Highlights from the Eleventh CEO/Innovators Roundtable
David G. Anderson, PHD, Barbara Anderman, MA, Larry Vernaglia, JD, MPH
Last June, we held our 11th CEO/Innovators Roundtable in Boston. This two-day gathering brought together some of the most dynamic minds in healthcare to share ideas, examine new opportunities, and debate solutions for improving quality of care and containing costs. As in the past, the group consisted of leaders of health systems, health plans, and medical groups, as well as legal experts, healthcare entrepreneurs, and investors. Importantly, these healthcare leaders are not only in the business of healthcare, but they are users as well, and many brought their own personal experiences to shed light on the issues.
This year’s roundtable was organized around the theme of implementing population health management (“PHM”). While the concept of PHM is not new, many more building blocks are in place today for health systems to manage performance, quality, and financial risk for defined populations. At this point, however, there is much more “sizzle” about PHM than “steak.” Unequivocal success experiences are rare, as CMS’ experience with its first two generations of Accountable Care Organizations (“ACOs”) illustrate. Success will require further development of analytical tools and clinical management approaches. It will also require more collaboration between provider and payer organizations to manage disease and health effectively and efficiently. Finally, it will require embracing patients as consumers – not just patients.
The two days were divided into five sessions, with each session led by a group of panelists. These panelists not only shared their own stories, but guided the conversation with an active and well-informed audience. The five sessions included:
Disruptive Approaches to Population Heath Management
Total Cost of Care (“TCOC”) Contracting
Measuring and Delivering Value
Developing and Managing Care Bundles
Population Health Management Infrastructure
Each panel dealt with a different set of critical issues that health systems, health plans, suppliers, and investors need to understand as our health sector evolves from pay-for-volume to pay-for-value. The first panel described the “irresistible forces” disrupting our current health sector (consumers, regulators, technology, and, on occasion, health plans, employers, and physicians) and the “immovable objects” they are aimed at – health systems and the markets and incentive systems that have historically sustained them. The TCOC contracting panel discussed the total cost of care gain-sharing arrangements used by CMS’ Shared Savings Program, many private insurers, and some large employers, concluding that it may not be possible for health systems to move incrementally into population risk – that health systems may need to transform themselves fundamentally in order to “cross the crevasse.” The panel on “Measuring and Delivering Value” examined the concept of value from the perspective of patients, payers, and providers and concluded that our obsession with measuring value is probably getting in the way of improving it. The panel on Bundled Payments prompted a spirited debate about whether these programs, developed by CMS and employer organizations like Pacific Business Group on Health, are likely to make sustainable new markets in bundles. And, finally, the panel on Population Health Management Infrastructure discussed a number of issues that are core to infrastructure development – e.g., whether EHRs can be relied on as the “single source of truth” or whether a pluralistic network of connected systems with distributed “sources of truth” is required to manage population health.
While much of the Roundtable discussion focused on the importance and centrality of integrated technology, we invited a dinner speaker to remind Roundtable participants of the risks inherent in our growing dependence on technology. Dr. Daniel Nigrin, Senior VP & Chief Information Officer at Boston Children’s Hospital, shared the experience of his hospital under a cyberattack from the internet group “Anonymous,” following publicity around the care of a patient with a controversial diagnosis and treatment plan. Nigrin’s remarks helped ground the Roundtable participants in the reality that crucial information systems have vulnerabilities, and offered a case study in how one well-regarded provider responded to the challenge.
DISRUPTIVE APPROACHES TO POPULATION HEATH MANAGEMENT
Amar Kendale, MS: SVP, Product Management, Livongo
Richard Popiel, MD: EVP & Corporate CMO, Cambia Health
Charles Saunders, MD: CEO, IntegraConnect
Tim Sayed, MD, MBA: VP, Physician Engagement, Interpreta, Inc.
Joseph Smith, MD, PhD: CEO, Reflexion Health
Our first panel depicted the current world of population health management as a fundamental tension between irresistible forces and immovable objects. Reliable “irresistible forces” of disruption in today’s environment are:
Regulators, especially CMS
Some players can be either disruptors (part of the solution) or obstacles (part of the problem), including:
The main “immovable objects” are our current health systems and provider communities, along with the markets and incentive systems that sustain their traditional business models.
Consumers as Disruptors
Amar Kendale, SVP Product Management of Livongo Health, sees consumers as the only group with the right incentives to disrupt today’s healthcare markets. Livongo helps consumers manage their own diabetes by providing a simple, easy-to-use blood glucose monitor and wrapping services around it that fit within consumers’ daily lives. He described Livongo’s approach as “putting doctors’ work in the hands of consumers.” In Kendale’s opinion, people don’t want to be “engaged” with their providers; they want to be empowered to reduce hassles related to their chronic conditions. To empower consumers, Livongo’s tools “close the loop with consumers every time they test their blood glucose.”
Joe Smith, CEO of Reflexion Health, described his company in similar terms. Reflexion’s goal is to bring physical medicine into the home, and its Vera System is a high-tech home-based rehabilitation kit for joint-replacement patients. The solution integrates motion-tracking technology, a “smart, responsive avatar,” and real-time feedback aimed at saving post-op hip replacement patients costly and time-consuming trips to physicians and physical therapists. While consumers love it, according to Smith, physical therapy providers are concerned that this home-based approach could reduce demand for their services.
Diverse types of consumers respond differently to various approaches, and, as a result, there is no one-size-fits all method for consumer engagement. Chuck Saunders of IntegraConnect said that when he was leading Healthagen, Aetna’s service provider for ACOs, the company identified eight different subgroups of consumers, each of which wanted to interact with their healthcare providers in different ways. Millennials, for example, may look to consumer healthcare apps to make managing their health fun. For these consumers, introducing “gamification” into mobile apps may be an effective approach. One participant commented that effective gamification requires giving patients frequent rewards. He cited Bing Gordon, EVP of Activision, the electronic game manufacturer, as saying their games were designed to provide players with a reward every 30 seconds.
While some millennials may be attracted to gamified apps for exercise or minor self-care interventions, panel participants felt that most people view healthcare as a chore, not a hobby. Consumer-driven healthcare needs to be adequate and convenient, not necessarily “delightful” or “fun.” Dr. Gene Lindsey, retired CEO of Atrius Health and a Roundtable participant, suggested that consumers want: (1) better access to information; and (2) workability – the ability to do things easily. One of Gene’s suggestions: “How about designing an explanation of benefits (“EOB”) that people can read and use to keep track of their families’ healthcare deductibles, like a bank account?”
