PANDAS

PANDAS: An acronym for Pediatric Autoimmune Disorders Associated with Strep.

PANDAS is the sudden onset of symptoms like those of obsessive-compulsive disorder or Tourette syndrome following infection with streptococcus bacteria, caused by an autoimmune reaction that affects the basal ganglia in the brain. If tics are seen, they may or may not be choreaform like those of Sydenham chorea, a closely related condition that can follow a bout of rheumatic fever, which is caused by the streptococcus bacteria.

Diagnosis is primarily by observation, although a blood marker for the disorder has been identified. Treatment is by plasmapheresis or intravenous immunoglobulin, prophylactic antibiotics, and/or medication for specific symptoms.

MedTerms (TM) is the Medical Dictionary of MedicineNet.com.
We Bring Doctors' Knowledge To You

Modifying Hospital Community Benefit Tax Policy: Easing Regulation, Advancing Population Health

Editor's note: This post is part of a periodic series of Health Affairs Blog posts discussing the Culture of Health. In 2014 the Robert Wood Johnson Foundation announced its Culture of Health initiative, which promotes health, well-being, and equity. These blog posts are being run in conjunction with the November 2016 issue of Health Affairs, a thematic issue on the Culture of Health which explores roles for individuals, communities, commercial entities, and public policy that extend beyond the reach of medical care into sectors not traditionally associated with health. 

How might regulatory tax policy constraints be eased while promoting greater health equity? The Trump Administration could take steps-under existing legislative authority-to broaden the longstanding definition of “community benefit” spending to promote fuller community-wide health improvement partnerships between tax-exempt hospitals and the communities they serve. These steps, discussed below, are described at length in a new report issued by the Milken Institute School of Public Health at the George Washington University and funded by the Kresge Foundation and the Robert Wood Johnson Foundation.

Origins And Evolution Of Community Benefit Policy

Under policies first adopted by the Internal Revenue Service (IRS) in 1956 under Revenue Ruling 56-186, nonprofit hospitals could qualify as tax-exempt charities if they “operated to the extent of [their] financial ability for those not able to pay for the services rendered and not exclusively for those who are able and expected to pay.” In other words, charity care was the basic prerequisite to tax exemption.

In 1969, the Nixon Administration replaced this earlier standard with Revenue Ruling 69-545, which recognized the promotion of health for the benefit of communities as a whole as a charitable purpose. Under the ruling, hospitals could eliminate their charitable care activities altogether without losing their tax-exempt status as long as they could demonstrate a benefit to their communities as a whole.

In 2009, the IRS first issued a special reporting schedule (Schedule H) that must be completed annually by exempt hospitals as part of their Form 990 filings. This schedule provides guidance on what constitutes community benefit spending.

The Community Benefit/Community Building Conundrum

Part I of Schedule H contains a relatively detailed definition of community benefit spending. To a considerable degree, the Schedule H definition represents something of an evolution from the activities first classified as community benefit spending by the IRS in its 1969 revenue ruling: financial assistance (often termed charity care) for patients, with considerable hospital discretion to define the terms of assistance; hospital expenditures to offset losses in connection with participation in Medicaid and other means-tested government health insurance programs that pay less than the cost of care; community health improvement and community benefit operations; health professions education; community-wide subsidized health care such as regional trauma units; and research. The term “community health improvement” is defined as “activities or programs, subsidized by the health care organization, carried out or supported for the express purpose of improving community health. Such services do not generate inpatient or outpatient revenue, although there may be a nominal patient fee or sliding scale fee for these services.”

Separate and apart from community benefit spending activities, Part II of Schedule H permits hospitals to report expenditures for certain “community building” activities. Part II activities encompass physical improvements and housing, economic development, community support, environmental improvements, leadership development and training for community members, coalition building, community health improvement advocacy, workforce development, and other activities.

The IRS reporting instructions note that Part II expenditures may be reported as community health improvement spending under Part I. At the same time, however, the same instructions note that Part II expenditures are “not reportable” under Part I. Furthermore, the instructions fail to provide guidance regarding the circumstances under which Part II spending can be reported under Part I and therefore count as support for the hospital's tax-exempt status.

