CMS should be a wake-up call.

Growing up, I naively believed that antipathy to science was relevant in Galileo’s time, but not modern times. Scientists as people were sometimes treated with disdain — as nerdy, awkward people wearing beanies with propellers, pocket protectors with slide-rules, and thick glasses with tape at the bridge of the nose. But that seemed playful and likely to be motivated as much by intelligence jealousy as hatred. I thought people were respectful of science itself out of gratitude for the obvious gifts of science, such as antibiotics, chemotherapy, refrigeration, telecommunications and air travel. It has only been recently that I realized there has all along been a substantial undercurrent of skepticism and malevolence for science itself within a broad portion of the population.

Looking in the mirror

I am unclear about the relative contribution of various causes of anti-science attitudes. Some of the problem is undoubtedly the result of sustained intentional efforts to discredit science by greedy people with particular interests in denying inconvenient truths, such as tobacco smoke carcinogenesis and fossil-fuel-driven global warming. Some of the problem is probably collateral damage in a more general war against “elites” of all kinds. But, within the scientific disciplines, we need to consider whether scientists’ actions have contributed to the problem.

In a recent editorial in the journal Science, Marcia McNutt, the President of the National Academy of Sciences, wrote:

Although scientists must never shirk their duty to provide the foundation of evidence that can guide policy decisions and to defend science and scientists from political interference, they must avoid the tendency to imply that science dictates policy. It is up to elected officials to determine policy based on the outcomes desired by their constituents. It is the role of science to inform these decision-makers as to whether those desired outcomes are likely to result from the policies being enacted.

I’ve made this same point in a previous post criticizing the CDC and other science and public health authorities during the COVID-19 pandemic for making pronouncements that were presented as being purely “science-driven” when the decisions included non-scientific considerations that were not explicitly acknowledged. For example, early assertions that masks were not necessary were undoubtedly influenced by underacknowledged considerations regarding the inadequacy of the supply of personal protective equipment for health care workers. Early dismissal of aerosol transmission of SARS-CoV-2 was probably influenced by underacknowledged considerations of the economic implications of ventilation system modifications. I also made this same point in another previous post in which I proposed a “harm index minimization strategy” for public health policy analysis, where people with expertise in various scientific disciplines develop models that predict health and economic outcomes of policy alternatives, and then leave it to political leaders who are better equipped and authorized to consider the values and interests of various constituents and to negotiate and approve the optimal policy alternative.

In McNutt’s editorial, she made a similar point, applied to climate policy:

Although science can affirm that climate change is happening and is primarily caused by anthropogenic greenhouse gas emissions, science can only predict the outcome of the various policies that might be enacted to address the problem. It is up to society and its elected leadership to decide how to balance these options, including the use of renewable energy, climate adaptation, carbon capture, or even various interventions that reflect sunlight back into space.

Social scientists are the worst offenders

The physical sciences and biological sciences have generally stayed within their objective, apolitical lanes, although they have had difficulty doing so in the domains of environment, nutrition, disease screening and prevention, mental health, epidemiology and public health. But, the greatest difficulties seem to occur in the social science disciplines, including health services research, economics, sociology, anthropology, psychology, and especially political science. From my perspective, over the past few decades, many social scientists seem to have given up even the pretense of objectivity, viewing their mission and purpose as “raising awareness” and outright advocacy for positions they take based on their sense of morality and justice, rather than based on observations, measurements and modeling. Just as some physician executives like to wear their white coats and stethoscopes to business meetings because they want to claim special authority, many social scientists wrap themselves in thick social scientific jargon that too often gets tangled up in other jargon used to signal social justice virtue. They use such jargon, along with their academic credentials and the prestigious venues in which they speak and write, to claim special authority when asserting their political positions. This drives common people crazy. People react by rolling their eyes and closing their ears to scientists.

