By Cindy Atoji

April 8, 2008 | Francois de Brantes is a staunch believer that health information technology (HIT) should be an integral part of pay-for-performance (P4P) programs and will benefit physicians. Yet a recent Massachusetts Blues study concluded that electronic medical records (EMRs) may not pay off for physicians, and a related study showed fewer health plans are including technology use in their P4P calculations.

de Brantes heads the not-for-profit Bridges to Excellence, a coalition of physicians, health plans, and employers dedicated to providing physician performance incentives to improve quality of care. Technology is an integral part of accomplishing P4P, insists de Brantes who recently spoke with Digital HealthCare & Productivity about why he’s right and the recent studies are wrong.

DHP: What’s your response to this recent study?

Francois de Brantes

de Brantes: BCBSMA (Blue Cross Blue Shield of Massachusetts) was determining what type of mandate to put on physicians and hospitals as a condition of participation in its incentives program. They decided that they could mandate the adoption and use of CPOE (Computerized Physician Order Entry) systems in hospitals because the hospitals have capital, and there is a return for the facility (not counting the patients). They also concluded that they could not mandate adoption by physicians because it would take far too long for physicians to recoup the cost of their investment, and that’s true.

However, the premise seems to be the wrong one, and certainly not one that BTE supports. Our data show very clearly that when physician practices go through a transformation which is evidenced by the adoption and use of HIT (including EMRs) and the production of good results in the management of patients in the practice, the quality dividend is huge — somewhere around $300 per patient per year — and accrues directly to the payer in terms of lower overall episode costs of care. 

As such, if the payer were to offer an incentive to the physician in the form of shared savings, there would be a huge incentive for the physicians to adopt and use the EMRs, and their ROI would materialize quickly.

DHP: What are the big issues around HIT incentives?

de Brantes: The issue with incentives around HIT are numerous: First, there’s the “free ride” effect. Say that Plan A, which has 20% of the patients in the physician’s practice, creates an incentive program to encourage the adoption of HIT in that practice.  If the incentives are big enough, the physicians will adopt the systems.  All the other plans will also benefit from this HIT adoption and they will have had a free ride because they didn’t have to do anything or pay anything for this change to happen. Many plans are reluctant to offer those incentives for this reason. Interestingly, employers are not hesitant at all.

Many plans have also failed to look at incentives for adoption and use of HIT as a necessary but insufficient part of an overall incentive program. They tend to carve the HIT portion out and since the ROI for that carve out isn’t clear, adoption is weak. When you explicitly tie the adoption to the use and demonstrate good results, then the outcomes are very different.

As far as collaboration, BTE’s studies clearly show that plans benefit greatly from collaborating with each other on the physician-focused quality incentives. Yet few do and as a result, their [individual] programs are often ineffective, which leads them to put small dollars at stake in incentives, which of course leads to further failure, and the cycle continues.

There are, however, many plans around the country that have gotten the message and are actively collaborating, tying incentives around HIT to the delivery of good results in the management of patients, and also understanding that creating incentives for the adoption and use of HIT is an essential building block to the transformation of care in physician practices.

DHP: What does that really mean; how can HIT help transform practices and produce fair pay for performance?

de Brantes: The broad definition around health information technology to me goes from a sophisticated registry to a fully compliant CCHIT electronic medical record system. But the base you need to have is that registry function: you need to have a database that systemically collects information on your patients.

The entire concept of Bridges and what we’ve been trying to do for years is to cause the re-engineering of practice of medicine and convert the management of patient care in physician practices from what is mainly a reactive patient management system today to a highly proactive management system. It doesn’t take a fully-compliant EMR system [to do that], but you need to have a database and registry. To me that’s the single biggest barrier to transformation in any physician practice in the country. I often think that in rush toward full EMR implementation we’ve forgotten the importance of that very this first initial step, of getting the physician practice an organized database.

The reason it is such a critical part of the transformation process is that the first report that comes out of that information system is almost a life-changing moment for the physician. For most of them, it is the very first time that they have a credible, reliable, and actionable report card on their patient panel. This is a transformational moment because for the most part that report card shows a level of performance that is below that physician’s expectation. They always think they are doing better than they are.