Hospital organizations have attempted to give examples to their members regarding the types of community building expenditures that might qualify as Part I community health improvement activities. Moreover, on occasion, and in response to specific queries (related to housing), the IRS has recognized community building as community health improvement. But the ambiguous IRS instructions create uncertainty regarding what community building activities qualify; indeed, the very existence of Part II effectively creates a presumption of exclusion of some or most community building activities. One might infer from the community benefit/community building distinction that activities focused on patient care (e.g., free lead tests for children) are reportable as community benefits, while activities that help community residents who are not patients (e.g., installing water filters in homes) amount to community building with a presumption of exclusion from Part I. This creates a disincentive for hospitals worried about maintaining their tax-exempt status to engage in the latter set of activities.

Hospitals As Community Health Actors: Eliminating The Community Benefit/Community Building Distinction

Health system transformation, spurred on by the Affordable Care Act (ACA), creates an opening for a different approach to defining community benefit, one that the IRS could adopt under its current legislative authority, just as it has done over the years in establishing the concept of community benefit itself. Using its authority, the agency could eliminate the community benefit/community building distinction, thereby incentivizing hospitals to move beyond patient care and assume a far more forceful role as full participants in efforts to improve the health of their communities.

Such a move would be consistent not only with the growing focus on underlying social conditions that affect health, but also with the “Triple Aim”-reducing the cost of health care while improving quality of care and population health-which has been part of the broader health policy landscape for years. This aim cannot be achieved by insurance alone. The Affordable Care Act has reduced the uninsured from 49 million to 29 million, from 16 percent to slightly more than 9 percent of the population, and its delivery reforms are structured to promote efficiency and incentivize hospitals and other health care providers to take a more holistic approach to patient health. While the ACA's insurance reforms face an uncertain future, to the extent that hospitals feel the effects of improved coverage, they can begin to focus on the broader community-wide health improvement role the Nixon Administration originally envisioned.

To be sure, the continued need for charity care remains great. This is particularly true in non-Medicaid expansion states for lower-income people who would be Medicaid-eligible in expansion states; it also holds true for privately insured patients, covered through individual and employer plans alike, who face steep, escalating cost sharing. But despite this continued need for financial assistance in connection with hospital care-which was established under the Internal Revenue Code as a basic condition for tax exemption (see page 1961)-the expansion of insurance can be expected to have a positive impact on hospital financing.

This positive effect can be seen in a 2015 study, which estimated a $6 billion decline in uncompensated care among hospitals in 2014. A separate study shows uncompensated care reductions as well, particularly in Medicaid expansion states. Added to potential gains from insurance expansion are payment incentives designed to reduce unnecessary readmissions and shorten the length of stay. These two developments, taken together, elevate the case for a hospital strategy that looks beyond the hospital door at the social conditions of health.

Tax Policy As Health Policy: Broadening The Meaning Of Community Benefit

The IRS possesses the power to broaden the definition of community benefit by eliminating the community building/community benefit distinction that the agency itself has created. Such a move-a recognition of hospitals' potential role in community-wide health improvement-would come at a time of transformation of hospital mission and practice and of growing awareness of the importance of the role of hospitals as community health anchors. We believe the IRS could take several actions to achieve this broad goal.

Eliminate The Separation Between Part I And Part II Of Schedule H

First, the IRS could eliminate the wall between Part I and Part II of Schedule H in terms of what the agency classifies as community benefits. This would encourage hospitals to contribute to clearly identified, high-priority health needs identified by hospitals themselves as part of their community health needs assessments (CHNAs), required by the tax code. The CHNA process emerges as a critical guide to such community-wide health improvement efforts. CHNAs must be conducted with input from the community served as well as public health experts and must be accompanied by annual implementation strategies that demonstrate how a hospital will respond to the needs that have been identified.

The George Washington University's (GW's) review of the most recently available CHNAs shows that, indeed, hospitals are themselves identifying population health conditions whose amelioration depends less on timely medical care and more on changing the conditions under which children grow and adults live, work, and age. By eliminating the Schedule H Part I/Part II distinction, the IRS could encourage hospitals to position themselves as full participants in community-wide health improvements that do not depend on a provider-patient relationship.