In designing healthcare applications and programs for consumers, it is easy to overreach and assume they are willing to do more to manage their own health than is actually the case. Tim Sayed of Interpreta attributed the failure of Google Health’s personal health information service to unrealistic expectations about consumers’ willingness to organize their own health information. When it comes to managing their healthcare, Sayed said, “Patients do not want to do heavy lifting.”
Regulators as Disruptors
All the panelists agreed that CMS, the principal agent of health reform, has disrupted the industry in major ways. CMS’s new approach to value-based physician reimbursement – MIPS / MACRA – is the latest in a long line of changes aimed at disrupting traditional fee-for-service payment systems (which CMS, of course, enabled for years). Gene Lindsey characterized MIPS / MACRA as “an exercise in behavioral economics.” Chuck Saunders described MIPS as “a zero-sum game that will create both winners and losers, with losers likely outnumbering winners.” Over the next few years, MIPS / MACRA payment incentives and penalties are expected to cause significant dislocation in physician practices, spurring integration, new relationships, and a flurry of transactions.
Saunders’ company, IntegraConnect, is aimed at another one of CMS’s disruptive programs – bundled payments for chronic conditions. IntegraConnect is a “clinical operating system” for patients with serious chronic conditions who need specialty care. Its “end-to-end, cloud-based” solution enables specialists to manage mid-and late-stage chronic conditions cost-effectively, while maintaining high quality. As he sees the value-based payment world:
Patient-centered medical home (“PCMH”) payment models are best for primary care and treatment of early stage chronic conditions
Risk-sharing arrangements with provider networks like ACOs are best for primary care and poly-chronic conditions
Episode-based bundled payments are best for paying specialists who manage well-defined procedures
Condition sub-capitation is the best way to manage mid- and late-stage chronic conditions
Several years ago, CMS recognized the benefits of sub-capitation for end-stage renal disease (“ESRD”) and created a capitated payment program for patients on dialysis. They have now extended this approach into cancer with their Oncology Care Model (“OCM”) pilot, and IntegraConnect is partnering with oncologists across the country to treat patients under this program. Next in line, according to Chuck, will be other high-cost, high-prevalence specialty episodes, including urologic conditions like prostate cancer, GI diseases like Crohn’s and Hepatitis C, and Rheumatoid Arthritis. He expects CMS to move into these areas in the near future.
Technology as a Disruptor
There is no doubt that information technology companies are becoming major disruptors of traditional healthcare. Over the past few years, Health 2.0 and other “incubator” companies have launched innumerable consumer-oriented healthcare and population health management applications and devices aimed at many conditions.1 Tim Sayed, VP of Physician Engagement for Interpreta (a “command and control center for precision medicine”) said that a fundamental question limiting the growth of PHM technologies is how centralized we want PHM information to be. “Do we want a single source of truth, or are we more comfortable with decentralized sources of information that we access when needed?” Sayed said there are no technological constraints to collecting all relevant PHM information in one place. Constraints to centralizing information are driven by differences in business processes, intellectual property considerations, security issues, and, of course, HIPAA.
Health Plans as Disruptors
For the most part, health plans have not been major health sector disruptors for two main reasons:
Most have been doing well financially in the current reimbursement environment; and
Their claims payment systems are based on old architectures and have grown so complex that they are difficult to adapt to new payment methodologies.
Some health plans, however, are making efforts to help providers take risk and manage population health. Dr. Richard Popiel, EVP and CMO of Cambia Health Solutions in Portland, reflected on his experience with Horizon Blue Cross Blue Shield of New Jersey, which had a strong PCMH program that paid PCPs to manage care and also experimented with a variety of bundled payment systems for different conditions.
Popiel has carried this innovative spirit to Cambia, which he described as a “consumer-driven health solutions company” whose “goal is to create a collision between the consumer and evidence-based care.” Cambia takes “a 360° view of members,” and tries to answer five fundamental questions they have about their care:
Do you know me?
Do you care about me?
Do you have the skills to help me?
Will you meet my needs?
Are you giving me reasonable value?
To get affirmative answers to all five questions, Cambia treats members as decision-making consumers, giving them choices and trying to create customer buying experiences similar to Amazon’s.
“We are a consumer-driven health solutions company that takes a 360° view of our members. Our goal is to create a collision between the consumer and evidence-based care.” - Richard Popiel, MD – EVP & Corporate CMO, Cambia Health
Employers as Disruptors
Several participants asked whether employers are likely to disrupt traditional healthcare. Clearly, a number of high-profile employers like Boeing, Walmart, and Safeway are disrupting traditional healthcare through “direct-to-employer” contracting (Boeing), retail care development (Walmart), and reference pricing (Safeway). However, most panelists weren’t convinced that employers will sustain this pressure over time. As one panelist put it, “Employers are a mixed bag. Large employers cater to their employees, and they have lots of constraints.” Another panelist observed that working with employers requires working with benefits consultants, whom he thought were rarely proponents of change.
Physicians as Disruptors
During the “physician practice management” era in the 1990s, physician groups and IPAs played a major role in disrupting traditional healthcare delivery. While IntegraConnect’s business model depends on independent specialty groups who want to take risk, most panelists and participants were not sanguine about organized physicians creating major pressures for implementing PHM. One difference is that many more physicians are employed by health systems today than in the 1990s. Also, these health systems are more concentrated, more diversified, and less vulnerable to “disintermediation” than they were in the 1990s. Finally, physicians have long memories, and many are leery of taking financial risk, especially when for many specialists, fee-for-service medicine can still be lucrative.
Perhaps for these reasons, our panelists did not identify physician entrepreneurship as a major disruptive force this time around. In fact, several saw “deconstructing” healthcare and relying less on doctors as a promising way to manage populations at lower cost. One panelist said that case managers generally know their patients much better than their doctors. Another panelist agreed, saying that care coordinators and case managers can reduce cost by 10% overall, and 30% for people with chronic conditions. Surprisingly, he said, RNs are only needed for 10% of these interactions.
In most industries, competition is the ultimate disruptor. In today’s healthcare sector, however, competition between health systems over who manages population health more effectively and efficiently is still rare. A handful of HMOs like Kaiser Permanente and provider-sponsored health plans like Providence Health Plan, Henry Ford’s Health Alliance Plan, and Sentara Health Plan, have developed some PHM capabilities, but most compete with traditional health plans and care delivery systems rather than with other PHM companies. Building more PHM companies was, in fact, one of CMS’ main reasons for building provider-driven ACOs, a movement that is still in its early stages.
Health Systems: The Immovable Objects
Despite all these powerful disruptors, our large health systems seem unperturbed. Joe Smith opened the panel by quoting Jonathan Bush, the CEO of athenahealth, who once said “Healthcare is one of the few industries where you can say the same thing for ten years in a row and still be viewed as a visionary.” Smith’s point is that scores of participants have been sharing groundbreaking ideas at these types of discussions for years, but adoption – wholesale change – has been difficult to achieve.
