Sunday, January 25, 2009

How the other half lives

It's good every once in awhile to rise above the surface of the health care market and see how the real economy does things. Today the Times had an article about the barriers to adoption of near field communication (NFC) technology, which would allow you to swipe your phone over a reader to make credit card transactions. The technology is already in use today in Japan and in the UK, but it's facing many obstacles in the US market.

As described in the article, the issue is not technology. By 2012, most phones are expected to have the technology built-in, yet the availability of the "wave-and-pay" function could take much longer. As an industry expert explained:

For that to happen, all the players will have to work together to define standards, determine revenue-sharing, expand the network of electronic readers and think through the other parts of what he calls "this 2,000-piece puzzle."

The expectation is that a trade association, the NFC Forum, which represents 150 stakeholders in this field, will forge the way to a solution. Yet, the same industry expert warns:
...it is completely possible that nothing will happen in mobile phones in the next five years if everybody keeps thinking only about their own piece of puzzle.

I have no doubt that they're going to figure this out and we'll be waving our phones all over the place relatively soon. Reflecting on the somewhat similar dilemma we face with respect to healthcare IT, I'm struck by two big differences that make health care harder.

First, we'd be lucky if we had only 150 stakeholders. Part of our dilemma in healthcare IT is that the demand- and supply-sides of the industry aren't just fragmented, they're atomized. On the demand-side, there are over 1000 health insurers in the US, and on the supply-side, almost 8,000 hospitals and 170,000 office-based physician practices. HITSP and CCHIT have done a nice job bringing together the technology suppliers (in the latter case, probably too good a job....), but they're only addressing the technical side of this issue. NeHC is supposed to be a forum to forge consensus on market-blocking issues, but they're a top-down creation of the federal government, not the result of the burgeoning demands of underlying grassroots contituencies.

Second, the benefits of health IT aren't as crisp and clear as easier credit card transactions, so our customers (ie, patients) aren't exactly clamboring for what health IT has to offer. Most of us use credit cards very often (all right, probably too often), so little tiny convenience benefits accrue in an obvious way. Most of us don't use the health care system that often, however, so the convenience factor isn't all that meaningful to a lot of us, and so the appeal has to be on less immediate benefits (safety, quality, etc) that are harder to grasp (and believe).

Like the "wave and pay" issue, the obstacle in health care IT is decidedly not the technology. If we can't get "wave and pay" into the market by 2012, what hope do we have of achieving the President's goal of universal EHR adoption by 2014? It's clearly going to take a much larger "forcing function" than the health care market will be able to muster on its own. The Congress' watered down version of the President's health IT vision clearly isn't going to provide that "forcing function", however, so it looks like we're going to have to place our hopes on health care reform.

Thursday, January 22, 2009

Piety in the House of Representatives

The House versions of the economic stimulus related to health IT are working their way through the Ways & Means and Energy and Commerce Committees. Then, it's on to the Senate. The House bills delay release of 90% of HIT funds until 2011, ostensibly on the grounds that the technology isn't ready for immediate investment. Paradoxically, the House approach works only if physicians start investing in systems right away, and it completely ignores the reality that the problem isn't the technology, it's the lack of a business imperative. The very same technology that they criticize is now expected to solve the mind-numbing flaws that currently plague our health care delivery system.

The House approach takes $20B and gives $2B to a government agency now, and $18B to physicians in the form of phased incentives starting in 2011. However, in order to get these incentives, physicians have to be already using these EHRs and HIE in "meaningful" ways (ie, electronical clinical quality reporting and care coordination) by 2011. Which would mean that for most physicians, they would need to start implementing within the next 18 months, because it takes that long to get up and running on these systems.

In taking this approach, the legislation assumes that each physician will make a roughly $50K investment now on the promise of being repaid for this by Medicare over a period of 5 years beginning in 2011. Assuming, of course, that they can pass Medicare's test on "meaningful" use, even though that hasn't been defined and at present there's little to no infrastructure to allow such meaningful use anyway.

Seems like a tough sell to me. There are very very few places in the country that have regional health information networks, and there are no places that have real infrastructure for electronic reporting of clinical quality data, so Medicare will have a hard time defining what meaningful use is, let alone certifying that physicians have successfully done it. They already tried to launch electronic quality reporting a few years ago in the DOQ-IT program, and it was an unmitigated disaster.

Then there's the problem of implementation. According to Medicare, 30-40% of EHR implementations fail. And the vast majority of the ones that don't fail aren't implemented to inter-operate with other systems or generate good clinical quality data.

The House approach glosses all of this over, however. It underinvests in a technological and organizational infrastructure to guide this massive makeover of 15% of our economy, and overinvests in a misplaced faith that IOUs to physicians will drive individual purchases of EHRs, and this, in turn, will induce demand for the network and implementation infrastructure needed for success. And according to success criteria that we're unable to define at present. And in time to meet the President's goal of ubiquitious adoption by 2014.

It reveals an almost religious belief in the power of incentives, however diffuse, and technology, however complicated, and markets, however dysfunctional, to solve the problems that have left 96% of physicians without a fully functional EHR up until now. Almost touching, really, this kind of faith, misplaced though it may be.