Identify Promising Community-Wide Health Improvement Activities

Second, the IRS could issue comprehensive guidance to help hospitals identify those community-wide health improvement activities that hold real promise. To do so, the agency clearly would need the help of public health experts, just as the agency turned to these experts to develop its CHNA rules. This expertise could be provided through a government-wide advisory group that could guide tax regulators on promising health improvement efforts, while also suggesting criteria that hospitals might apply as they seek to introduce additional community health improvement innovations.

Typically the IRS waits until it receives queries regarding the lawfulness of particular practices, such as supportive housing. A far better course would be for the agency to affirmatively engage with experts across the government-drawn from the Departments of Agriculture, Education, Health and Human Services, Education, Housing and Urban Development, Transportation, Veterans Affairs, Labor, Commerce, and other agencies-in developing community health improvement policy. The National Prevention Council, whose mission is to promote lifelong health, potentially offers such a starting point for the agency.

Establish Metrics To Guide Reallocation Of Hospital Community Benefit Spending

Third, the IRS could establish basic metrics to guide hospitals as they consider reallocation of community benefit spending. Federal law establishes no minimum community benefit spending floor, but hospital community benefit spending is considerable. According to a January 2015 IRS Report to Congress on Private Tax Exempt, Taxable, and Government-Owned Hospitals, in 2011, hospitals governed by the community benefit spending requirement spent $62.4 billion on community benefits. About half went to charity care and Medicaid shortfalls; only about 4 percent of this total (about $2.7 billion) was spent on Part I community health improvement activities, a figure that also includes expenditures on hospitals' own community benefit administration operations.

Moving forward, expenditures to support direct patient care undoubtedly will and should continue to play the major role in community benefit spending. But this leaves considerable room for spending on a far broader set of activities that can improve community health and that are undertaken in response to health priorities identified through the CHNA process. The IRS could set as a specific goal the greater alignment of community benefit spending and the priorities set through hospital CHNAs; the agency could offer suggested performance measures by which hospitals could, over time, slowly begin to reallocate their own community benefit spending toward these priority needs. Such actions could help ensure that as hospitals realize gains from broader insurance coverage, their contributions to community health improvement grow, and community benefit spending does not shrink.

In sum, eliminating the distinction between community building and community benefit is the essential first step. For maximum effect, however, this step should be coupled with broad guidance on community health improvement strategies that show promise, as well as guidance for hospitals that could, over time, help them reallocate spending toward CHNA-established priorities.

There are many efforts to improve the health of communities that will require new legislative authority coupled with increased spending. But the transformation of community benefit policy is not one of them. These reforms can be set in motion through executive action, led by the agency that oversees federal tax policy but deeply informed by the many sources of public health expertise within government. This approach to community benefit policy offers the potential for integrating tax and health policy in new ways, so that federal policy levers are not only being pulled, but pulled in the same direction.

Authors' Note

Funding from the Kresge Foundation and the Robert Wood Johnson Foundation support the report on which this blog is based.

Science in the Age of Trump

By DAVID SHAYWITZ, MD

As Donald Trump seeks to build his administration, he will likely struggle with creating a science infrastructure, given his estrangement from the nation's scientific community.

The distance between Trump and scientists seems to reflect mutual disdain. Trump famously trusts his gut over more data-driven methods. Trump's success, against most expectations, can be read as a triumph of instinct over science, or at least to reveal the perils of data-driven overconfidence. Trump's apparent intention to appoint a climate-change skeptic to lead the environmental protection agency suggests to many a disregard for the vast weight of scientific data here, while his comments about vaccinations during one of the Republican debates were charitably described by Steven Salzberg as “wildly inaccurate” and “thoroughly discredited.”

For their part, most scientists take a very dim view of Trump: Science prides itself on being inclusive, international, objective and collaborative–not generally the first adjectives used to describe Trump. (Whether science in practice lives up to these ideals is another question.) Add to this Trump-specific distaste the left-leaning bias of universities (well documented by centrist academics like Jon Haidt, among others–see here and references therein), and the result is a community that seems solidly united against the President-elect. (That said, I'd note that I've met more than a few political independents within academia who chafe at current norms.)