“Healthcare is one of the few industries where you can say the same thing for ten years in a row and still be viewed as a visionary.” - Joseph Smith, MD, PhD (attributed to Jonathan Bush, CEO – athenahealth)
Gene Lindsey cited research by Bob Wachter of UCSF that it takes 10 years for digitized solutions to become embedded in institutional workflows. One Roundtable participant gave an example: the slow progress of adopting glucose monitoring kits for diabetic patients. “Why isn’t it going faster? It’s because these ‘thousand points of light’ require health systems like ours to tie them together.” Culture also acts as an anchor, because culture change is so time consuming. As Lindsey pointed out, “We have a culture of obesity, which has enormous implications for our health. How do we change this?” Another participant described ER visits as “hacks” that reflect failures of population management. Emergency rooms aren’t going away, however, and changing patient behavior will require development of suitable substitutes, as well as changes in incentives and our ER-centric culture.
Intrapreneurship as a Path Forward?2
Innovative leaders in some health systems have built new population health management skills that can be very useful when environmental incentives shift. Intermountain Health Care, Providence Health & Services, Henry Ford Health System, Geisinger, Sentara Healthcare, Texas Children’s Hospital, Memorial Hermann, Montefiore Health System, and many others have capitalized provider-sponsored health plans or other risk-bearing entities within their organizations. As a result, these systems have developed core competencies in population health management that others lack.
Building provider-sponsored health plans is not the only way to cultivate “intrapreneurship” within health systems. Many health systems have developed ACOs or clinically integrated networks (“CINs”) with significant financial incentives to manage population health. Leaders of these systems are sometimes surprised when their incentivized doctors steer patients away from hospital-based outpatient services to less expensive, independent doctor-owned facilities. Rather than browbeat these physicians, health systems should see them as “canaries in the coal mine,” and move to make their prices more competitive.
Relying on intrapreneurship, however, may not produce change that is fast enough to achieve “escape velocity.” Resistance to change is entrenched. As one participant put it:
“Why would a provider adopt a new treatment model if he or she will experience a dramatic drop in patient visits or fees for services performed? Why would a hospital system that invests heavily in its physicians, plant, staff, and equipment embrace a healthcare model that strives to keep consumers (and their dollars) outside the hospital?”
The discussion of cultural and economic resistance to change evolved into a discussion of “creative destruction.” Gene Lindsey summed up many participants’ feelings by saying, “You can’t get to the new world incrementally. You have to pull the rug out.”
TOTAL COST OF CARE (“TCOC”) CONTRACTING
Fredric Leary, MD: CACO, Palo Alto Medical Foundation
Scott Powder, MBA: SVP & CSO, Advocate Health Care
Katherine Schneider, MD: President & CEO, Delaware Valley Accountable Care Organization
Sanjay Shetty, MD, MBA: EVP, Corporate and Business Development, Steward Health Care
This panel focused on two main questions:
Is risk migrating from insurers to providers, and, if so, how fast?
Are “total cost of care” shared risk contracts like Medicare’s Shared Savings Program (“MSSP”) stepping
stones to greater risk-sharing?
Four panelists with markedly different experiences with risk contracting discussed these questions:
Scott Powder, SVP & Chief Strategy Officer of Advocate Health Care. At $5.5 billion in revenue, Advocate is the largest health system in Chicago, with over 1,500 employed physicians and a “narrow network ACO.” Altogether, Advocate has 850,000 lives under some type of risk-based arrangement. About 350,000 of these are in capitated or “close-to-capitation” contracts, and this number is expected to increase to grow. The other 500,000 lives are in upside-only performance bonus or shared risk total cost of care contracts.
Dr. Fredric (“Rick”) Leary, Chief Accountable Care Officer for Palo Alto Medical Foundation, a subsidiary of Sutter Health. PAMF employs about 1,300 providers, approximately half of whom are primary care practitioners and half are specialists. These providers care for approximately 1 million patients, generating about $2 billion in revenue. PAMF has five commercial payers and has entered into value-based reimbursement arrangements with three of them. Rick pointed out that in the 1990s, PAMF was a major player in managed care contracts, with almost ¾ of its revenue coming from capitated contracts, especially Medicare risk. Since then, however, PAMF has throttled back on risk-taking, and today only about ⅓ of its business has significant risk.
Dr. Katherine Schneider, President & CEO of Delaware Valley Accountable Care Organization (“DVACO”). DVACO is jointly owned by five Philadelphia-area health systems. It began sharing risk in the Medicare Shared Savings Program (“MSSP”) in 2014 and more than doubled in size to over 100,000 Medicare beneficiaries in 2015. DVACO earned substantial bonuses in 2014, but not in 2015, because new practices and patients brought the TCOC benchmark down by $1000. (Interestingly, total cost of care continued its slow downward trajectory, generating savings for Medicare.) In addition to MSSP lives, DVACO contracts with three private payers for another 100,000 covered lives. All of DVACO’s contracts are upside gainsharing only contracts. Most of its participating physicians are PCPs, about half employed by DVACO’s owners, and half independent.
Dr. Sanjay Shetty, EVP, Corporate and Business Development for Steward Health Care. Steward is the largest fully integrated healthcare services organization and community hospital network in New England, with 10 hospital campuses and approximately 3,000 employed and affiliated physicians participating in its Accountable Care Organization. According to Shetty, Steward is deeply committed to value-based contracting and has about 350,000 lives under Medicare or commercial risk contracts. Steward was an original Pioneer ACO and is now “graduating” to Medicare’s NextGen ACO program. Steward’s largest risk-sharing commercial contract is Blue Cross Blue Shield of Massachusetts’ “Alternative Quality Contract,” which shares significant upside and downside risk with providers within defined risk corridors.
How Much Risk is Shifting?
On the first question – how much risk is actually migrating from insurers to providers – the general viewpoint of the panelists was, “Not much yet.”3 Scott Powder described a meeting he attended recently with seven other “peer” health systems, where he said only two systems – Advocate and one other – currently have substantial numbers of privately insured capitated lives. Panelists had several explanations for this slow uptake. One is simply long memories of the risk-taking overreaches of the 1990s. As one participant put it, “The HMO scares of the 1990s are still discouraging capitation. We need a new, improved capitation 2.0.”
Another explanation for the slow migration to risk is the sheer complexity of managing the health of different populations with markedly different cost structures and health behaviors, as Dr. Al Daugird, Chief Value Officer of UNC Health Care and a Roundtable participant, pointed out.