Separate your scientist and non-scientist persona

When any scientist — physical, biological or social — strays into the political lane while wrapped in the vestments and jargon of science, they should feel guilty for cashing in the trust of science in the eyes of the public. That trust takes generations to build up, but can be squandered much more quickly. No matter how strong the scientist feels about the righteousness of their political causes, they should honor and defend science itself, which has great long term value to humanity. Sciences can be defended in small ways, such as by making counter-arguments to specific anti-science rhetoric and misinformation. But the big way to honor and defend science is to produce scientific outputs that are clear, objective and apolitical. We expect judges and generals to refrain from political speech to defend the appearance of objectivity of the justice system and the military. Likewise, scientist should be reluctant to speak politically, and if they choose to do so, they should make it clear to all that they have taken off the vestments of science and dropped the jargon of science and are speaking as a citizen or resident.

In the November 1, 2024 issue of JAMA, Ciara Duggan and others from Harvard and Brigham published a paper evaluating the utilization of “low value services” across major Medicare Advantage insurers and Traditional Medicare. Low value services utilization was defined based on the Choosing Wisely campaign and the US Preventive Task Force, and incorporated in metrics by the actuarial firm Milliman within their MedInsight Health Waste Calculator software. The authors used a dataset from 2018, so the results are already a little stale, but the scale of the dataset with 2.3 million Medicare beneficiaries provided the power to dive into plan-level performance and to break out performance into service categories. They found that, compared to Traditional Medicare, some plans such as Centene, Humana and United Healthcare outperform Traditional Medicare, while other plans such as CVS Aetna, Cigna and Blue Cross Blue Shield affiliates actually perform worse than Traditional Medicare, with by far the worst performance coming from Anthem, the largest of the Blue Cross Blue Shield affiliates — with total utilization of low value services 9% higher than Traditional Medicare.

The study also broke out the low value services into categories, and as shown above, comparing the performance of Medicare Advantage PPO and HMO plans and Traditional Medicare. In every category of services except diagnostic and preventative testing, the MA HMO plans outperformed the MA PPO plans. MA HMO plans, with their narrower networks and stronger influence by the plan on the beneficiaries’ relationship with their primary care physicians, performed particularly well in reducing the prescribing by those physicians of low value common medical treatments such as antibiotics for upper-respiratory infections or ear infections — leading to a 29% reduction compared to Traditional Medicare. MA HMO plans also achieved a 24% reduction compared to Traditional Medicine in the utilization of low value procedures and surgeries, which could be a function of gatekeeping or value-based contract incentives of referring primary care physicians, network selectivity of specialists with prudent and evidence-based practice patterns, and/or pre-authorizations and associated coverage denials.

The authors made an important point: MA plans have been treated as if they were homogeneous, when in fact they are heterogeneous. I think the implication of this is that we need not only to routinely update this type of analysis, but we need for the low value services utilization metrics to be associated with the programs, incentives and network characteristics that are effective in avoiding low value services utilization.

Critique of the presentation of the results

I usually do not write blog posts just reporting the results of papers I read. This humble blog is not going to meaningfully extend the reach of JAMA! I write only when there is something worthy of critique and commentary. In this case, my critique is about the way our scientific publications present data. I did the graphical analysis above to present the data based on what I think the “real” question is. Journals like JAMA present themselves as “scientific” even when the subject matter is policy decision-making. Science is about coming up with a theory regarding cause-effect relationships, and then making observations and conducting analysis intended to support or refute the theory (which are described as hypotheses in the context of particular studies). The premise of this scientific process is that a theory that was plausible before the study is more likely to be true if an observed association between some cause and some effect is stronger than some accepted threshold. In more precise words, the hypothesis is supported if the association is “statistically significant.” Science is a fantastic tool when applied for its intended purpose. But scientific methods were not designed for policy decision support. Policy decision-making is based on comparing outcomes and associated uncertainty across decision alternatives. In policy decision making, uncertainty remains relevant, but the particular method of statistical significance testing is not. In my opinion, the subject matter of the report by Duggan and colleagues tells me that the real purpose is to support policy decision-making. But in order to publish their analysis in JAMA, they needed to frame it as a scientific paper, for which they needed to state a hypothesis before conducting the analysis and then conduct significance testing, and then report the conclusions based on that significance testing. When an observed relationship is found to be not statistically significant, the expected virtuous behavior in the scientific tradition is to report it as “we observed no differences” which gets interpreted by most readers as “there was no difference.” That is a great tradition when the purpose is science because it avoids wasting the limited pages of journals and the limited time of abstract presentation attendees and journal readers on associations too likely to be observed by chance. But, policy decisions are made on a best-information-available basis. Confidence intervals are informative for policy decision-making, but the “no differences” tradition is just wrong for that purpose.