The problem, of course, is that science matters to America and to Americans. The U.S. scientific ecosystem has advanced knowledge and powered a range of industries, from biotech to aerospace. America has a wonderfully robust tradition of supporting scientific exploration, and it's hardly surprising–but worth noting–how many great scientific advances, including many resulting in Nobel Prizes, were made by researchers originally from other countries attracted to the United States by the freedoms of our nation.

Many Trump critics argue passionately against participating in the Trump Administration, essentially suggesting that joining the team would help legitimize an unworthy (some would say execrable) regime. (In a small, informal and unscientific Twitter poll I did over the weekend, about a quarter of respondents were in this category.)

The alternative view is that in a Trump administration, it's especially important for rigorous scientists to participate and be heard–provided, of course, that their voices would be heard, rather than their prestige co-opted. Because so few scientific innovators seem to be drawn to the Trump team, those who participate may find themselves–like Silicon Valley VC Peter Thiel – with an opportunity for influence far greater than they would have had in a more traditional administration.

To this end, one name that immediately comes to mind as a potential biomedical advisor or potential appointee is Dr. Jeffrey Flier, who recently stepped down as Dean of Harvard Medical School, after nearly a decade in that role.

(Disclosure: I know Flier as an endocrinologist and colleague, but do not have a personal or business relationship; I do have a long history with the university, including a current adjunct/visiting scientist appointment.)

In addition to deep expertise in medicine and science, and a range of honors including elected membership in the National Academy of Medicine, Flier has three attributes that would presumably appeal to the Trump administration.

First, Flier has encouraged the university to embrace entrepreneurship while remaining true to its academic foundation–an incredibly difficult needle-threading challenge that he somehow seems to have pulled off, a testament to his vision, independent thinking and political aplomb.

Second, Flier's striver background is arguably similar to Trump's: on the way to becoming Dean at Harvard, Flier went to City College of New York, and received his MD and subsequent training at Mount Sinai School of Medicine. He rose through the ranks at Harvard's Beth Israel Hospital, which might be viewed as Boston medicine's version of coming from the Outer Boroughs (see this Samuel Shem classic for more detail).

Finally, Flier's love of medicine is clearly part of the family business; his wife is also a professor of medicine at Harvard, his brother as an internist in the Boston area, and his two daughters are physicians as well.

While Flier has suggested on Twitter he has no interest in being considered for a role in the new administration, Trump's team would do well to do whatever they could to bring him into the fold.

Our nation's continued leadership in biomedical science is vitally important for our collective future. The new administration must seek to recruit the best talent–even NeverTrumpers, like Flier; similarly those who are tagged–even NeverTrumpers–must consider serving, for the benefit of the country.

The Current 21st Century Cures Legislation Is Still A Bad Deal For Patients

A lot can happen in 18 months.

In July 2015, the House of Representatives passed the 21st Century Cures Act, which contained public science funding and numerous provisions aimed at adjusting how the Food and Drug Administration (FDA) evaluates new medical products and new indications for existing prescription drugs and devices. Early in 2016, a Senate committee approved several individual bills that addressed some of the issues targeted by the initial House bill, but did not consider the full package of changes.

Finally, on November 25, late Friday over a holiday weekend, Congressional leaders released an amended version of the 21st Century Cures Act. After some final tweaks, the legislation was expected to pass in the House on Wednesday, November 30th. It would then move on to the Senate for consideration. At 996 pages, the revised bill (without the final tweaks) is 90 pages longer than the Affordable Care Act (ACA) and over 600 pages longer than the original bill. More than 1,400 lobbyists have worked to shape the proposed legislation behind the scenes.

Uncertain Funding

Whereas the original bill was lauded for providing National Institutes of Health (NIH) funding for the discovery of new medical advances, the current bill provides significantly less certainty in its support for prevention, research, and regulatory science. In 2017 and 2018, $500 million would be placed in a fund to help states combat the opioid epidemic. Over a 10-year period, $2.8 billion and $430 million would be placed in funds to support NIH research and FDA innovation, respectively. However, this funding would not be guaranteed; Congress would have to vote each year to actually make the money available as part of the appropriations process.

Additionally, most of the proposed NIH funding would come at the expense of cuts to the Prevention and Public Health Fund, which might otherwise vanish if the new Congress repeals the ACA next year.

