Wednesday, January 21, 2009

Five "Shovel-Ready" Health Care Reforms

Microsoft Health Vault’s leader Peter Neupert has a wonderful blog post that makes two important points really well. One message is that health care reform is about the outcomes, not the technology. We should think expansively about which technologies to invest in, based on the results we want to get.

The other message is the economic stimulus package is different than the reform effort. It is moving at hyper-speed through Congress, and it may be difficult for staffers and other advisors to sort through and incorporate what may seem like opposing Health IT views against a backdrop of traditional ideology and extremely forceful special interest lobbying.

Even so, there’s consistency among the health care professionals who worry about these issues all the time. Peter unexpectedly discovered that the messages of his fellow panelists from the Health Leadership Council, the National Quality Forum, the Permanente Federation and the General Accounting Office were remarkably in sync with his own testimony to the Senate Health, Education, Labor and Pensions Committee.

Congress is about to make some big moves in health care that will require immense resource expenditures but, depending on what we pay for, may or may not bear the fruits we hope for. They should move carefully. Not all health care reform has to be labyrinthine. Not all ideas must require huge cost or take years to come to fruition and gain market traction. There are relatively simple actions that are available now, and that the Obama Health Team could tackle to effect tremendously positive, immediate impacts on the system.

Of course, right now the Health IT industry is focused on the promise of a huge stimulus windfall that would be dedicated to their products. But the opportunities we describe below follow principles that have broad support among students of the health care crisis. Two would change the way we pay for health care services, tying payments to documented results. Three are based on how we pull together and make use of the data that can drive clinical and financial decisions, and they overlap, though not perfectly, in their potential. Still, if any system adjustments can be passed through policy initiatives that focus on what’s best for the common rather than the special interests, these should be among the most straightforward.

Re-Empower Primary Care
There is general agreement that primary care is in crisis, the result of years of abuse and neglect by the medical establishment and by CMS. In simple terms, the primary care/specialist ratio in the US is 30/70. In all other developed nations, its about 70/30. And our costs are roughly double theirs.

We should allow primary care physicians to do the jobs they were trained for, changing their roles from “gatekeepers” to “patient advocates and guides.” We should immediately start financially rewarding them for collaborating with specialists to manage patients throughout the full continuum of care. Keep in mind that, as the Dartmouth Atlas and other studies have made clear, most health care waste is concentrated in the sub-specialties and in inpatient settings, incentivized by a fee-for-service reimbursement system that rewards more procedures, independent of their utility. One very thoughtful approach to invigorating primary care has been advanced by Norbert Goldfield MD and colleagues.

Of course, truly re-empowering primary care will require more than just paying primary care physicians more. Higher reimbursements will help them afford to spend more time with each patient, yes, but PCPs also need help acquiring tools that can help them better manage those patients. And they need the authority to work collaboratively with specialists. Challenging, but certainly doable and important!

Changing America’s current imbalance between primary and specialty care should drive significant downstream waste from the system, dramatically improving quality and reducing cost.

Increase the Incentives For Programs That Tie Payment To Outcomes
Projects like the CMS/Premier Hospital Quality Incentive Demonstration (HQID), in which 250 participating hospitals got 1-2 percent bonuses for achieving quality improvements, have clearly demonstrated that incentives work. The hospitals that pursued the incentives made greater strides in quality improvements than their peers who did not work toward the incentives.

But we need to make the financial incentives large enough to drive real paradigmatic change. Too many programs offer incentives that are trivial in the minds of providers. Does it make sense for physicians in small, busy practices to rework their office flows to try to meet the challenges associated with hitting targets in exchange for a 1 or 2 percent financial bump, tied to a fraction of their patient population?

Now that there’s no question that incentives work, we could easily give these programs teeth by raising the incentive antes to 15 or 20 percent, while also demanding commensurate levels of savings. And we should go in, understanding that the goal is to drive out unnecessary care, and create expectations that, by managing better upfront, the total spend will be lower.


Establish a National All-Payers Database
Data sets, including those comprised of health care claims, must be large to generate credibly useful information.

But health care is financed through many different payer streams and by many players within each stream. Nearly all treat their data as proprietary, and information remains fragmented. So, for example, physicians rarely receive useful information on their complete pool of diabetic patients: instead, they get small slices of data from each payer, each analyzed using a different proprietary methodology. Or, we fail to accumulate adequate sample sizes to identify which treatments, interventions, drugs, devices, health plans, physicians or facility services provide the best value.