The authors knew that to get accepted in JAMA, they needed to twist their decision-oriented analysis into a proper scientific study. The authors presumably were rightly worried that even with their enormous sample, they might lack the power to achieve statistical significance in subanalysis. Therefore, they chose the safer route of framing the “key questions” of their analysis into broad questions for which they they had sufficient sample size to give them the power to have a good chance of achieving statistical significance:

“First, are there significant differences in the use of low value services between Medicare beneficiaries enrolled in MA vs Traditional Medicare at a national level?”

“Second, are these differences consistent across major MA parent insurer categories?”

Based on reading the paper, it is clear that the authors also wanted to understand if there were differences between MA PPO vs. MA HMO plans, and they wanted to know if those differences were of different direction or magnitude across the categories of low value services. But they presumably did not want to have more than two “key questions.” So, those got relegated to secondary status, and were therefore unwelcome for inclusion in the “findings” or “abstract” sections. Presumably to conform to the requirements of the particular statistical methods they used for significance testing, they chose to report “adjusted absolute differences” in the utilization rates per 100 beneficiaries — even thought those absolute rates are of no intuitive meaning to virtually all readers. Percentage differences are far more intuitive when the reader is just trying to inform policy decision-making by getting a sense of the magnitude of the effect that MA PPO and MA HMO plans might have on reducing waste on low value services of different types.

One interesting finding is that both PPO and HMO Medicare Advantage plans perform poorly compared to Traditional Medicare on use of low value diagnostic and preventive testing, about which the authors offered no explanatory theories. MA PPO plans had 37% higher use and MA HMO plans had 50% higher use compared to Traditional Medicare. Perhaps these low value services are a side effect of increased primary care vigilance and visit frequency, leading PCPs to order low value services that are not sufficiently economically consequential to justify utilization management attention by the plans. The study was not intended to answer that question. But, perhaps more importantly, when one of the service categories that makes up an overall metric behaves so differently than the other service categories, to the point that it partially cancels out observed differences in the overall measure, it makes curious minds wonder if the services are really alike enough to treat as being part of the same “population” of services within the aggregated “total low value services” metric. The rigors of JAMA’s science-based framing preclude the authors from presenting information from any less formal information gathering they might have done to shed light on that issue or from offering comments that might come off to JAMA reviewers and inappropriately speculative.

Therefore, the purpose of this blog post is to share my re-do of the results presentation as I think it should be presented to be most useful for the purposes of supporting policy decision-making, without the misapplied limitations imposed by the publisher who demands pretending it to be for purposes of science.

—

In a paper published in the August 23, 2024 issue of JAMA, Claire Boone and Anna Zink, from the University of Chicago Booth School of Business (my MBA alma mater!), collaborated with Bill Wright and Ari Robicsek from the Providence Research Network to disclose something that usually remains a secret. 

They analyzed employment contracts of 809 primary care physicians (PCPs) affiliated with a health system (presumably part of Providence) and the value-based payer contracts relevant to the patients attributed to those PCPs.  The payer contracts included commercial payers, Medicaid and Medicare.  The authors counted the number of unique quality metrics to which each of the PCPs was being held contractually accountable.  By “unique” they meant that if a particular quality metric appeared in multiple contracts, it was counted once.  They found that the PCPs were held accountable under an average of 7.6 value-based contracts, and the PCP were accountable for an average of 57 unique quality metrics, with a range from 0 to 103 unique metrics.  Each Medicare contract contained an average of 13.4 metrics, compared to 10.1 for commercial contracts and 5.4 for Medicaid contracts.

