Dr. David Brailer may be the most important man in health care these days. Tapped last year as President Bush's national coordinator for health information technology, the physician/economist is the point man for the President's plan to help hospitals develop electronic health records. All the care a patient gets -- from hospitals, doctors, or anyone else -- would be tracked in a single, privacy-protected database so that health-care providers everywhere would have the information they need give safe, effective care without duplicating each others' efforts.
The Administration is also tinkering with pay-for-performance experiments that pay hospitals more if they rank in the top 20% in quality, as rated by their compliance with established medical practices. Brailer recently spoke to BusinessWeek E-Business Editor Tim Mullaney. Edited excerpts of their conversation follow:
Q: In the 1990s, people expected health care to get a big productivity boost from the Internet. Most experts agree it didn't happen. Why not?
A: Well, health care underinvests in information technology. We invest on average 2% or 3% of revenue for the typical hospital and less than that for a physician's office. Many industries invest 12% or 15% of their top line in their IT infrastructure.
Health care is an inordinately complicated industry that really lacks the business model for productivity improvement because it's so fragmented, and there hasn't been a force that has been able to consolidate and integrate the industry in a better way.
Q: What's the outlook for changing that over the next 5, 7, 10 years?
A: I think there's a good outlook. People are aging, and the people who are aging have been [familiar with IT] in their professional productive years and they expect to continue [to use IT].
Q: Experts have claimed the industry can easily get $85 billion a year in cost savings out of things like electronic health records. One big Washington lobbying group, citing you, put that number at $140 billion a year. Is that realistic?
A: Well, $140 billion is 8% of health-care spending now. I think that number is actually well founded by estimates that have come in. The Center for Information Technology Leadership is one group that estimated numbers in that range. Other extrapolations have estimated numbers that are, frankly, much larger.
We're trying to unleash consumer choice and better information for consumers. We'll get the gains that happen if a doctor gets better information. Plus consumers will be able to better evaluate their options and more able to make the industry accountable in terms of less duplication or more timeliness of service.
Q: How does what you do help lay the groundwork so these kinds of gains can be realized?
A: We have called for four things to accomplish the President's agenda. One is that clinicians use an electronic health record. There is very very strong evidence that when clinicians use electronic health records there is a substantial health-status benefit.
The second thing we've called for is portability of information. The idea is that information flows with the patient, and when they show up in an emergency room or at a doctor's office or an imaging center, their information goes with them unless for whatever reason they don't want it to.
The third piece is personal health records. We want to see every person have access to a personal health record and be able to communicate with their clinician using it.
Finally, we want to streamline the various federal systems that collect data. The ability to collect data for pay-for-performance is an important thing, but we also collect data for adverse effects for drugs and bioterrorism and clinical trials. That becomes very burdensome to the private sector if we don't have a mechanism for making that more streamlined.
Q: What is the $125 million the President has proposed spending each year to promote health care information technology going to be spent on? And how do you answer critics like one hospital exec we met who called it a spit in the ocean?
A: The estimates of what it costs to [install electronic records at all hospitals and doctors' offices] go from the low billions to as high as $10 billion. What we need to do is determine what it really costs.
The federal government isn't going to give people electronic health records or pay for it or set up these networks. We don't want to see just a regulation that requires the industry to do that. We're trying to figure out how to set up a market where electronic health records are cheaper and more valuable to doctors so we can really harness their buying power.
The money we're spending this year and next year is to set up the market institutions. One good example is a group called the Certification Commission for Health Information Technology. What it does is inspect electronic health records and say this one meets all the reasonable criteria for what's a good system and this one doesn't.
Q: What are the best hospitals doing that the smaller hospitals aren't?
A: They have electronic data repositories that can keep track of what's happening to patients enterprisewide. They have bar-code scanning used by nurses when they're giving drugs to prevent administration errors. They have robotic systems in the pharmacies that reduce the errors of a manual pick. They have fully integrated imaging into their electronic health record. They have personal health records. I just gave you [a few] examples of probably 30.
Q: You see the Institute of Medicine report saying 44,000 to 98,000 lives are lost to medical errors in hospitals each year. HealthGrades, a private consulting firm, says it may be as high as 195,000. What kind of reduction is realistic as we put something closer to the kind of technology corporations have into hospitals?
A: Well, it's hard to know because we don't know the full denominator of how many people really are injured. But I do believe that the number is an underestimate.
So what savings could we get of whatever the number is? Well, if you look at the studies with automated prescription ordering, implemented in ideal circumstances, there is a more than 80% reduction in errors that relate to prescribing choice, dosing, and fulfillment and administration. But I don't think that necessarily means that 80% of the deaths get reduced. My rough calculation is we could reduce inappropriate deaths by 50%. I think the numbers are similar for ambulatory errors.
Q: But drug errors aren't the only kind of faulty medical care. What other mistakes can IT help prevent?
A: I think that [there's a larger issue]. There's very good evidence that these systems improve preventative compliance -- computers remind us to make sure that a patient is on aspirin if they've had a heart attack or to check the stool for blood of someone over a certain age. No one has really ever forecast the death saving that comes from [that] because the deaths [that are counted] are only from errors, which means you do something but do it wrong.
We accept that errors and mistakes and missed things and omissions are just part of health care. I think the lesson from the leaders is no, it can be different. And we have to reset doctor and consumer expectations to expect a much higher level of care.
Q: Medicare is experimenting with a system in which hospitals that provide the highest-quality care will get paid more. Why will that work, and what must hospitals do to keep up?
A: We definitely want to see a market that competes on quality, and we haven't been able to do that because we've never been able to measure quality or health status. We now can not only measure quality but we can measure the leading indicators: We know that if a doctor gives patients aspirin after they've had a heart attack, they will have a substantially lower death rate and fewer heart attacks later.
There may be 10 to 12 diseases where it seems like these golden truths have gotten pretty well evident, that if we follow them we're going to do much better. That's evidence-based care, and [compliance with evidence-based care procedures is] what pay-for-performance is based on. But this is really where information technology comes in. It's impossible to [keep track of this data] without IT in place. I think pay-for-performance will drive the adoption of health IT because it's necessary to measure and report that data. Edited by Patricia O'Connell