Another reason for caution is the inherent challenge of managing MSSP and PPO risk arrangements with limited utilization controls, patient freedom of choice, and, in many cases, “black box” attribution methodologies. Taking risk with HMO contracts that have primary care gatekeepers and positive utilization controls is inherently less risky than taking risk in MSSP and PPO arrangements. While Advocate and Steward both operate in markets with substantial commercial HMO populations, most markets are dominated by PPOs, and providers are understandably slow to adopt higher levels of risk.
Another reason for the slow migration to risk is the ambivalent attitude many insurers have about sharing risk with providers. Drs. Shetty and Schneider both said their insurers were willing to allow providers to take more risk. Steward is more enthusiastic about taking risk than DVACO, perhaps because they have more direct control of their physicians. As Shetty said, “We’ve seen real change in our commercial contracts, and we have been able to craft value-based contracts that suit providers’ needs as we gain more expertise and experience over time.” Another participant in the Boston market with considerable risk-taking experience agreed, saying, “We love capitation and taking full risk. We do the right things for our patients, and it’s a better place to work for the docs.”
On the other hand, Leary and Powder questioned whether the health plans they dealt with would ever negotiate generous enough contracts to cover the cost of managing the risk of the populations their systems take risk for. Leary said that Sutter has decided to start its own health plan and compete with other health plans in California in order to win with risk. As he said, “We’ve gotten financial returns from shared savings programs, but there are lots of costs there, too, and the revenues don’t often keep up with the costs.”
The reality is that while the ACA’s regulation of the medical loss ratio (“MLR”) “floor” has shifted incentives a bit, payers and providers are still fighting over what share of the revenue each will get. We had a real-time illustration of this dynamic when Dr. Leary said PAMF’s enthusiasm for risk depends on the capitation rates they can negotiate with the plans. “If you give me an MLR of 86%, I’ll take it on.” At which point, a health plan executive in the room responded, “In that case, I’ll give you 84%.”
Most panelists were united in their view that “payer-provider partnerships” aren’t very “partner-ly.” Shetty said that for the most part, payers don’t help providers manage risk: “In many cases they’re dealing with antiquated systems that can’t easily produce the claims data and reports required to manage different populations. As a result, we don’t get timely reports, and we have had to make large investments to create and operate our own infrastructure to manage risk.” Shetty is not necessarily complaining: “We believe that as providers, we are in the best position to translate these investments into better care for our patients.” But the fact remains that health plans aren’t helping much. Schneider reminded the group that DVACO’s members are not only “partners” of the health plans through DVACO, but also suppliers, and “the fear of retribution if we compete with them is real.”
Of course, competing with health plans is only possible if health systems spend enough on marketing to attract members and reserve enough risk-based capital to meet insurance regulations. Dr. Gene Lindsey, who is the Board of The Guthrie Clinic in central Pennsylvania, said that Guthrie was considering whether or not to start their own Medicare Advantage plan. While many health systems have done this, the requirements are considerable, and the risk is significant – e.g., CHI’s recent decision to sell QualChoice Health and exit the insurance business.4
TCOC Contracts: Transition Vehicles?
The ACA’s conception of Total Cost of Care gain-sharing contracts, as reflected in the MSSP, is that by taking on controllable amounts of risk, ACOs will learn how to manage population health. Our panelists, however, were not convinced that risk-sharing TCOC contracts were very effective in promoting organizational learning and building population health management skills. They expressed several concerns:
Sustaining gains from TCOC gain-sharing contracts is very difficult, unless you start with substantial excess costs. Even then, the string runs out quickly. (Schneider)
Gains from risk-sharing are likely to be offset by loss of margin on reduced health system revenue.
Health plans’ antiquated systems are often incapable of providing useful information to help ACOs manage
total cost of care. (Shetty)
The interaction between FFS and risk-sharing contracts creates internal tensions. As one participant said,
“Fee for service has twisted primary care into a revenue-generator for the system. We need Medicare to pay
us a primary care bundle for our PCPs to manage population health.”
Physicians who need to live in a “hybrid world” experience tensions As one participant said, “Physicians are
good at checking the boxes to get the dollars, but if the direct incentives go away, they’re not very interested
Gains from shared-savings contracts don’t usually cover the cost of managing population health. (Leary)
Powder expressed a clear point of view that capitation or “near-capitation” contracting strategies are the most effective ways to change physician and health system behavior. Risk-sharing that falls short of this “threshold” may not provide enough incentive to motivate the significant changes required. Advocate’s TCOC contracts, he said, have had relatively little impact on behavior. He does not believe that CMS’s new MACRA program will be successful in moving health systems to alternative payment methodologies. Shetty agreed, saying, “MACRA won’t itself promote the adoption of APMs. As it stands, the rules are the rules. You should assume you’re doing MIPS.” Marci Sindell, Chief Strategy Officer of Atrius Health and a Roundtable participant, sees increasing regulation like CMS’ MACRA / MIPS program as inevitable: “40% of our State budget is healthcare, and our [Massachusetts] Health Policy Commission has already added lots of regulations.” Micro-management of health systems by government purchasers and agencies may be inevitable, but it seems unlikely to create the conditions for health systems to become effective population health managers.
MEASURING AND DELIVERING VALUE
Terry Carroll, PhD, MS: CDO, SPRING Network
Ray Herschman, MS: Former President, xG Health Solutions
Charles Kennedy, MD: CPHO, Healthagen
Michael Palmer: Former Chief Innovation and Digital Officer, Aetna
This panel organized the topic of measuring and delivering value around four different perspectives: (1) the patient; (2) the consumer (who may or may not be a patient); (3) the purchaser (health plan or employer); and (4) the provider.
Value for the Patient
Ray Herschman kicked off the panel by taking the patient perspective, telling the personal story of a family member who is dealing with eight known serious diagnoses requiring the involvement of eleven different specialists. From the patient’s and family’s perspective, Herschman said, value isn’t about cost or savings. It’s simply about quality of care, which his family experienced as confidence or lack of confidence in their physician providers. Really sick patients are scared, and for them value is defined in large measure by their communications with their doctors. Herschman then described the daunting task of coordinating care, especially getting all the specialists involved in his relative’s case communicating with each other. As he said, “We had to force a meeting to get the doctors to talk to each other. We had to call a ‘Code Red.’ Afterward, they all said, ‘We’ve never had a huddle like this before.’” This patient-driven “multi-disciplinary clinic” not only reassured the patient, but actually surfaced several new ideas for treatment that weren’t being considered before. However, it shouldn’t take an extraordinary effort by a highly sophisticated family member to create this type of value for patients.