Easier Pathways, Greater Risk

The current bill also retains some controversial provisions affecting how regulators evaluate certain types of medical products. Despite the insertion of clauses that purport to keep the FDA's approval standards intact, these provisions could be interpreted to afford the Health and Human Services (HHS) Secretary considerable discretion in applying those standards. The Act, for example, would require the creation of a program to use “real-world evidence” in support of new indications for existing drugs. Defined as information on drug outcomes that are derived from sources other than clinical trials, real-world data are less likely to have been uniformly collected and therefore risk being less reliable. Attempted analysis of real world information is potentially further plagued by systematic differences between populations of interest that may not be captured and, thus, cannot be adequately controlled, which can distort the true benefits and risks of a drug.

Another concerning provision involves a new “limited population” approval pathway for antibiotics treating patients with unmet needs based on a sliding benefit-risk scale, factoring the severity, rarity, or prevalence of the infection to be treated and the availability of alternative therapies. Manufacturers would be required to affix disclaimers flagging this pathway in the drug's labeling and to submit promotional materials to the FDA for inspection. But research has shown that consumers rarely read, let alone heed, health disclaimers, and the FDA has been increasingly hampered in its ability to regulate pharmaceutical promotion for unapproved or off-label uses.

Numerous sections in the current bill relate specifically to medical device regulation. For example, one section would formalize and expand an ongoing pilot program that entitles “breakthrough” high-risk medical devices to priority review, in which approvals could be made despite greater risk-benefit uncertainty. Yet, as defined in the current legislation, a device could be classified as a breakthrough even if its possible advantages are not “clinically meaningful.” Therefore, many new, high-risk devices would qualify for this pathway, and faster reviews of high-risk medical devices are associated with increased reports of safety problems after approval.

Such provisions will inevitably lead to widespread use of new and costly treatments before they are shown to work or before their risks to patients are adequately known. In many cases, government payers like Medicaid and Medicare would have no choice but to cover these treatments, while private insurers would be forced to play a more pronounced gate-keeping role.

Is It Worth It?

Of course, there are some positive provisions in this 996-page piece of legislation. The bill retains sections from the original legislation that would strengthen the FDA's ability to hire, train, and retain experienced staff scientists needed to evaluate greater and increasingly complex data. It also adds a requirement for manufacturers to publicly disclose how, if at all, they may make experimental therapies accessible to patients with serious or life-threatening conditions who do not qualify for clinical trials. Such transparency could help physicians and patients learn about the all-too-rare cases in which manufacturers offer such expanded access programs.

Still, we believe the bill's potential benefits and modestly enhanced funding for scientific advancement are far outweighed by the increased risk of patient harm, and added pressure on the FDA to rush new products to market without adequate evidence. The current 21st Century Cures Act would not drive innovation. Instead, it represents a poor deal for patients.

People Post: Staff And Board Changes At Health Foundations

Mary Backley, vice president for finance and administration at Grantmakers In Health (GIH), departed from the organization at the end of September. According to an announcement from Faith Mitchell, president and CEO, “Mary has been a core member of GIH's staff since 1993, helping us grow from a small circle of Funding Partners to our current 250+.” Backley was well known for event coordination-particularly the GIH Annual Conference on Health Philanthropy.

Meredith Sullivan Benton is the new senior program officer at the Healing Trust, a foundation in Nashville, Tennessee. (The funder was formerly called the Baptist Healing Trust.) This philanthropy funds nonprofits in forty counties in Middle Tennessee. Benton was previously chief of staff and assistant commissioner for external affairs at the Tennessee Department of Environment and Conservation. According to her biography, “her career highlights include serving in the office of a U.S. senator, two Governor[s'] administrations and for a former President's international organization.”

Lisa M. Fasolo Frishman has joined the Syracuse, New York, office of the Health Foundation for Western and Central New York. She is senior program officer and directs the funder's work in community health capacity and policy, according to an October press release. She was previously executive director of the NY (New York) Funders Alliance. Her past positions include program coordinator at Syracuse University's Maxwell School of Citizenship and Public Affairs.

Jake Grindle has been promoted to program officer I at the Maine Health Access Foundation (MeHAF). According to his biography, he “administers the Discretionary Grant and Meeting Support Programs and is helping to lead MeHAF's efforts to deepen its focus on health equity.”