Wednesday, January 21, 2009

I'm speechless...


"In reaffirming the greatness of our nation,
we understand that greatness is never a given.
It must be earned."

Monday, January 19, 2009

Message to Congress: It takes a village to implement an EHR

Well, the health IT legislation is starting to take shape, and it's a little more sobering than the initial speculations of tens of billions of HIT dollars being unleashed on state governments in the next few months. In many ways the recent turn of events is an about-face from the early speculations. The House Appropriations and House Ways & Means Committees approaches have the following policy underpinnings:
  1. Separates HIT spending from the economic stimulus
  2. Focuses first on creating a framework for how to handle billions of dollars of HIT funding
  3. Drives the vast majority of money (90%) through Medicare/Medicaid reimbursement channels
  4. Focuses the role of state governments on areas that require local coordination, tailoring, and governance
  5. Moves ONC beyond "coordinator" to actual owner of administrative infrastructure, with all of the programmatic and fiduciary responsibilities that such functions imply
  6. Makes the Federal government the decision-maker on issues such as technical standards, with input from advisory committees on policy and HIT

I'll admit that I was among those who was getting a little dreamy and even woozy at the thought of billions of dollars flowing into health IT over the next year. Compared to that somewhat heady vision, the House language is surely a disappointment. Yet, like most compromises, it represents progress in certain key areas.

Things I like about the approach are:

  • Balance of state-led and federally-led approaches. I like the idea of a network of regional HIT Extension Centers that work directly with ONC rather than through states. State governments have a role as well, but mostly in the areas of coordination, galvanizing health information exchange, promoting quality improvement and public health, and making sure that under-served communities don't get left behind. I like this approach because EHR adoption is not nearly as state- or local-specific as is HIE, which really does need to be tailored to local markets and conditions. Thus, it makes sense to let the Feds drive EHR adoption through regional organizations, and have states focus on state- and local-level HIE concerns.
  • Incentives for doing stuff, not just for buying stuff. Focus on incentives that require participants to use the technology, rather than just having systems that are "certified". I like that the incentives are tied to quality reporting and health information exchange because I don't believe that inter-operability standards are enforceable without having activing monitoring by certified HIEs, public health entities, and quality data aggregation entities.
  • Resources and authority to ONC to get on with it. Gives the clear message that the federal government has to take a stand on key policy decisions in order for us to move forward. This is not ideal, particularly for standards in a fast-moving, decentralized technology space, but it's not clear to me that other approaches are obviously better. The Federal government needs to set standards for Medicare and Medicaid, so that much makes sense regardless of how standards get determined generally.

Things that I think would improve the House language are:

  • Develop a programmatic overlay to the EHR implementations. Inter-operability and robust reporting don't just happen, they get done. And they won't get done if there isn't an implementation program behind the effort, because the systems are too complicated for individual physicians to do this on their own. There's also too much coordination required with other entities, which can only be coordinated by a formalized program. Therefore, we should cement the link between EHR incentives and the HIT Extension Centers. EHR implementations should be executed through or certified by the HIT Extension Centers, otherwise we'll end up with a lot of really bad retail implementations, just like we have today, because we'll only find out about them ex post (ie, after they've failed and can't deliver on their quality and HIE requirements).
  • More HIT funding should be made available before 2011. Not necessarily the whole $18B, but there are some parts of the country that are ready to meet the new requirements right away, and we should make funds available to them to build on their momentum while the overall program catches up.
  • We should try to go "wholesale" rather than "retail". The current approach to the incentives is to go "retail", meaning physician-by-physician, but there's much more value to be had by going "wholesale", meaning market-by-market. Retail implementations will only mimic, or worse, amplify, the existing entropy of care delivery. Putting a programmatic overlay to "communities" or "markets", such as New York and Massachusetts are doing, creates more effective and efficient vehicles for getting providers to work together, which they do too little of today, and ease the path for them to focus on how to best use technology to improve care across the system, not just in their individual offices.

So, concrete ways to accomplish these goals might be:

  • Designate a couple of HIT Extension Centers right away
  • Formalize the role of HIT Extension Centers so we get more proactive interventions in government-funded EHR implementations to get better assurance that they get done right the first time, rather than trying to rescue them after they've failed
  • Provide additional funding to these HIT Extension Centers for them to provide implementation services to physicians up-front
  • Accelerate Medicare and Medicaid incentives to the markets that these HIT Extension Centers cover
  • Allow aggregation of incentives by community according to a formula that allows providers who share the same patients to implement in a coordinated way, and perhaps provide a "sweetener" to those who organize themselves this way

One thing we should recognize is that by putting most of this into Medicare/Medicaid incentives, and by delaying most of the money until 2011, HIT could be on a collision course with health care reform. In some ways that's good, because we shouldn't be using technology to try to solve the intractable problems of the current system, we should use technology to enable and enhance a better system. Yet, the reality is that we could get to a point where we push off the 2011 date to align it with health care reform. That would get us even further away from the President-elect's goal of ubiquitious EHRs by 2014.