The apparent purpose of the study was to call attention to what the authors referred to as “saturation of the quality measure environment.” The authors asserted that a clinician would be unlikely to be able to “reasonably optimize” against 50 or more measures at a time.

But, as noted in a published comment I wrote for JAMA, for almost three decades, I have disagreed with the premise that we need fewer measures.  On the contrary, I believe that we actually need a lot more measures to finally convince health care provider organizations to stop trying to “study for the test” and to actually learn the subject and improve the core processes that drive health care quality.

The Outcomes Management Movement

Back in the 1990s, I served as both a member and a technical advisor for the Committee on Performance Measurement (CPM) of the National Committee for Quality Assurance (NCQA), the committee that oversaw the HEDIS standards, which were the dominant quality measures at the time.  I was on the CPM as the representative of the American Group Practice Association (AGPA), which has since been renamed the American Medical Group Association (AMGA).   During our lengthy committee meetings in Washington and elsewhere, we had endless debates about competing priorities regarding the future of the HEDIS standards. 

We were concerned that the dominance of process of care metrics could be construed as micromanagement and overreach on the prerogatives of health care professionals and provider organizations to set their own guidelines and policies regarding evidence-based care.  The AGPA was an early advocate for the use of measures of patient experience and outcomes, collaborating with InterStudy and later the Health Outcomes Institute in the development of “TyPE specifications” – condition-specific outcomes measures.  Julie Sanderson-Austin, VP of Quality Management and Operations of AGPA enlisted many AGPA member medical groups to refine and test the TyPE specifications.  At Henry Ford Health System, we implemented TyPE specifications for hip and knee replacement, cataract surgery, diabetes, and asthma, and we organized the SCORE Consortium to develop and test outcomes measures in 6 multi-specialty group practices.  The thinking was that providers should be held accountable for outcomes, so we could be free to innovate on different treatment and care management strategies to achieve improved outcomes.   

Although I was an early advocate for outcomes measurement, I learned that there were two big problems that would make it infeasible to use outcomes survey data as a basis for accountability. 

• First, the measures were too delayed.  For example, we measured whether patients experience a reduction in bodily pain one year after a hip replacement surgery. Including the pre-operative baseline and the one-year follow-up, and including additional time to analyze and report data, outcomes data is always a year and a half delayed.  Five year cancer survival outcomes are delayed at least five years. 
• Secondly, the subjective experience measures had far too much variance. Even for common surgical procedures, the sample size at the level of individual physicians or practice locations was just too small to overcome random noise. 

Long delays and high variance are fine for large multi-center studies to evaluate which treatments work, but they don’t really work for physician-level accountability, no matter how lovely the idea seems on the surface.

There was also great concern among the HEDIS committee members and within the medical group community that quality and outcomes measurement might become too burdensome.  This sentiment, which seems to be shared by Boone et al in the recent JAMA paper, was that we needed to reduce the number of measures to which physicians and physician organizations were to be held to account for two reasons; to reduce the expense of measurement and improvement, and to allow physicians and organizations to focus their efforts on a small number of metrics to increase the chances of “moving the needle” to improve quality. 

The Quality Indicators Movement

But, even back then in the late 1990s, there was a dissenting school of thought, most eloquently articulated by Beth McGlynn, a PhD health services researcher from RAND Corporation who was serving as one of the other technical advisors to the HEDIS committee.  (Beth is currently VP for Research at Kaiser Permanente and executive director of the Kaiser Permanent Center for Effectiveness and Safety Research.)

