Information technology has the potential to transform the way health care is delivered and received. It can help to reduce errors, improve efficiency, and empower both patients and providers. In the first study of its kind, Ruben Amarasingham, M.D., M.B.A., the associate chief of medicine at Parkland Health & Hospital System in Texas and an assistant professor of medicine at UT Southwestern Medical School, and his fellow researchers directly measured physicians' use of health IT in hospitals. From their survey of 41 Texas hospitals, he and his team found that those with automated clinical information systems—like electronic notes, test results, and decision support—had fewer complications, lower mortality rates, and lower costs. But, according to the researchers, improvements like these can only accrue if physicians are properly trained and invested in using the systems implemented by their hospitals. We talked to Dr. Amarasingham about the way technology and culture can intersect to help providers and hospitals deliver high-quality care.
Can you describe some examples of how physicians and other providers use IT to make a difference in how care is delivered?
Ruben Amarasingham: Medicine is extraordinarily complex. Every day new studies and new techniques emerge. On top of that information explosion, you have an increasingly fragmented delivery system. You may have four or five consulting physicians, a primary physician, different sets of nurses, ancillary staff, physical therapy, nutrition, pharmacy—all taking care of the same patient and all operating on different schedules and potentially with suboptimal communication.
Electronic medical records have the potential to be a coordinating center for all that activity. In addition, electronic decision support helps manage the information explosion by offering additional ideas about what could be done for patients based on best-practice literature or local policy. Electronic systems have great potential to provide an additional “set of eyes.” For example, a cardiac patient may be admitted and the cardiac problem will be the appropriate focus of the team. But additional problems might emerge—for example, dysfunction in other organ systems. An electronic information system may be able to pick that up and present it to the physician before the condition significantly deteriorates. That’s vital—particularly now, with physician teams sometimes managing 15 or 20 or 25 patients.
Information systems also help standardize care. There’s been criticism that American medicine is highly variable. There are, of course, lots of exceptions and subtleties in medicine, but information systems can bring the level of care closer to a standard practice, and perhaps that way, fewer errors will be made. Over the years, my hope is that important practices do become standardized, with variation only seen in highly complex cases.
As you were surveying the hospitals, what kind of benefit did you expect to find for those hospitals that used clinical information technology?
Amarasingham: I did expect positive findings, because I’ve worked in hospitals with information systems and have been struck with the benefits of well designed systems. But we weren’t expecting the consistency and strength of the findings—that was impressive.
Across a variety of functions, patients admitted to hospitals that had carefully automated their information systems had significant reductions in odds of mortality, complications, and cost. These results were the most impressive. But it was also interesting to see, in certain cases, increased odds of risk of complications associated with electronic documentation. There are two possible reasons. One is that despite the benefits of electronic clinical documentation, there may be unintended consequences that we need to more fully explore, but the second, and in my opinion more likely explanation, is that electronic documentation enables hospitals to find or extract evidence of complications that weren’t noted before in the paper record. Ultimately this will help hospitals improve performance. We’ll have to explore this carefully as we approach an era of widescale implementation of electronic health records.
I can say from my own experiences that the wealth of data that becomes available when you have electronic medical records is staggering. There are a lot of things you can measure and identify that you were not able to do previously.
Your work showed that investing in software and hardware is not sufficient—the technology must work in concert with people and culture. How can hospitals create the right kind of environment and help providers interact effectively with technology and information?
Amarasingham: That’s the most important question, and it will require more research. The first issue is that hospitals need to fully involve clinical staff, which is very difficult. Clinicians are extremely busy and may take a dim view of organizational reengineering efforts, perhaps because they’ve seen them before and may perceive them as the “flavor of the day.” Sometimes they feel their views are not fully appreciated or understood. When a hospital administration decides to embark on electronic health records, it really needs to go to extra lengths to get staff involved.
The issue of aligning technology with the organizational culture and workflow is critical. In practical terms, this means that if a technology imposes a harmful change to a good workflow, then the technology has to be modified or sacrificed. In the worst-case scenario, this means abandoning the approach or significantly modifying the technology. When hospitals purchase information systems, they may think they’re done or halfway there. And really, it’s just the beginning.
Have you seen examples of this happening successfully?
Amarasingham: Two of the hospitals I’ve had the privilege of working in—Johns Hopkins Hospital in Baltimore and Parkland Hospital in Dallas—have been exemplary in this regard. They’ve been able to establish early staff engagement and identify physician champions who are well respected by the staff. They also have clear appreciation for clinical workflows—leaving true and tested clinical workflows alone and molding the technology around it.
Sometimes that’s not possible, and that’s where physician and nurse champions play a role. For instance, in cases where workflows or processes are very dated, poorly constructed, or based on bad habits or “work-arounds” in the paper environment, you need a colleague to shape the discussion and move people forward, and it’s very difficult for non-clinicians to do that.
In addition to publishing your results, are you feeding information back to any of the hospitals?
Amarasingham: Absolutely. We provided scorecards for each hospital and showed them their IT scores in relation to other hospitals. Most agreed with the findings, and many responded that they were going to use the results to change their strategic IT plans. There was recognition among the hospitals that did poorly that they needed to reach out to their physicians and change some of their IT objectives. I think the instrument we used could be valuable to hospitals to gauge how usable their systems implementations are.
What do you feel is the most important thing President Obama can do to encourage the adoption and use of health IT?
Amarasingham: I think the $19 billion stimulus was an outstanding start. I think it would be wise to tie the funds to specific requirements that can be measured. For example, if hospitals and clinics receive federal money, it should be premised on achieving very specific outcomes, including reporting measures that could only be collected and assessed by well-run information systems. In addition, systems should be built for interoperability and the capability to share and extract information across systems. This may be the element that has been least explored: how this vast amount of data can assist in policymaking and the creation of a high-functioning, efficient, cost-effective system. Having all this data, properly analyzed and interpreted, will dramatically improve our capability to realize a high performance health system.