But merging those data across payers and making the aggregated set freely available would create the basis to identify true evidence-based best clinical and administrative results. Based on hundreds of millions or billions of records, we might be able to credibly identify which professionals, services or approaches most consistently produce the best results within value parameters. The data set would always be building, providing an always slightly-new base for answering our most difficult questions. Together with the analytical tools that are also becoming stronger and more refined, the potential is vast.

Of course, health plans, always politically formidable, might fight tooth and nail to maintain the competitive advantage they believe is inherent in their data. But health care is a special enterprise, with objectives that are ultimately rooted in the common interest, so they have no real excuse to refuse this. And health plans, like the rest of us, would gain access to much larger data sets that can be mined to advantage.

There also are precedents here. Several states have already begun to establish all-payer databases. At a June 2008 meeting, a presentation on Maine’s experience highlighted 3 fundamental, telling principles that are challenges to any effort.

1. Nobody wants to pay to develop and manage the database.
2. Nobody wants to contribute their data to the database.
3. Everyone wants the aggregated data that develops in the database.

The solution: make it a national effort, paid for by CMS, and with mandatory participation, user fees, and open access to the data.

Create Uniform Nationally Accessible Disease Registries

Many physicians have come to appreciate the value of disease registries. Registries allow clinicians to count all active patients with distinct conditions, e.g. hypertension or diabetes. They can track characteristics within a patient subset, e.g. diabetic patients on a particular medicine. They can monitor and stratify patient status and progress within each group, and generate reminders and alerts to assure guideline level care. And they can identify trends in performance and, with relative ease, get a sense of what works and what doesn’t.

Even so, many registries are still in silos, meaning that the sample sizes remain small and that the parameters that define the registries’ characteristics often vary between implementations.

What we need are freely available, Web-based registries with easy data entry and easy querying capabilities. The impact on our management of patients with chronic illness, who consume 70 percent of our health resources, would almost certainly be powerfully positive.

Release Medicare’s Physician Data
Nearly a year and a half ago, the consumer advocacy organization Consumer Checkbook sued the US. Department of Health and Human Services (HHS) for the Medicare physician data in four states and DC. HHS argued that physicians have a right to privacy, even though, in the case of Medicare and Medicaid, they are vendors taking public dollars, and even though hospitals do not enjoy the same protection from scrutiny. In August 2007, the court held with Checkbook, and on the AMA’s “advice,” HHS promptly appealed, locking up the data for the duration of the Bush Administration.

The large commercial health plans have traditionally considered their claims data proprietary and so have not made their data sets publicly available. Self-funded health plans, administered by Third Party Administrators (TPAs), develop sizable data sets but have resisted collaborating, and have also not expressed an interest in making their data available.

So for those outside the health plan community, there are few, if any, data sources with sample sizes large enough to accurately evaluate and profile physician performance. This is significant, since studies have shown that there can be profound differences, 6x-8x, in resource consumption (i.e., cost) between the least and most expensive physician (within a specialty and market) to obtain the identical outcome.

In other words, not all doctors perform equally. While more patients are paying out-of-pocket for a larger portion of care, there is still virtually no credible information to guide their physician choices.

The American people could quickly learn which physicians within a specialty and a market consistently get the best outcomes at the lowest costs if Medicare physician data were made publicly available. Releasing these data would also put pressure on physicians everywhere to understand their own numbers, and to improve if their performance values are lacking. We see this as beneficial to the great majority of physicians who seek excellence in their work.

Smoothing the Way

American health care is a vast enterprise in which millions of professionals and hundreds of thousands of organizations vie for an ever larger portion of what has historically been an always growing resource pool. The chaos and dysfunction that has developed in health care is largely due to two system characteristics. One is the fee-for-service reimbursement system that has rewarded more rather than the right care. The other is a lack of transparency that prevents us from knowing and understanding performance, even when that performance is dangerous: what works and what does not, which approaches are high and low value, who does a good job and who does not.

The five action steps outlined above would allow us to better identify the problems and opportunities in our health system, as well as the strongest solutions to drive decision-making. Then they would leverage that information to create strong incentives for the right care, organically changing the dynamics of care and reimbursement and, to the degree possible, smoothing the transition required to heal the way we supply, deliver and finance care in America.