Friday, January 09, 2009

The National eHealth Collaborative

The AHIC successor organization is now the National eHealth Collaborative. As a friend reminded me, imitation is the sincerest form of flattery. We at the Massachusetts eHealth Collaborative thus couldn't be more flattered. And we'll get over the name thing. Really. We will.

In all seriousness, congratulations to NeHC -- we wish you every success!

Tuesday, January 06, 2009

Ready for prime-time

The New Year’s Day issue of the Boston Globe had an article discussing some objections by people concerned that HIT systems aren’t ready for the large-scale investments being advanced as part of the economic stimulus package (“Letter highlights hurdles in digitizing health records”). The critics advocate investing at a slower pace and focusing investments not on purchasing current technologies but on creating new technologies to fix perceived shortcomings in current systems.

I understand the concerns – after all, we’re talking about spending billions of hard-earned taxpayer dollars, and as a citizen and former federal government employee, I see that as a sacred trust. From what I’ve seen though, these concerns are either misplaced or readily addressable and therefore don’t warrant delaying large-scale investment.

It is certainly true that current EHR systems are complicated, cumbersome, and barely inter-operable. They are that way for a reason: US health care delivery is complicated, cumbersome, and barely inter-operable.

The supply-side of health care is unbelievably fragmented. According to the AMA, there are about 670K practicing physicians in the US. Roughly 150K of them are hospital-based and practice in 7500 hospitals, two-thirds of which are community hospitals. The other 520K physicians work on the ambulatory side and, according to the CDC, they are spread across 170K office-based practices nationwide, 80% of which are solo or 2-physician practices. This is a cottage industry where the individual businesses face little market pressure to standardize around anything except billing codes. Not surprisingly, when they purchase technology, they don’t demand standardization either, and indeed, they demand the opposite, namely, that the technology be able to adapt to their non-standardized and idiosyncratic workflows and clinical decision-making processes.

This fragmentation among so many small and independent providers has three negative effects on health care delivery that federal HIT funding can help resolve. First, care is difficult to coordinate. Second, basic reporting for public health and performance measurement does not exist. And third, clinical documentation and data standards are impossible to promulgate and enforce. Federal HIT funding can help overcome these obstacles by giving all users the tools to document and communicate key information according to national standards, and requiring that they do so as a standard of care.

Getting back to the main point then, the critics have it all wrong. We shouldn’t be waiting for better technology, because technology is an ever-moving target driven by technical and scientific improvement and user demand. If we had insisted that Tim Berners-Lee anticipate live streaming of HD video from the likes of YouTube and Netflix, we’d still be waiting for the World Wide Web. Nor should we be spending a lot on “innovation” or “simpler, easier” technologies, because we’ll almost assuredly get that wrong. Governmentally-directed innovation spending would never have come up with Google, Twitter, Facebook, YouTube, Hulu, Yelp, Sermo, and craigslist, and we’d be much worse off for it.

So, government funding is needed, but spent the wrong way it can stifle innovation and just plain waste a lot of money. What we need to do is first recognize that this will take a long time to get right, it’s wrong to try to architect it perfectly in advance, and it will only become mature when more users engage in using technology to accomplish real business needs. With such a decentralized user base, fast-moving technology, and a dynamic, complicated field such as medicine, we should specify as little as we can get away with technologically but create a flexible architecture that can efficiently accommodate changes into the future. New York is working on just such a model.

EHR technology has gotten as far as it can in a thin market – what’s needed now is more bottom-up pressure from more users, and more top-down pressure from policy-makers and businesses to align these users. Federal dollars can facilitate this by creating a large user base and imposing a policy and programmatic overlay to what would otherwise be a funding free-for-all.

My personal recommendations for an economic stimulus funding program would be:

  • Establish goals focused not on technology, but on what we want people to do with technology, such as coordination of care, adherance to guidelines, reduction of medical errors, and improvement of population health
  • Each state should designate an HIE entity (or entities) to broker and enforce statewide health data exchange, and make Medicare and Medicaid data available to authorized users through this HIE infrastructure
  • Require that all clinical entities use the state-designated HIEs to provide patient-specific post-visit reports to each other
  • Require that all clinical entities use the state-designated HIEs to regularly report public health and quality/safety data to state-designated public health and quality data entities
  • Require that all clinical entities use the state-designated HIEs to populate patient health records (PHRs)
  • 90% of funding be earmarked for EHRs, and 10% for state-designated HIEs, quality data warehouses, and public health reporting infrastructure
  • Require that state-designated HIE, public health, and quality data entities monitor and enforce health data exchange according to existing HITSP standards for data exchange and existing quality and safety measurement standards established by AQA, NQF, HITSP, and others, and penalize states that don’t do this

Is it scary to spend so much taxpayer money so fast? You betcha. But that’s true for every part of the economic stimulus package, not just health care. The need is great, however, so we need to roll up our sleeves and put in place the right vision, leadership, and management. The health IT infrastructure and experience base is perfectly poised to make excellent use of such funds to accomplish the goals of immediate economic stimulus and improvement in health care. By outfitting physicians with modern tools, and requiring that they use them to achieve societal goals, our federal stimulus dollars will provide returns to the country for years to come.