Long ago, driving through the Arizona desert with Beth during one of our HEDIS committee trips, Beth explained her point of view that we should not think of specific process quality measures as objects of our focus in improvement initiatives.  Rather, we should treat the measures as merely “indicators” of the performance of the underlying care processes.  Beth recognized that health care quality was the sum of doing a bewildering number of small things right.  Yes, we needed to remember to order mammograms and pap smears.  But those are just two of perhaps five dozen clinical preventive services that we needed to deliver more consistently.  Beth’s team at RAND went on to develop the RAND Quality Assessment Tools system, including a whopping 439 “indicators of quality of care” covering 30 acute and chronic conditions as well as preventative care.  If one were to propose a panel of measures that large today, peoples’ heads would explode. They tested these measures through the Community Quality Index study, which was done as part of the Community Tracking Study conducted by the Center for Studying Health Systems Change.  Beth and her many collaborators published the results in a famous paper in the New England Journal of Medicine in 2003, in which they reported that 6,712 study subjects from 12 metropolitan areas received only 54.9% of the “recommended care” specified in the quality indicators, a result that was widely referenced as a call to action to improve healthcare quality.

Core Process Focus

Beth’s thinking resonated with my own observations within the Henry Ford Health System and the Health Alliance Plan.  We were one of the pilot sites that tested early versions of HEDIS, including the mammography quality measure.  In the Health Alliance Plan, we tried to improve mammography quality through a program to offer health plan members a HAP-logo gym bag if they got a mammogram and sent us a card saying they had done so.  I wondered if we were going to offer gifts for every preventive service. Perhaps baseball caps for pap smears?  Within the Henry Ford Medical Group, our radiology department tried to improve mammography quality by creating a new “Mammo-Track” computer system.  I wondered if we were going to create separate tracking systems for every radiologic test. What about laboratory tests?  Like Beth, I believed strongly that we were thinking too narrowly – studying for the test, rather than really trying to think through how we could improve the performance of our care processes across the board.   If we focused only on a small number of measures, even if we were successful in achieving dramatic improvements, we would only be addressing a tiny slice of the improvement opportunity space.

From my perspective, the only feasible way to address a sizable portion of the opportunity space is to recognize that there are a smaller number of “core processes” for clinical decision making and care delivery that drive overall quality of care.

Some examples:

• During every encounter, there is (or should be) a core process to determine which preventive services and treatment monitoring interventions are due and to assure they get added to the care plan.
• Between encounters, there is a core process to determine when such interventions are due and to initiate outreach if there is not an upcoming encounter scheduled.
• When any item is added to a care plan (an order for a test, a referral for a consultation or procedure, a prescription, or even “return to clinic in 6 weeks”), there is a core process to track the care plan item to completion, escalate and remediate if it gets stuck, and initiate necessary follow-up plan of care changes.   
• There are also core processes for ambulatory encounters, abnormal test result follow-up, care setting transitions, and care relationship transitions, among others. 

Over the years, I’ve developed and patented technology focused on enabling these core processes, and I’ve introduced these concepts to many health care organizations and clinically-integrated networks.  The notion that we should focus our improvement efforts and IT investments on fundamentals rings true to many health care leaders, although I’ve also observed that the urge to focus on a small number of narrowly-defined opportunities is a “zombie-idea” – an idea that will not die, no matter how many times it is disproved.

In this context, it may be the case that contractually holding professionals and provider organizations accountable for a larger number of metrics – Beth McGlynn-style – might actually be helpful to force clinicians to realize that the metrics are merely “indicators” of the performance of the underlying core processes, and that the only way to achieve good overall performance is to establish and continuously improve those core processes.

We need to stop just studying for the test (narrowly focusing on a few things being measured) and start really learning the subject and improving core processes.

The popularity of shifting site-of-care

For decades, initiatives to encourage shifting care from hospitals to lower-priced non-hospital settings have been popular among health plan medical directors tasked with “bending the trend” in health care benefit costs.   In response, physician groups and other entrepreneurs have constructed ambulatory care facilities such as ambulatory surgery centers, urgent care centers, infusion centers, post-acute care facilities, medical imaging centers, and various types of home care services which offer lower prices than what hospitals charge.   This post will explain why the seemingly straightforward “grocery store” approach to calculating savings from health plans’ site of care shifting initiatives is deceptive because it fails to account for fix costs, a type of “squeeze the balloon” fallacy.