Tuesday, January 6, 2009

Let’s Reboot America’s Health IT Conversation Part 2: Beyond EHRs


Yesterday we tried to put EHRs into perspective. They’re important, and we can’t effectively move health care forward without them. But they’re only one of many important health IT functions. EHRs and health IT alone won’t fix health care. So developing a comprehensive but effective national health IT plan is a huge undertaking that requires broad, non-ideological thinking.

As we’ve learned so painfully elsewhere in the economy, the danger we face now in developing health care solutions is throwing good money after bad. We don’t merely need a readjustment of how health IT dollars are spent. We need to reboot the entire conversation about how health IT relates to health, health care, and health care reform. To get there, we need to take a deep breath and start from well-established and agreed-upon principles.

Most of us want a health system that, whenever possible, bases care on knowledge of what does and doesn’t work - i.e., evidence. We want care that is coordinated, not fragmented, across the continuum of settings, visits and events. And we want care that is personal, affordable and increasingly convenient.

Most of us also agree that, so far, we have not achieved these ideals. In fact, health care continues to become costlier, quality is spotty, and the gap between the health care we believe possible and the current system is widening.

We believe that most health care professionals are acutely aware that more health IT alone cannot resolve these problems. Despite billions of dollars in health IT investments by health care professionals and organizations, the gap persists and is widening. Many physician practices have expanded their health IT functions, moving beyond electronic billing systems - a necessary asset to be paid by Medicare - toward EMRs and from paper to software systems. About a quarter of US physicians use EHRs from commercial vendors. Hospitals and health plans - larger, corporate organizations with more dedicated capital resources - have implemented health IT more quickly. Even so, the tools implemented have typically been focused on record-keeping and transactional processing, not decision-support. Health care clinical and administrative decisions have not yet become more rational, less tolerant of waste and duplication, or more congruent with evidence.

We don’t need simply more health health IT; instead, we need an array of specific health IT functions and capabilities that can facilitate better care at lower cost, and the adherence to evidence-based rules.

What would those empowering health IT products look like, and what would they do?

Focusing on Decision Support

Most important, new health IT would help patients, clinicians, managers and purchasers make the best possible clinical and administrative decisions. This includes identifying risks and following the best path to lowering them whenever possible. Health IT should help people stay healthy and avoid illness through active clinical decision support, and make sure that the system recognizes value. Which patients, according to past data, have acute or chronic conditions that need care? Which, do the data show, are the most effective (or high value) doctors, hospital services, treatments and interventions - so that the market can work to drive efficiency. Given a particular set of signs or symptoms, lab test results, or genetic test, what is the best next step in care?

Technology and information engineering is readily available to do this. Car technologies now help drivers understand when a problem is occurring, or is likely to occur, monitoring and communicating fluid levels, tire pressure, maintenance appointments, and location in case of emergency. Banking technologies can flag suspicious credit card purchases and can instantly invalidate charge cards. Recently, Google trended flu searches to help estimate regional flu activity; their estimates have been consistent with the CDC’s weekly provider surveillance network reports.

By comparison, most health IT is relatively unsophisticated. In general, the prevailing front line tools do not yet help clinicians identify individual- or population-level health risks. They do not yet provide guidance with evidence-based approaches that can best mitigate those risks, create alerts and reminders, or help monitor adherence to care plans, even though the data are now clear that most Americans die and we pay the most money due to easily preventable and managed conditions.

In short, we monitor our cars and bank accounts better than we do our health. We can change this.

Untethering Patients with Easily Accessible Personal Health Information
High value health IT would improve care by making summary personal health information available to providers and patients, increasingly independent of location and time. Most health records are still tied to a health care organization’s data center, supporting an outdated business model in which the patient must come to a centralized, expensive location for even the most routine tasks, like history-taking or lab testing. Most current EHRs don’t change this, in large part because they aren’t connected to the Internet yet. Web-enabled patient information would untether the patient, and make increasingly standardized care more readily available anywhere. De-coupling health information from health care providers is the first step in the development of new business models that will offer team-based care services wherever one is located, saving money and increasing convenience.

Empowering Patients Through Online Linkages with Clinicians and Other Patients
High value health IT will link patients with clinicians, will match problems with the most appropriate solutions, and will use social networking to increase access to patient- and condition-specific information, knowledge, and guidance. This class of health IT applications and services will be particularly useful with chronic illness, shifting more of the condition’s monitoring and management to the patient and his/her family and peers, with diminished reliance on the office-based physician and the single visit model of care. Bringing advances like these to fruition will require much broader implementation and access to broadband and mobile technologies, as well as standardized health record formats that use XML, like the Continuity of Care Record (CCR).

