Entering the Era of Electronic Medical Records

Electronic medical records (EMRs) have been garnering more public attention of late, particularly since President Bush listed their wider adoption as one of his health care goals for the next decade. EMRs have the potential not only to improve efficiency, but to improve the quality of patient care by reducing the chance for errors and facilitating coordination of care. Their adoption by physician practices remains slow, however, and doctors who have implemented EMR systems do not always make the best use of them. We talked to Robert Miller, professor of health economics at University of California, San Francisco, about his recent work with solo and small group primary care practices who have recently adopted EMR technology.

What has motivated the physician practices in your study to use EMRs?

Robert Miller: I think physicians are realizing that they can do more electronically than they've done in the past. In some cases, they were motivated by financial gains or desire to improve quality. Other doctors thought it would save them time. In a couple of cases, we encountered "technodocs" who were simply happy to play around with technology.

Of the practices in your study that have adopted EMRs, what is the most striking benefit you've seen?

Miller: The financial benefits were fairly substantial. We estimated that benefits were around $33,000 per FTE [full-time equivalent] billing provider, per year. About half was due to efficiency gains, most of which came from FTE cuts, especially reduction in medical records and transcription staffing. Slightly more than half of the gains came from increased levels of coding of office visits. We observed a big shift to the next highest level of coding, which in some areas pays $30 or $35 more per visit. The net effect is $17,000 per billing provider, per year, in coding benefits. We didn't expect it would be as big as that. That should give the actuaries at CMS [Centers for Medicare and Medicaid Services] something to think about.

What are some of the specific costs and obstacles physicians are facing?

Miller: The initial costs are substantial. We came up with about $44,000 per billing provider [e.g., physician, nurse practitioner, or physician's assistant]. Time costs can also be substantial: we found practices and providers were working longer hours on average, once they had put in EMRs. In some cases, it was one to two hours per day, and that could last for over a year. Providers have to change their basic work processes, since electronic documentation is a radical change for physicians.

What about physicians' use of EMRs for quality improvement (QI)?

Miller: One of the things that I was really interested in doing was looking at the obstacles for using EMRs to improve quality. Looking at it that way, the payment problem really comes to the forefront. Once physicians have EMRs, certain things will happen automatically, but substantial quality improvement does not happen automatically. The question is: what incentives do physicians currently have to use EMRs for QI? And the answer is not much, not under the current reimbursement system.

Not surprisingly, we found the QI gains were limited. You get some automatic benefits from greater data legibility, accessibility, and organization. And using the templates can improve quality because they act as a prompt—providers are reminded to ask certain questions. But, in general, we did not find many EMR-enabled QI activities. Only five of 14 practices had practice set reminders, which are used, for example, to remind providers to ask diabetic patients if they've had a hemoglobin test in the past six months. And only two of 14 practices had provider performance reports, which show providers real data relative to national averages or peers. This is an issue, since one of the reasons to be excited about EMRs is the potential to transform quality of care.

If most of the benefits are in "operational efficiency" and not in quality, why should physicians be subsidized for using EMRs?

Miller: Physicians should be paid for using EMRs well, and that's where pay-for-performance comes in. Practices with EMRs are very well situated to gain from pay-for-quality performance because they are in a better position to capture and report on data than are paper-based practices, and have more tools to improve quality. The larger issue is that any type of information technology, including EMRs, is not a silver bullet. EMRs are really a tool to help good things happens, but they are placed in a reimbursement environment that doesn't reward quality. The key is to make other changes that complement the EMR.

Given the large financial commitment EMR systems require, can physicians expect to see a return on that investment? And if so, how soon?

Miller: On average, we found providers paid for their initial costs and cumulative ongoing costs in 2.5 years. After that, it was $23,000 per billing provider per year in net gains, on average. But that's the average. There were three practices that, at the rate they were generating benefits, would take over nine years to pay for EMRs. One practice almost went bankrupt due to billing problems associated with implementation. So, while the average payback periods is 2.5 years, there's also risk in making the move.

What kind of variations in costs and benefits did you observe?

Miller: The costs varied less than the benefits did. Benefits ranged from over $50,000 per billing provider per year to very little. That's the difference in having a payoff in one year and having a payoff that may never happen. Some practices were more focused on obtaining coding benefits than others and some champions were simply better at process change than others. In one practice, the champion was very good at reorganizing care to create quality improvement. He was less focused on efficiency or coding gains. A lot depends on personality, skills, and interests of the champion. We have to think of ways of decreasing the importance of the champion, so practices can reap the benefits, even without an ardent champion.

July 2005