Value for the Consumer (Not Just the Patient)
Today’s healthcare buyers include a wide spectrum of small, medium-sized, and large businesses, as well as individuals and families who purchase individual health insurance, many through the health insurance exchanges. What these consumers look for, according to Michael Palmer of Aetna, is: (1) convenience; (2) results – they want their problems solved – and (3) affordability. On the insurance exchanges, affordability is the bedrock. As Palmer said, “You have to be on the first page of www.healthcare.gov [i.e., among the least expensive plans] to be considered.”
As an example of a more convenient, consumer-friendly health plan, Palmer described a transitional plan Aetna recently launched that includes an online appointment scheduling function, simple payment instructions (e.g., answering the question, “Should I pay this bill?”), and a smartphone app that connects a member to a live person at their call center.
Delivering value by producing better results will require much more powerful interventions than most health plans have in their arsenals today. As Palmer said, traditional health plan-sponsored weight loss programs have minimal impact on members: “2% answer our phone calls, and 3% of these do anything at all to change behavior.” Palmer went on to describe a new, more powerful “metabolic syndrome” risk reduction pilot program that Aetna undertook with Newtopia, a Canadian company. Newtopia provides a health engagement platform that inspires individuals to develop healthy habits to meet their own unique lifestyles.
Initially, 500 of 1,900 members in the pilot joined Newtopia’s voluntary program. Participants were asked to measure their ability to do a series of push-ups, wall-sitting exercises and other tasks. Based on their ability to perform these tasks, a personal program was developed for each individual that included logging their exercise and food consumption and participating in weekly Skype video calls with coaches. 60% of subscribers lost 6-7% of their weight in the first year. Participants reported feeling great, and they were able to maintain their weight loss over two years. By keeping their weight down, Aetna estimated that participants saved an average of $1,500 a year in medical costs. For employees on high-deductible HSA insurance plans, this $1,500 was immediate value to the consumers.
Value for the Payer
Charles Kennedy of Healthagen began his talk by recognizing the inherent conflict between the “wholesale” business of designing and marketing a commercial health plan, and the one-to-one relationships between patients and their physicians that creates value for members. Providing value-based care means leveraging the power of averages, while understanding that there is no such thing as an average patient or consumer. He identified a few of the ways averages are used in care management, including managing length of stay and readmission rates, defining clinical guidelines, designing benefits plans, and even defining quality scores (with high and low scores tossed out). These averages hide unique patients with unique circumstances, and simply applying the average solution to all patients will not create success.
To illustrate this idea, Kennedy told a story of the introduction of jet fighters by the United States Air Force in the 1950s. Pilots complained that despite the amazing technology in the planes, they were unable to make them perform to the fullest because the cockpits were too small and limited maneuverability. To remedy the problem, the Air Force set out to measure each pilot including waist, height, and even neck circumference. The data was sent to the manufacturers who built new cockpits based on the average measurements. The result was a cockpit that was no better than the original cockpit. The Air Force then re-measured its pilots to determine how many of them fit the average measurements and, according to Kennedy, not a single pilot was “average.”
Healthcare systems, Kennedy argued, are at risk of making the same mistake—delivering care based on the average. “By optimizing for the average, we will sub-optimize for every individual.”
Kennedy also described several major challenges that health plans are facing as they modify their businesses for value-based care, including:
The need to update legacy data systems that, while functional, were built for fee-for-service medicine and not for population health.
Developing new data sources to help manage care (“We haven’t done enough to use data to manage care.”).
Overcoming cultural diversity within the commercial payer segment
Convincing members to see their health plans as partners in managing their health (rather than adversaries to contend with).
“We’ve seen examples where health plan personnel have had to be moved into different roles because they couldn’t switch from a volume to a value focus,” Kennedy said. “Payers are starting to shift, but it won’t be a rapid or easy process.”
Managing Population Health = Managing Individuals
Charles comments about individuals and population were seconded by a number of participants. One participant restated this idea by saying that what’s best for the population has to start with what’s best for individuals. Terry Carroll of SPRING Network said that insurance companies should focus less on spreading risk over individuals and more on managing individual conditions. “We’re too focused on averages. We need to unlock the value that’s tied up in individual genomes.”
Another participant reminded the group of Clay Christensen’s principle that 80% of care is very standard, but 20% is much more customized. Sindell of Atrius Health picked up on this theme, saying we need to transition caring for individuals from an art to a science. One way to increase process consistency and reliability, she said, is to remove routine work from doctors and let them spend more time managing unique cases.
Value for the Provider
Providers value useful information about their patients, which is how they distinguish individuals from averages, according to Terry Carroll of SPRING Network. He highlighted a relatively new problem providers are having with current electronic health records: The various indicators and “alerts” EHRs generate are beginning to overwhelm them and interfere with their ability to pay attention to what really matters. “Should we be asking different questions?” he asked. He then described how the emerging remote monitoring industry was designing their systems – measuring only what is needed to take action – and suggested this as an approach EHR vendors should consider.
Carroll’s comment spurred a general discussion of the proliferation of metrics from CMS and others that attempt to measure value. Herschman said that CMS’ metrics-intense approach to population health management “is a rational way to drive affordability.” One health plan executive participant observed that if half of all payments are going to be value-based, “Providers will need all the specialty-specific metrics they can collect to hit their numbers.” A number of participants, however, felt that the use of value metrics is getting out of hand. Powder of Advocate said that while we have a lot of metrics, our understanding of what value really means to our customers is woefully inadequate. Kennedy emphasized the importance of having the best knowledge relevant to the patient’s specific clinical condition, not just general patient experience metrics. Daugird of UNC reminded the group that severity adjustment is still in its infancy, and that performance metrics without valid severity adjustment can produce flawed – even harmful – results.
Gene Lindsey captured the group’s gestalt by saying, “We tie ourselves in knots trying to measure everything we can,” and he argued we need to back off on micro-measurement and become much more precise and limited. “We need to design metrics for management, not for reimbursement.” Another participant suggested that accreditation surveys that aggregate multiple metrics may be much more useful than lists of metrics. Citing the ANCC’s Nursing Magnet Recognition accreditation program, she suggested that rather than worry about Leapfrog and HEDIS scores, the business community should pressure hospitals to get magnet-certified. “Nurses drive patient experience in hospitals,” she said, “and patients really see a difference is a Magnet hospital.”