Supporting Participatory Medicine: Bridging the Medical Home and Web-Based Care
As Kibbe and Kvedar recently wrote, much of the health IT we’re describing here bridges the divide between two powerful trends: Health 2.0 (or user-generated health care ), and “the medical home.” It is now clear that, while most health care consumers want to be more actively engaged in their own care management - e.g., using Web-based search and joining patient communities - they also want to be connected to their physicians for questions and care when appropriate. The way forward here is Participatory Medicine that combines and remixes health information and knowledge - some from experts and some from the crowd - in the interest of helping us live healthier lives. Here is a very good description from Neal Kaufman, MD, a practicing pediatrician and the CEO of DPS Health, about how this will work:

…organized medicine needs to provide the day-to-day support patients need to prevent disease and to self-manage their conditions if they are ill. In the connected era that means just in time delivery of the personalized and up-to-date data and information a person needs to have the knowledge to make wise choices. It means supporting patients to easily and accurately keep track of their performance. It means providing tailored messages and experience that speak to each person based on their unique characteristics, their performance on key behaviors and their needs at that moment in time. It means helping patients link directly to family and friends for critical support, and link to their many providers to help integrate medical care with everyday life.

Making Data and Accountability the Routine By-Product of the Use of Health IT
Health IT can help make all health care professionals and organizations - physicians, hospitals, other providers, health plans, drug firms, device firms - more accountable stewards for quality, safety and cost results, and for the engineering required for continuous improvement. We can learn from our current supply, care delivery and finance processes in the same ways that Toyota and Wal-Mart monitor their internal business processes.

But we need to design data aggregation into the products from the start, not as an afterthought. The problem is not just that we lack some important data elements to carry out these analyses now. More to the point, we have not committed nationally to aggregating, analyzing, and reporting the massive amounts of health data that we already have. Similarly, due to a lack of incentives and competing interests, most professional and organizational health care players have resisted using data to improve the quality, safety and cost of American care.

Interoperabilitiy of various EHRs is absolutely critical to the ability to cost-effectively collect, manage, and report outcomes data. All health IT products used in the care of diabetic patients, for example, ought to be required to export performance data relevant to care of diabetes in standardized formats. All research of any kind depends on this capability.

Removing the Complexity and Cost Associated with Multi-Payer Claims Administration
Health IT ought to make claims payment, eligibility look-up, co-pay verification, and other administrative processes simpler, easier, and faster for providers, patients, and family members. There is no good reason why we don’t currently have an all-payer clearinghouse for patient administrative and financial information that is standards- and web-based. There also is no good reason why, in the era of PayPal, physicians and hospitals experience Days in Accounts Receivable of 36 and 55, respectively. As Rick Peters has written recently, it is time for us to build a scalable, XML, and cloud-based claims adjudication, public health, and quality reporting system to replace the entire archaic mainframe systems at CMS and their fiscal intermediaries. “Make the winning solution open source, implement it for Medicare and the CDC, and offer it free to every state Medicaid program and all the commercial payers,” he says, and we agree it is time to use updated technology to resolve the inexcusable claims administration mess.

Closing the Collaboration Gap
Finally, a new generation of health IT platforms and services will close the “collaboration gap” that exists between the system’s many sequestered players, who as a result perform so much less effectively and efficiently than they otherwise might. Clinicians, for example, diagnose disease and set up treatment plans but often are isolated from helping patients cope, manage, or adhere to these plans. Patients, once diagnosed, are motivated to manage their illnesses but often have few tools or methods to assist them. Purchasers and payers want to see clinicians use the most efficacious resources, but typically do not have a way to inform and reward evidence-based purchasing processes. In every case, health IT can facilitate a more collaborative experience that is tailored to the user’s purpose, no matter what role that user plays in vast health care space.

Health IT presents enormous, unprecedented opportunities to improve the quality of care, to dramatically reduce the waste and cost inherent in our current approach, and to culturally transform physicians and patients so both become more actively engaged in improving health and health care. Bringing the fluidity of health information and knowledge that is just starting to fruition will allow us to leverage the true power of information engineering, and that can take many forms. We think the name “clinical groupware” is more appropriate to this new class of health IT products and services than is the term “EHRs.” In any case, the real health IT challenge to the Obama health care team is to step back, take stock of the kinds of applications that are emerging in the domain of health IT, including EHRs, and create an expansive, open policy structure that can leap beyond the status quo and really change the way American health care, in all its facets, works.