Kristen Keely-Dinger, executive vice president of the Healing Trust, has been promoted to be its new president and CEO. A licensed advanced practice social worker, she will succeed Catherine (“Cathy”) Self, who will retire. Keely-Dinger will become the trust's new leader in December.

Elizabeth (“Beth”) McGlynn was named chair of the ABIM Foundation's board, according to a July press release. She is director of Kaiser Permanente's Center for Effectiveness and Safety Research.

Diane Oyler has joined the Health Foundation for Western and Central New York as a program officer. She started there in July. Oyler was previously with the Erie County (New York) Department of Senior Services. At the foundation, Oyler will focus on vulnerable older adults and caregivers. She continues as a lecturer at Buffalo State College in its public administration department.

According to an October press release, Robert (“Bob”) K. Ross, president and CEO of the California Endowment, will join the Weingart Foundation's board of directors, effective January 1. The Weingart Foundation, located in Los Angeles, California, funds in six Southern California counties. Read its announcement about its “Fiscal Year 2017 Program Plan.”

Brandon Skidmore has joined Sunflower Foundation, located in Topeka, Kansas, as a program officer, according to an October e-alert. Skidmore most recently was director of the Bureau of Health Promotion at the Kansas Department of Health and Environment. According to the e-alert from Billie Hall, Sunflower president and CEO, Skidmore will manage the foundation's Health Care program area, including its Integrated Care Initiative. The program focuses on “attainable and sustainable integration of primary and behavioral health care at the community and statewide levels.” Skidmore also will work closely with Sunflower's Healthy Living and Active Communities and its Advocacy and Policy program areas.

Charlie Venti, executive director of the Nicholson Foundation, located in Newark, New Jersey, will retire at the end of December. He “has led the day-to-day operations” of the foundation since 2010, according to his biography. His successor will be named soon.

Pamela Browner White is now senior vice president of communications at ABIM and the ABIM Foundation, according to an August press release. On the foundation side, she is supporting its mission to advance medical professionalism. For example, she will strive “to expand conversations about avoiding inappropriate or unnecessary tests and treatments” through the foundation's Choosing Wisely campaign. White most recently was senior vice president of corporate communications and strategic development at Esperanza Inc., “the largest Hispanic faith-based nonprofit” in the United States. Before that, she was vice president, public affairs, for Cancer Treatment Centers of America.

Honors

Ralph Fuccillo, president of the DentaQuest Foundation, received the 2016 Albert Schweitzer Prize for Humanitarianism. He was “honored for his work as a servant leader in public health and his lifelong commitment to health equity,” according to an announcement from the foundation. He received the award on November 4. Read the original announcement from the Albert Schweitzer Fellowship organization and an article in the Somerville (Massachusetts) Times for more information.

Grantmakers In Health announced in September that eighteen foundation staffers were named to the 2016 class of fellows for the Terrance Keenan Institute for Emerging Leaders in Health Philanthropy. The Aetna, Blue Shield of California, Empire Health, and Robert Wood Johnson Foundations fund the institute this year. Two of the new fellows have written for GrantWatch in the past. Jennifer Chubinski of Interact for Health (formerly the Health Foundation of Greater Cincinnati) coauthored a GrantWatch article, “For Two Regional Health Foundations, Returns from the Kentucky Health Issues Poll Are Worth the Investments,” in the September 2014 issue of Health Affairs. Kimalon Meriwether of the Cleveland (Ohio) Foundation wrote a July 2015 Health Affairs Blog post titled “A Perfect Combination: The Cleveland Neighborhood Model for Urban Health Care Education.”

Risa Lavizzo-Mourey, president and CEO of the Robert Wood Johnson Foundation (RWJF), was one of three people honored by the United Hospital Fund “for their work to improve health and health care,” according to a September press release. Lavizzo-Mourey “received the Health Care Leadership Award for her visionary work to build a national 'culture of health.'” She has announced plans to step down from her position at the RWJF. Read a GrantWatch blog post about this news.

Read Health Affairs' Editor-in-Chief Alan Weil's interview with Lavizzo-Mourey and Alonzo Plough, vice-president, research-evaluation-learning, and chief science officer at the RWJF, in the November issue of Health Affairs. The interview is titled “Building A Culture Of Health.”