DEVELOPING AND MANAGING CARE BUNDLES
Bill Kramer, MBA: Executive Director, National Health Policy. Pacific Business Group on Health
Sunny Ramchandani, MD, MPH: Deputy Chief Medical Officer, Aetna and former CMO, Federal Employee Benefit Program
Dave Terry, MBA: CEO, Archway Health
Dave Terry, CEO of Archway Health, began this panel by providing the rationale for CMS’ effort to “make a market” in bundled payments:
Specialist care drives most healthcare costs
When people get sick, they go to specialists, and PCPs are largely out of the picture
Specialists have significant control over procedure costs
Bundled payments affect specialist compensation
Specialist compensation drives specialist behavior
He illustrated this by describing how an orthopedic surgeon can triple his income for a hip replacement procedure under bundled pricing:
Assume payment for the total joint replacement bundle is $30K, and total costs are $25K, including a payment of $1,200 to the surgeon
Assume that aggressive cost management could save 20% of costs ($5,000) – “very realistic,” according to Terry
If the surgeon gets to keep half the savings in addition to his fee (“typical”), he or she would be paid $1,200 + $2,500 = $3,700
Simple instructions to the patient can often make a big difference in cost and quality. For example, scheduling a return visit in two weeks vs. six weeks can allow the surgeon to review rehab results and modify protocols if necessary. Or, simply instructing the patient experiencing increased redness, swelling, or pain to call the practice rather than go to his or her PCP or an Emergency Room, can save the cost of a primary care, urgent care, or ER visit.
Bundled payments also create transparency around costs, allowing the surgeon to see all the cost elements of the procedure. Terry described a situation where one surgeon found out the ambulance company was being paid more than he was ($1,500 vs. $1,200) and put a system in place to replace non-emergent ambulance transports with nursing transport vans or other reasonable alternatives.
Bill Kramer then described Pacific Business Group on Health’s (“PBGH”) Employers’ Centers of Excellence Network (“ECEN”) program. PBGH represents 60 public and private organization that spend $40 billion annually on healthcare services for 10 million people. Large, self-insured employers, Kramer said, are frustrated with the slow progress health plans are making in controlling healthcare costs and improving quality, and, as a result, PBGH has contracted with a number of health systems around the country to provide specific procedures at competitive bundled payment prices. Several large PBGH members have signed on to the program (e.g., Walmart, Lowe’s).
Currently, PBGH has developed networks to deliver hip and knee replacements and spine surgery, and it is developing a network for bariatric surgery. Kramer stressed that PBGH is not looking for the cheapest provider. As he said, “We are looking to find elite partners on the cutting edge of surgical care that will collaborate with us.” The value to the patient includes zero deductibles and coinsurance, free travel, excellent surgical outcomes, and “white glove” customer service. The value to the employer is excellent surgical outcomes, avoidance of surgery where unnecessary, shorter time away from work, and employee retention. He cited a Lowe’s employee who told his supervisor about the excellent service he received for his hip replacement at Virginia Mason MC in Seattle, and concluded by saying, “I just love working for Lowe’s.”
The discussion focused on five major issues regarding bundled payments:
How well do bundles work?
How do bundled payments fit with an ACO strategy?
How ready are providers for bundled payments?
How can bundles engage consumers?
What is the likely trajectory of bundled payment systems?
Each of these questions is discussed below:
How Well Do Bundles Work?
One participant who used Remedy Partners to develop a CHF bundle said his system’s experience has been “incredibly powerful.” “It was a lot easier for the physicians to control care than we thought.” Dr. Howard Landa, CMIO of Alameda Health System and a Roundtable participant, said that simply sharing data about bundles has helped them control costs: “When the surgeons found out they were making less than the pathologists,” he said, “It was quite a revelation.” Some participants, however, advised taking a cautious approach. “Health system providers need to be more than a portfolio of bundles,” said one, ““Don’t let the perfect be the enemy of the exceptional.” Terry summarized the discussion by acknowledging that while bundles may be part of the solution, they are not the whole solution. For example, he said bundled payments don’t deal with over-utilization of bundles. He concluded this discussion by saying, “But there’s a ton off waste. Let’s start somewhere.”
How Do Bundles Fit with an ACO Strategy?
At their core, the reimbursement philosophies of bundled payments and ACO capitation are fundamentally different ideas about how the healthcare market should evolve. As one panel member said, “If you siphon off high-acuity procedures, what is left for the ACOs to manage?” Michael Porter of Harvard Business School, Elizabeth Teisberg, and Robert Kaplan are three of the most articulate proponents of bundled payments. In a recent pair of Harvard Business Review articles, Porter and Kaplan squared off against Brent James and Greg Poulsen of Intermountain Health Care, each of which argued for one or the other payment system.5
Terry took a different view – that bundles are not a stepping stone to capitation, but that bundles and ACOs will likely co-exist. CMS is encouraging both approaches in the marketplace to see how they interact and which gains more traction. Gene Lindsey, former CEO of Atrius Health said that, in theory, bundled payments fit with open access PPOs and Medicare Advantage plans that don’t or can’t require referral authorizations. Dr. Jonathan Niloff, former CMO of McKesson Connected Care & Analytics, said that the effectiveness of bundles depends on the maturity of the provider system. Mature capitated health systems, he said, could theoretically create their own ‘proxy bundles’ by bonusing providers out of the total cap using risk-adjusted episodes of care.
One important question is how ACOs are likely to view bundles – as opportunities or threats. ACOs may avoid working with outside bundled care providers if they have to pay for these bundles out of network. Sindell of Atrius Health said that in 2014, “bundlers” would have taken more than $1M away from its Pioneer ACO savings, which would have pushed this contract from a gain to a loss. Terry argued that even large ACOs won’t always be the best in their market at everything. He believes that focused specialist bundlers with meaningful upside and downside incentives can create greater value for the ACO as well as themselves. Lindsey, Sindell’s former CEO, said he thought bundles should be considered part of the “value stream” of providing care and should be used as a “learning lab” by at-risk ACOs.
Are Providers Ready for Bundled Payments?
One panelist questioned whether more than 30-40% of providers are ready to manage bundled payments. He also wondered whether the financial upside in bundles is enough to attract physicians to make take them on, especially given the need to manage post-acute services, which represent a major risk factor in the cost of care.
Terry replied that while some specialty groups don’t want to participate in bundles, many do. The former group was represented by one participant who asked, “Why sign a fixed price contract when I can get a time and materials contract? Bundles represent an administrative hassle and create conflict between specialists. Who needs them?” On the other hand, Terry explained, the price difference for a virgin hip replacement at two demonstrably high quality providers could easily be 60%, creating a substantial opportunity for primary care groups or health systems to gain from shifting their referrals. Furthermore, according to Terry, specialty groups often know better and cheaper ways of doing things. For example, a large oncology group in Texas, found that 100% of their small-cell lung cancer patients were admitted to hospitals because they were in pain. Since alleviating pain does not require hospitalization, they changed their protocols to treat pain more effectively at home, significantly reducing hospitalizations.