Until you get to that point, it’s actually difficult for the physician to think about “how do I reorganize my processes so that I can close the quality gaps that have been identified.”

And logically, you do end up with a fully functional EMR system with clinical decision support, but the way you create that ultimate system is by starting to understand where are my own failings, where are my own quality gaps, and then what help do I need, what types of clinical decision support tools, and what should they to be focused on, which then lead to decisions as to what type of EMR system do you want.

DHP: BTE recently announced a physician performance assessment program. How will that work?

de Brantes: Assessment of physician performance is usually done regionally by either using claims data or through data from paper-based medical charts, which can be laborious and expensive for physicians and health plans. These methods create long lead times for review, analysis, and performance reporting. It’s been thought that assessing physician quality through the use of medical records was impractical and too costly but as EMR adoption grows, direct submission of data from EMRs can work and yields better measures than claims information alone.

Up until now no one has attempted to stream performance assessments in a nationally-standardized way. Only regional or localized efforts have been implemented. We are working with performance assessment organizations such as MassPRO, the Minneasota Community Measurement Collaborative, and the National Committee for Quality Assurance to measure physician quality using medical record data.

DHP: Which EMR organizations and Health Information Exchanges will initially participate in the initial program?

de Brantes: We’re creating a series of real-time pilots with performance assessment organizations using EMR data. Five EMR organizations and two Health Information Exchanges are participating. We’re working with EMRs such as NextGen, E-clinical Works, GE Healthcare’s Centricity, patient registries, and other information systems in doctors’ offices including BioSignia’s Know Your Number disease-specific risk assessment tool.

So the economics and potential impact of medical record-based performance assessment is changing. On the community side, the existence of regional health information networks such as the Massachusetts eHealth Collaborative,  HealthBridge in Ohio, the Indiana Health Information Exchange, and the Taconic Health Information Network and Community in New York also create a way through which EMR data can be aggregated for performance assessment.

DHP: What are some best practices in pay-for-performance?

de Brantes: The first one is credible and reliable data. If the data isn’t credible, you’re not going anywhere — so if it’s two-year-old data, no one cares. You really need to have what we refer to as rapid and frequent performance feedback loops. If I continue to refine treatment regimes, I should be able to see the results every quarter, not once every so many years. And the more HIT is deployed, the greater the precision on the performance measurement.

The second is around the predictability and the size of the incentive itself. Physician practices are businesses, and even if you’re not a huge for-profit business, you still make decisions on the relative benefit derived from a specific investment. I summarize it to health plans by saying the juice has to be worth the squeeze for the doc. They’re not going to spend money on an EMR system unless they know the benefit it will provide.

The final one is giving the physician the opportunity to actively participate in quality improvement efforts because there’s nothing simple about fundamental practice re-engineering and there are not a lot of resources available to physicians to re-engineer their practice. It’s not like I can call the Geek Squad for physician offices. The payer needs to be sensitive to the fact that the more demanding the program, the more successful it is in causing fundamental change in behavior of the practice, the more difficult it is for practices to figure out what to do.

DHP: How can health-IT vendors work closely with P4P programs?

de Brantes: There’s a huge variation in what the vendors offer to practices. Most of time when you have EMR implementations, you start at time zero and move forward. Very seldom is there a look back, because it requires a huge amount [of effort] to look at all your paper records and start translating them into an EMR database. But that to me is a necessary and critical step in an implementation by any vendor.

I don’t see many EMR vendors talking about this. They talk about their clinical decision support, audits, things that can help the physician ultimately achieve good scores on their performance measures, but I think they really need to sit down with the practices, and say, “unless you have a good starter set of data in your EMR database, it might take a year before you can apply for and qualify for incentives under a pay for performance system.”

DHP: On a 1-5 maturity scale, where is pay for performance, and how can it grow?

Brantes: I think it’s at a 3 1/2. So it’s way past infancy but still maturing. To me the signs of maturity are the fact that more health plans have stopped designing their own mix and are really standardizing around a core set of programs. But old habits are hard to [break.] And at the physician level, you don’t get the practice transformation unless the signal to the doctors is really strong. Otherwise they don’t pay attention. And so we’re still evolving.