Next Avenue's 2016 Influencers in Aging, who “continue to push beyond traditional boundaries and change our understanding of what it means to grow older,” included some names from philanthropy: Bruce Chernof of the SCAN Foundation, John Feather of Grantmakers In Aging, Terry Fulmer of The John A. Hartford Foundation, and Carol Levine of the United Hospital Fund.

What Does the Recent Election Mean For Predictive Analytics In Healthcare?

By ROBERT ROWLEY, MD

The outcome of the recent election caught many people, and many forecasters, by surprise. How could their predictions have missed the mark so significantly? Granted, there were a number of people who predicted the outcome more accurately, but many of those who used data models to analyze the likely outcome are left now with head-scratching and postmortem analysis in order to improve their methods.

In their book Superforecasting, The Art and Science of Prediction, authors Philip Telock and Dan Gardner describe a subset of people who, on average, are significantly more accurate in their ability to predict upcoming events. “What makes them so good is less what they are than what they do-the hard work of research, the careful thought and self-criticism, the gathering and synthesizing of other perspectives, the granular judgments and relentless updating.”

What does this mean for healthcare? I'm not talking about the impact of the new presidency on health policy and healthcare delivery (that's another discussion) – I'm talking about whether predictive analytics is really all that accurate in the first place. Where does it fail?

The strengths and weaknesses of predictive analytics

Predictive Analytics in healthcare, a buzzword in the industry for a couple of decades, is the science of determining which populations are likely to become ill, what the health and cost implications of that are, and what might be done by way of pre-illness intervention to change things. About 20% of the population consumes 80% of health care dollars. But those who are catastrophically ill this year are not necessarily the ones who will become catastrophically ill next year – the high-cost cohort, though a consistent finding year after year, will be comprised of different individuals each year. The goal of medical Predictive Analytics is to figure out who will likely drop into that high-cost bucket next year, and what can be done to reduce that risk.

Much of risk stratification (the core of medical Predictive Analytics) is focused on populations. Taking people with certain health risk parameters in aggregate, as a population, is something that is predictable, can be measured and studied, and is the basis of what we have now. But drilling that understanding down to an individual patient becomes much more uncertain. Should this diabetic patient, controlled with medications but not on a statin, and who has at-target LDL cholesterol levels for a diabetic – should this patient be prescribed statins anyway? Doctors will have differing opinions, will do different things, and will look to supporting data (which may be sparse in a more granular analysis) to justify their choices.

How can we get better at individualizing medical recommendations? How can we take the current state of Predictive Analytics, which concerns itself with population management, and move it forward to something more precise?

AI: the next step in prediction

This is where Artificial Intelligence (AI) in healthcare can be very powerful. AI is the intersection of Machine Learning (ML) – a set of self-teaching algorithms that can identify patterns in data without being pre-programmed on what to look for (therefore without “pre-analysis bias”) – and the application of that ML to very large data sets. The shortcoming of medical AI so far is not so much a shortcoming in ML algorithms, but is the lack of very large, normalized data sets on which it can work. Medical data (clinical data) is historically fragmented into institution-centered silos, and claims data is segmented into payer silos. Aggregating this data into huge data sets is the task at hand in order for AI to become meaningful.

From this effort, our Medical Knowledge Graph (MKG) can be extraordinarily useful. The Flow Health Medical Knowledge Graph is the organized result of AI insights built in a way that can be used on-the-fly by a variety of medical applications, such as Electronic Health Records, population management and reporting tools for value-based care, web tools, and patient-facing apps. For the patient described above, the individualized recommendation can be made for that given person, and take into account all the diagnoses, lab values, medications used and discontinued in the past, and genetic markers if known.

Does this technology, once it matures, make the doctor's role obsolete? No. It makes the doctor's role more precise, more accurate, more consistent. In clinical medicine, we use clinical judgement based on recognizing a pattern presenting in a given patient, and we try to match that against similar patterns from our learning and our experience. We then use that pattern-matching to make recommendations. In the case of AI and the MKG, the pattern can be described in more detail, and the comparison is done against the entire body of data available to the ML engine. It becomes a tool that can make clinical judgement much better informed.