Terry said that they have learned to divide doctors in medical groups and ACOs into three groups:
A group – Pioneering doctors who want to push the envelope on quality and cost, and who are excited and committed to bundling
B group – Doctors who aren’t necessarily enthusiastic, but who will respond to changes in incentives
C group – Doctors who won’t change, regardless of the incentives
One key question is whether CMS will ever allow provider incentives to reach the threshold necessary to motivate change. Cynthia Hare, President of Vizient Southeast, said her membership had 16 health systems that were participating in a collaborative around Lower Extremity Joint Replacement as a result of CMS’ mandatory Comprehensive Care for Joint Replacement (“CJR”) program. The initial range of target pricing received from Medicare for these systems was $18,000-$23,000 for a virgin joint replacement, including 3 days pre-op and 90 days post-op care. To hit these targets, the first thing the health systems did was to reduce and/or eliminate SNF care. Furthermore, she said, Medicare target pricing is expected to continue to decline and may well reach $12,000 in five years. At these pricing levels, providers may not get very excited about the opportunity to take risk. The idea that CMS is using bundled payments programs as just another way of extracting greater discounts from providers (similar to the MSSP) was widely shared.
How Can Bundles Engage Consumers?
Most of the companies participating in Pacific Business Group on Health’s Employers’ Centers of Excellence Network offer PBGH’s selected providers as a voluntary option to members, albeit with savings in their deductibles and coinsurance. Kramer said that using this approach to marry bundles with benefit design innovations creates a greater incentive for members to use this feature. Employee uptake appears to be less in companies with rich benefit packages, where the patient has less financial incentive to choose the ECEN program.
One Roundtable participant suggested that if you want to accelerate creative destruction, “you should stick the patient in the middle, make the cost of the bundled service the ‘reference price,’ and let the market move.” Kramer acknowledged that reference pricing – where the member bears all financial risk above the reference price – has proven quite effective for CalPERS and some other employers. However, not all employers are willing to embed reference pricing in their benefit plans. Picking up on this theme, one medical group executive said their group had found that midwives were better providers than OBs for normal deliveries, and asked whether a bundle couldn’t be created that would encourage moms to choose midwives as attendings for their normal deliveries. Kramer mentioned that PBGH has been experimenting with a reference pricing program to reduce unnecessary C-sections, but “it has been slow going.”
Trajectory of Bundled Payments
Terry emphasized that bundled payments programs “are still in the first inning,” with a lot more of the game left to play. One participant described the efforts of the Health Transformation Alliance (“HTA”), a health policy association of 26 large employers, similar to PBGH, aimed at cutting healthcare costs. (Dr. Glenn Steele, chairman of xG Health Solutions, is Vice Chairman of the Alliance.) HTA has four initiatives underway:
Data pooling and analytics
Developing provider networks (including direct contracting and possibly bundling)
Drug cost control (especially generics)
Consumer / member engagement
While HTA, like PBGH, is making progress, organizing groups of employers to accomplish these goals has been difficult, he said.
One obstacle to the adoption of bundled payments by private health insurers their administrative requirements. Processing bundles require significant changes in health insurers’ claims processing, actuarial, and repricing processes. As one panelist said, “Modifying systems for joint replacement bundles is fairly simple; modifying them for COPD is not.” As a result, relatively few insurers have jumped onboard the bundled payments train.
Another question raised by Terry relates to the future of CMS’ Center for Medicare & Medicaid Innovation (the “CMS Innovation Center” or “CMMI”) after the Fall elections. CMMI has been a major promoter of bundled payments, but the program has not evolved enough to begin producing a “flywheel effect.” If the wind goes out of CMMI’s sails, it could well deflate bundled payment efforts in both the public and private sector.
Assuming that CMS maintains its base of political support, CMMI plans to expand its bundled payments program significantly. Five years from now, CMS could be reimbursing many acute and chronic conditions using bundles. In general, Medicare bundled payment programs will probably continue to grow in markets where: (1) Average costs for procedures are relatively high (and variable); and (2) Bundled care providers are organized and motivated to bid for this business. On a national scale, the pace of development will depend on whether the large health insurers see bundles as a major source of potential savings and can get their arms around their infrastructure requirements. The future of “medical tourism” programs such as PBGH’s ECEN program, direct outreach by provider systems like the Cleveland Clinic, or direct employer contracting with bundled providers is unclear.
POPULATION HEALTH MANAGEMENT INFRASTRUCTURE
Jeff Cunningham: CTO & Chief Strategy Officer, Informatics Corporation of America
Kirk Elder: CTO, Wellcentive
Vik Kheterpal, MD: Principal, CareEvolution
Michael Reagin, MBA: CIO, Sentara Health System
Rick Ward, MD, MBA: Senior Advisor, BDC Advisors and Former VP for Clinical Programs & Medical Informatics, BCBS of Michigan
Rick Ward of BDC Advisors began this discussion by suggesting panel member’s address three questions about population health management (“PHM”) infrastructure that health systems need to understand:
Should health systems insist on a single tool at the point of care to standardize care processes across the network or try to establish sufficient interoperability with different point of care tools?
Should they focus primarily on self-service reporting applications and dashboard reports or invest in analysts with talent and training to formulate good questions, understand analytic methods, and tell actionable stories supported by data?
Should they outsource to meet their need for healthcare informatics and analytics expertise or build these core competencies internally?
On the first question, Ward said his point of view had shifted significantly:
“Twenty years ago, at Henry Ford, I thought we should build care management capabilities internally around a single system at the point of care. Since then, I’ve changed camps. We must knit together disparate systems. I’ll be dead by the time we standardize everything.”
Mike Reagin, CIO of Sentara Healthcare, agreed with Ward. He explained that Sentara has a clinically integrated network with over 3,000 MDs, with 54 different EHRs, and 20 different payer systems, and that “using a single tool is out of the question.” He went on to describe his view of the different functions PHM infrastructure needs to support:
Organizing physician workflow
Stratifying the population
Collecting and using population metrics
Understanding where gaps in care exist
Implementing care management and utilization management processes
No single system can possibly support all these functions.
Kirk Elder, CTO of WellCentive, a PHM system, agreed with Ward and Reagin, making the point that you need multiple “sources of truth” because data is in multiple places. “We can’t build everything in a monolithic mode. We need an agile platform, and we have to be flexible regarding source systems and mapping to value-based contracts.” Elder described WellCentive’s business model as using value-based contracts to define data needs. (This isn’t the only possible approach. Others systems utilize more general data models.) WellCentive collects data directly from source systems like EPIC, does the requisite analysis, and submits data to payers to get their clients paid. Data quality is always an issue, and the company spends considerable effort employing data quality tools in the service of the contracts.
Ward agreed with WellCentive’s emphasis on cleanliness and timeliness of data. He sees the need for clean, timely data as a common area of misunderstanding between CEOs and CIOs. CIOs assume that clean, timely data is critical to decision-making, but many CEOs don’t appreciate the investment required to achieve this.