Hospitals offer both inpatient and ambulatory services within their facilities.  Therefore, the shift of care from hospitals to non-hospital settings includes two main types of shifts: (1) shifting inpatient hospital care to ambulatory settings, and (2) shifting care from hospitals’ ambulatory care departments to other ambulatory settings, such as shifting an ambulatory surgical procedure from a hospital operating room to an independent ambulatory surgery center, or shifting a chemotherapy infusion from the hospital’s infusion room to an independent infusion center or possibly to a home setting.

Health plans encourage the shifting of services to those non-hospital facilities through both patient-facing and physician-facing incentives.   Patients can be motivated to seek care from lower priced facilities through deductibles, coinsurance and copayments.  Physician pay-for-performance initiatives and accountable care gain-sharing initiatives can motivate physicians to refer their patients to non-hospital settings. 

Unlike prior authorization programs that aim to reduce the utilization of services, site-of-care initiatives do not involve fighting with doctors and health plan members about whether or not a service is medically necessary.  You just have to overcome objections that the quality or safety of care might be higher in a hospital setting because it might be easier to access emergency care if there are complications and easier to coordinate care.  When proposing to shift inpatient hospital services to non-hospital settings, those objections are strong, and one could expect that only a minority of inpatient care could safely be shifted, at least with today’s telemedical capabilities.  When proposing to shift hospital ambulatory care to non-hospital settings, those objectives are relatively weak, and one could expect that the majority of care could be safely shifted to settings that charge less.

Site of care shifting initiatives are also attractive because the savings seem straightforward.  Unlike care management initiatives, for which savings calculations are complex and require counterfactual assumptions regarding what would have happened without the care management intervention, savings from site of care initiatives seems on the surface to be far simpler.  You just multiply the volume of services that shifted by the price difference – just like calculating how much you saved from buying the generic rather than the brand name product at the grocery store, right?

No, not right, at least not in the long term.

New McKinsey / Harvard / JAMA analysis says site-of-care shifting could save billions

Before I explain why, let me first summarize a recent, prominent example of how to do it wrong.  In this week’s JAMA, Nihil Sahni and colleagues from McKinsey Consulting, Harvard, and Washington University published a paper entitled Potential US Health Care Savings Based on Clinician Views of Feasible Site-of-Care Shifts, for which I submitted a published comment. The authors had two aims. 

The first aim was to address the objection I described above about the quality of care or safety differences that might make it inappropriate to shift care to “alternate care sites.”  They interviewed 1,069 clinicians, 75% of whom were physicians with the remainder being nurse practitioners, physician assistants, nurse anesthetists, radiology and imaging techs and clinical psychologists.  They asked about 312 services (“care activities”) and 16 sites of care grouped into seven types, shown below.

They showed the clinicians the current distribution of services across the sites of care and asked them to propose an alternative distribution of care (still adding up to 100%) that could be safely achieved.  The clinicians’ responses suggested that, among the 33-38% of care activities delivered in inpatient hospital settings, a little more than 10 percentage points could be safely shifted from inpatient hospital settings to various ambulatory care settings.  The proposed shifts were higher for services that involved procedures, rather than just evaluation and management.  The clinicians proposed the greatest shifts for certain specialties, including physiatry and rehabilitative medicine, radiology, general surgery, and orthopedic surgery.

In my opinion, this strikes me as an impossible mental exercise for clinicians to do intuitively, given the diversity of cases, risks and capabilities across the sixteen sites of care.  Nevertheless, I suppose it is always a good idea to ask frontline people what they think.  To take this cognitive challenge to an even more ridiculous extreme, they also asked the frontline clinicians to imagine advancements in telemedical technology that might occur in the next 7-10 years, and to opine on a safe distribution by site of care that those advancements might enable.  I must have missed the class in medical school in which we were trained how to figure that out in our heads!