Monday, January 5, 2009

Let’s Reboot America’s HIT Conversation Part 1: Putting EHRs in Contex


On Dec. 19, we published an Open Letter to the Obama Health Team, cautioning the incoming Administration against limiting its Health Information Technology (IT) investments to Electronic Health Records (EHRs). Instead, we recommended that their health IT plan be rethought to favor a large array of innovative applications that can be easily adopted to result in more effective, less expensive care.

The response to that post was vigorous. We received many comments and inquiries from the health care vendor, professional and policy communities - urging us to provide more clarity. One prominent commentator called to ask whether we, in fact, supported the use of EHRs. We both have been active EMR and health IT supporters for many years. Dr. Kibbe was a developer of the Continuity of Care Record (CCR), a de facto standard format for Electronic Medical Records (EMRs), and has assisted hundreds of medical practices to adopt EHRs. Dr. Klepper has been involved in EMR projects for the last 15 years, and the onsite clinic firm he works with provides every clinician with a range of health IT tools, including EMRs.

That said, we are realistic about the problems that exist with health information technologies as they are currently constituted. As we described in our previous post (and contrary to some recent claims), most products are NOT interoperable, meaning licensees of different commercial systems - each using different proprietary formats - often find it difficult to exchange even basic health care information.

Most EHRs are bloated with functions that often are turned off by practitioners, that are promoted politically through the current CCHIT certification process, and that drive up costs of purchase, implementation and maintenance. Despite moving toward Web-based delivery models that have MUCH lower transactional costs than old-fashioned client/server approaches, most commercial offerings are still extremely expensive, especially compared to the revenue flows of the relatively small operations they support. (Dr. John Halamka’s recent recommendation that the Fed invest $50,000 per clinician for rapid implementation of “interoperable CCHIT certified electronic records with built in decision support, clinical data exchange, and quality reporting” provides an idea of the resource allocations that are on the table.) The very wide range of choices in the market currently raises the question of whether the implementation of a national EHR infrastructure MUST be so costly.

Many health care professionals still think of health IT as a compartmentalized function within health care organizations. But health IT has increasingly become the glue between and across all health care supply chain, care delivery and financing enterprises. In the past, it was enough for health IT to facilitate information exchange inside organizations - in which case a proprietary system would do - but we now expect information to be sent and received seamlessly, independent of platform, including over the Internet. Most of the currently dominant EHR technologies don’t even begin to get us there.

Nor, despite the rampant optimism about its potential, can a focus on health IT alone - or even more emphatically, EHRs - resolve health care’s deeper problems. As the noted health care economist Alain Enthoven wrote in a December 28 New York Times editorial:

[President-elect Obama]… has suggested, for example, that electronic medical records could save Americans nearly $80 billion per year. But information technology cannot bring meaningful savings if it is used in a health care system that regularly rewards waste and punishes efficiency, as ours does.

In other words, as the recent reports from the Congressional Budget Office and the Dartmouth Atlas point out (yet again), real reforms will require an array of significant changes, many of which will face withering opposition from entrenched interests. One of those interests is the established health care information technology sector, which stands to finally win handsomely from huge Federal investment in their current products.

The good news is that this is the position held by Peter Orszag, the incoming Director of the Office of Management and Budget, the current Director of the Congressional Budget Office, an astute student of health care dynamics, and a key member of the Obama health team. In July 18, 2008 testimony before the Senate Finance Committee, he said:

The bottom line is that research does indicate that, in certain settings, health IT appears to facilitate reductions in health spending if other steps in the broader healthcare system are also taken to alter incentives to promote savings. By itself, however, the adoption of more health IT is generally not sufficient to produce significant cost savings.

In other words, it is fair to be skeptical about how we should proceed with a national health IT build-out effort. The health IT industry’s current product/service offerings are analogous to the auto industry’s obsession with SUVs, as much the problem as the solution. Just as the auto industry can be re-purposed to build lower-energy, less wasteful vehicles, so too should the health IT industry be encouraged to offer smarter products that serve the interests of an affordable, convenient, and evidence-based health care system.

A smorgasbord of Health Information Technologies is available to help us build a far better health system. Part 2 will describe some functions that a national health IT infrastructure renewal effort might consider.