Like WellCentive, Informatics Corporation of America (“ICA”) is at the center of interoperability issues with its CareAlign toolset, according to Jeff Cunningham, ICA’s CTO. “Getting the right information to the right people at the right time is really difficult,” said Cunningham. Achieving interoperability across an eco-system requires: (1) Technical connectivity of transaction-based systems; (2) Data standardization; and (3) An overall context for the data – Who needs what data, and how data links to problems (“a big area of frustration”). Cost is very important for small practices, and making connectivity affordable for them is challenging. “Federation” may be a reality, but today’s supplier systems like EPIC and Cerner have their own data models that aren’t aimed at connecting with others. While standardization of some data (patient IDs, demographics, clinical normalization, etc.) has been improving, understanding the many relationships and linkages between data elements is still difficult. It’s also important to understand what data are missing – critical data gaps – to identify new data sources that may be needed.
Vik Kheterpal, Principal of CareEvolution, stepped back from the interoperability issue to give his overall assessment of PHM infrastructure requirements. In his opinion, today’s tools and approaches are not up to the task of helping “pay-viders” (payer/providers such as ACOs) manage population health. “Big data,” he said, “is long on promise, but short on execution.” Typical approaches pit fiscal/financial requirements against clinical requirements, rather than harmonizing them. Risk scoring, for example, has its roots in pricing. Since “clinical care is delivered one patient at a time,” clinically relevant risk scores could drive front-line processing. However, that’s not the way they’re used. Instead, increased attention to Risk Adjustment Factor (“RAF”) scores has simply motivated providers “optimize” RAF scores by documenting more diagnoses. Similarly, benchmarking motivates fights over metrics, rather than stimulating performance improvement. Network management focuses on “reducing leakage” to enhance revenue, rather than eliminating handoffs and reducing coordination problems. Overall, Kheterpal believes that pay-viders need better connectivity – but it is difficult to get them to invest enough in “plumbing” to make a difference.
“Big data is long on promise, but short on execution.” - Vik Kheterpal, CEO, CareEvolution
One “Source of Truth” or Many?
All the members of this panel were strong advocates for establishing interoperability among multiple sources of truth, rather than relying on a single source of truth (e.g., EpicCare). Ward emphasized that interoperability needs to go beyond matching patient identifiers and sharing lab results to reflect plausible population health use cases. He is working with a “Super-CIN” comprised of ten different CINs – Clinically Integrated Networks – to implement a collection of standardized “population health transactions” that supports consistency in care processes across the network, rather than trying to get everyone to adopt the same EMR or implement a common disease registry for PHM across the entire network. A single source of truth in this structure would be impossible to build.
Rather than a single source of truth, Kheterpal said, you need traceability and uniformity – what he called “data liquidity.” In his view, “MUMPS and EPIC have become digital islands of automation with intentional wide moats around them.” You can build a federated database, Kheterpal said, if you have plumbing, infrastructure, and common data definitions. “The best databases are peer-based, not top-down. And the need for precision depends on the application. To drive clinical care for patients, for example, you can’t afford a 2-3% error rate, but you can if you are aggregating data to detect overall population trends or reporting provider performance.”
One participant summed up the situation by pointing out that while suppliers like EPIC and Cerner try to maintain their own proprietary interfaces, customers aren’t interested in overpaying these companies for access to their own data.
Self-Service or a Cadre of Analysts?
Ward addressed this question directly by saying that organizations that take on population health responsibilities need access to people training in analytic methodology disciplines such as epidemiology, biostatistics, actuary science, health economics and data science. These people also need to have some familiarity with healthcare and talent in formulating good questions and effectively telling actionable stories with data. People with these skills and talents are difficult to recruit, particularly because they are in great demand across many industries. “But they can be found,” he said, “At BCBSMI, we had 34 of these people strewn throughout the organization, including 12 PhD-level resources. They were extremely valuable.” Business intelligence or clinical decision support staff can sometimes be retrained as Data Scientists, but most have significant knowledge gaps that must be filled, requiring ongoing investment in coaching and professional development.
Cunningham agreed with Ward, explaining that ICA trains its staff to focus on translating data analytics into action, marrying processes and outcomes. Jonathan Niloff agreed with Ward and Cunningham: “Turning on a system gets you no ROI. You need a mix of systems and people. Very few clinicians will log in and look at canned reports. They get overwhelmed with data. The missing component is someone with domain expertise who can set informational priorities.” Kheterpal also agreed, emphasizing that you have to worry about the audience. “You can’t overwhelm clinicians with more and more raw data. You have to make it contextually relevant and useful.”
Despite this consensus, Ward pointed out that this is still the minority view. Many population health leaders and the technology vendors that sell to them still focus on the need for “dashboards” and “self-service reporting applications” and believe that developing trained and talented analytic human resources is optional or unaffordable or both.
Outsource or Develop Skills Internally?
Elder answered this question by acknowledging that companies can add value by creating subscriptions around a continuum of services. For example, an outsourcing company can offer clients access to analytic experts on a part time basis – “fractional data resources” – and help them develop a hiring plan to build internal analytic talent. By and large, however, panelists thought that PHM skills were too important for health systems to rely on outside companies to provide these resources over the long term. As Reagin of Sentara explained, “Our approach is to buy systems and integrate them internally.”
identified data in our data warehouse are cleaned, normalized, and standardized before they are made available for research. Real-time data isn’t necessary for the research we do at OptumLabs, and our data scientists are adept at working with EHR data that isn’t always perfect.”
Reagin of Sentara, on the other hand, emphasized the importance of dynamic data-sharing: “It takes 90 days to complete an episode of care. You have to keep the data in the pipe and use analytics to determine what happened, what is happening, and what should happen in the future. He also advocated building customer relationship management capabilities like Amazon: “New patients want to be touched.”
There was no particular consensus on this issue, probably because both data management strategies play important roles in managing population health. As data systems develop, however, and real-time data becomes more available to real-time analysis, the distance between these approaches will shrink. Cunningham described the architecture of Fast Healthcare Interoperability Resources (“FHIR”) as a major step forward in this development that presents “tremendous possibilities.” FHIR is a cloud-based Applications Programming Interface (“API”) of HL-7 that isn’t locked into a specific use case; rather it provides a model for consistently accessing and sharing data across clinical systems, similar to how iPhone apps are built on a common API framework to utilize mobile phone functions. Cunningham claimed that the underlying technologies that make up the FHIR standard are easily understood by software developers new to healthcare and, once adopted, will make it easy for companies to start building apps that address common needs of EHRs and unmet needs for population health overall.