The authors’ second aim was to estimate how much money would be saved if those proposed shifts were implemented.  They used the traditional method.  For each service, they used claims data to calculate the average “allowed amount” – the amount that Medicare or a commercial health insurance plan was contractually obligated to pay for each alternative site, without considering adjustments and without netting out subsequent incentive payments.  Since they only had claims data for Medicare and commercial payers, they extrapolated by 0% to 50% to estimate the impact on shifts for people covered by other payer classes.  They took the average of the clinician-proposed distributions by site and compared them to the current actual distributions to determine the magnitude of site-of-service shifts.  They then calculated the savings multiplying the volume of cases shifted by the associated price differences.  Grocery store logic.  They calculated the total annual savings opportunity from implementing the proposed site-of-care mix to be between $113.8 billion and $147.7 billion, including an overall 5.2% net savings for commercial and 8.9% net savings for Medicare.  Based on the “imagine future technology” interview results, they estimated that the savings opportunity would grow over the next 7-10 years to 10.2% of commercial and 13.9% of Medicare cost.

Why grocery store logic is wrong

At the end of the day, health care organizations must charge prices that cover their costs and provide a little margin to allow the organizations to make necessary capital investments over time.  If an organization is for-profit, it must charge prices that also provide a reasonable return on the investment made by their owners.

Consider the example of an elective orthopedic surgical procedure.

Patients can be admitted as in inpatient and receive the procedure in the hospital operating room, then in the recovery room and then receive some additional recuperation and monitoring upstairs in a hospital room.  Or, they could receive the service as an outpatient in the same operating room, recover in the same recovery room, and then go home with instructions about what complications to look out for.  Or, they could receive the service in an ambulatory surgery center, with a similar operating room and recovery room.  In all three sites of care, the patient arrives at the facility, walks by a security guard, and spends a little time with a staff member in registration to enter some data into a computer.  The doctors and nurses spend about the same amount of time with the patient in the pre-op room, the operating room and the recovery room examining the patient, executing the procedure and typing notes and orders into their computers.  The operating rooms are heated to the same temperature.  The same pre-op and post-op tests are ordered.  The same surgical supplies are consumed.  In all three sites of care, the operating room has to be cleaned after the procedure.  If the site of care was inpatient hospital, a floor nurse spends some additional time checking in on the patient periodically, and there are some additional “hotel” expenses to heat and clean a room with a bed and a chair and to provide a few trays of food at a cafeteria level of quality.  No wine list.  Compared to the compensation for the orthopedic surgeon and all the costs incurred in pre-op, the O.R. and recovery, the compensation for the floor nurse check-ins and the hotel costs are small.  If the health care organization sets prices to cover their costs and provide a little margin, one would expect the prices to be about the same for hospital outpatient and ambulatory surgery facility settings, and only a bit higher for the hospital inpatient setting.

So why are the prices charged to the health insurance company so different among these three alternative sites of care for the same orthopedic surgical procedure? The main answer is hospital fixed cost allocation.

Hospital campuses have substantial fixed costs.  The hospital accounting team must keep track of all expenses and revenues.  The hospital departments that deliver care, like the inpatient nursing floors, the operating rooms, the emergency room, and the laboratory and radiology departments, are “revenue centers.”  The revenue centers have associated cost accounts to keep track of the “direct costs” of the clinical operations that provide the care.  Many of those direct costs are to pay for staff time and supplies that are “variable” based on the volume of care that is provided during a month.  But some of those direct costs are “fixed” such as the salary cost of the head nurse in the operating room.    In addition to direct fixed cost, hospitals also have substantial indirect fixed costs, also described as “overhead” or just “indirect” cost.  Hospitals have large buildings and grounds.  They must buy and maintain all the equipment needed to provide a broad set of services.  They have a big IT department and have to pay substantial fees to software vendors.  They have finance people, a legal department, a human resources department, a security team, and a team that recruits medical professionals and handles credentialing and continuing medical education.  They might have to pay staff to administer residency programs.  All of these costs are tracked in indirect cost centers. 