Predictive analytics, and the AI tools now becoming available, predict the odds of success, or the odds of something occurring. They deal in probabilities. However, as noted by many forecasters, nothing is truly certain (until it happens). Failures of accurate prediction teach the learning engines. This is true in political outcome prediction, and it is true in medicine. Leaders, whether in government, in the military, in business, or in healthcare, need to be well-advised, but must make executive decisions. In healthcare we call that making a surgical decision (there are no erasers on the ends of scalpels). Clinicians need to be decision-makers, informed by the best analytics available. In health IT, we need to build the best analytics engines we can, so as to inform medical decision-making in the best way that technology allows.

Robert Rowley, MD is Chief Medical Officer at FlowHealth

Calculating Medicare Advantage/Fee-For-Service Price Differences Is Harder Than It Looks

Understanding whether fee-for-service (FFS) Medicare and Medicare Advantage (MA) plans pay different prices is important - but more difficult to determine than it seems. In “Medicare Advantage Plans Pay Hospitals Less Than Traditional Medicare Pays,” we reported that MA plans paid hospitals 8 percent less, on average, than did FFS for a standard basket of hospital services. Our colleagues at the Medicare Payment Advisory Commission (MedPAC) have since alerted us that we did not account for differences in the way that hospitals receive Indirect Medical Education (IME) payments for MA, as compared to FFS, beneficiaries. In this post, we investigate the extent to which these differences affect our key result.

The Problem

The fundamental problem is that payments to hospitals from FFS have IME included in them, but payments from MA plans do not. According to MedPAC, for each MA admission, CMS makes an IME payment directly to the hospital equal to what would have been included in the payment for the admission had it been for a FFS rather than an MA beneficiary. These IME payments for MA admissions are over and above whatever payments are made by the beneficiaries' MA plans and the beneficiaries themselves.

The Centers for Medicare and Medicaid Services (CMS) calculates the IME payments for MA admissions based on claims that hospitals are required to submit to CMS. These claims are sometimes called dummy or “no pay” claims, because they are separate from and in addition to the primary claims for payments that hospitals submit to MA plans. In our analysis, we did not include the IME payments made on behalf on MA beneficiaries through this separate channel. Thus, although we accurately calculated the difference in payments made by FFS and MA plans, we may have overstated the difference in payments received for FFS and MA admissions.

The Solution

The correct way to account for IME payments would be to obtain from CMS the IME payments made on behalf of MA beneficiaries and then match them, at the admission level, to the payments made by the beneficiaries' MA plans and the beneficiaries themselves. The sum of payments from these three sources should then be compared to the sum of payments from FFS (including the embedded IME) and FFS beneficiaries themselves for the common (standardized) basket of admissions.

Unfortunately, we were not able to make this calculation. However, we were able to stratify hospitals by whether they were “teaching” hospitals or not, and then calculate the MA/FFS payment gap for a fixed basket of admissions for non-teaching hospitals only. Because non-teaching hospitals have no (or low) IME payments, the MA/FFS gap in payments we calculate for them would provide an (approximately) unbiased estimate of the difference that we might expect (at least for non-teaching hospitals) if we were able to account for IME payments for MA admissions.

We found that in 2012, the MA/FFS price gap for non-teaching hospitals-analogous to the 8 percent gap we calculated for all hospitals-was 4.7 percent. The bottom line: the exclusion of IME payments from MA claims account for part, but not most, of the MA/FFS difference we originally calculated.

Looking Forward

More generally, our MedPAC colleagues are raising a crucial issue: What does it mean to compare claims payments made from two different systems when the systems may have different side payments not captured at the claims level? In particular, CMS is now giving bonuses to (and imposing penalties on) hospitals with certain FFS claims patterns, over and above (or under and below) regular FFS. According to recent work, these payments are on track to amount to almost 6 percent of all inpatient hospital FFS payments in 2016. By the same token, MA plans may also be exploring the use of bonuses and penalties to provide stronger performance incentives. Because the MA/FFS gap is of approximately the same magnitude as these side payments, accounting correctly for them needs to be a key part of future analyses.

Authors' Note

The authors thank our colleagues at MedPAC and Amanda Frost of the Health Care Cost Institute for helpful comments.

Laurence Baker and Daniel Kessler have received fees for speaking, consulting, or both from hospitals, insurers, or both.