When a hospital finance team is negotiating the fee schedule and other reimbursement contract terms, they need to make sure that the revenue they project based on the contract terms is more than sufficient to cover all their direct and indirect costs.  One way they do that is to calculate for each of the revenue centers a “contribution margin” by subtracting direct variable costs from revenue, and then making sure that the total contributions of all the revenue centers can more than cover all the direct fixed and indirect fixed cost.   Another way the do that is to allocate the indirect fixed costs and direct fixed costs to each of the services offered within each of the revenue centers.  For example, the allocation for security services and building maintenance services might be done in proportion to the square footage of the space occupied by each clinical department.  Then, the “fully loaded cost” for each service is calculated as the sum of indirect fixed, direct fixed and direct variable cost, divided by the units of service.  Then, the finance team can compare the fully loaded cost to the projected revenue to determine if the service is profitable, and they can adjust the prices as needed to ensure economic sustainability of the enterprise.

So, what actually happens if you shift care from inpatient to outpatient settings, or if you shift ambulatory care from hospital campuses to non-hospital providers?

Over a relatively short period of time, the variable cost should shift from one site of care to the other.  But the fixed cost is, um, fixed.  You could continue to allocate the same amount of fixed cost to the revenue center from which care was shifted and raise the prices for the shifted services to make up for the smaller volume, thereby creating an even stronger incentive to shift more of that care away from the hospital.  Or, more likely, you end up allocating more fixed cost to all the other services, and raising the prices of all those other services a little bit to cover it.  The same thing happens to costs for uncompensated care provided to patients without insurance that cannot afford to self-pay, and for losses from services provided to patients covered by Medicaid, which systematically under-reimburses for care.  Unless the community and the health plans are OK with the closing of the hospital, the prices across the board must increase to cover the costs which are truly fixed.  It is a classic case of squeezing the balloon.  The fixed cost air inside has to go somewhere.

Site of care shifting is beneath us

One could argue that, from the perspective of one payer, if you can shift more care away from the hospital than your competing payers, you get to keep the savings from the price difference, but the compensatory increases in prices for all the other hospital services might fall slightly disproportionately on the competition, giving you a slight strategic cost advantage.  That’s the same kind of rationalization that is used when health plans spend lots of time and money on programs to game diagnosis coding slightly more than their competition for patients covered by Medicare Advantage and other risk-based reimbursement arrangements.   The point is that, when health plan medical directors focus their time on site-of-care shifting (as implicitly advocated by the McKinsey/Harvard/JAMA paper), they are unintentionally deceiving themselves about the savings and participating in a race-to-the-bottom that does not add real long term value. 

In fact, the push to shift care may achieve the opposite.

Health plan initiatives to shift care incentivize the construction of new non-hospital facilities for ambulatory surgery, physical therapy, imaging, infusions, etc.  All those facilities have their own security guard and check-in staff and many other duplications of hospital fixed costs. By spreading care among many facilities, there is a loss of opportunity to achieve savings from economies of scale, specialization and care coordination.  If the new facilities have smarter leaders or use more efficient care processes compared to hospital organizations, then by all means, we should replace the hospital leaders and adopt the innovations.  But intentionally promoting the duplication of infrastructure to create the appearance of savings by evading hospital fixed cost allocation, when the health plans are still paying all the fixed costs one way or the other in the long run, should be considered beneath those on a professional mission to increase population health care value.

To achieve true improvements in value, it is necessary to do the hard work of figuring out more effective and efficient ways to deliver necessary health care services and convincing people to stop delivering low value services.  Admittedly, such work can generate controversy and conflict, particularly since people tend to resist change and defend their interests.  And, such work admittedly requires more complex analytics and modeling than the grocery-store logic used for site-of-care shifting.  But population health professionals can and eventually will overcome those barriers.