Here Comes the Stimulus Money, Now Spend It Wisely

Now that the $787 billion American Recovery and Reinvestment Act has been signed into law and billions of dollars are about to be funneled into HIT, the promise of improved care through technology has become something of a mantra: If you implement an electronic medical record, you will save money and more of your patients will survive.

Seems like a fairly simple equation, and a recent study from UT Southwestern certainly bolsters that notion, concluding that hospitals that use EMRs, CPOE, and clinical decision support systems saw a 15% decrease in the odds of in-hospital deaths. But one of the study's lead researchers warns that simply acquiring and installing these systems won't be enough.

The study compared 41 urban hospitals in Texas using an instrument created by the researchers that measures physicians' interactions with information systems. The researchers examined the rates of inpatient death, complications, costs, and length of stay for 167,233 patients older than 50 who were admitted to the hospitals for a variety of conditions during the same time frame in 2005 and 2006.

What they found was that, for all of the medical conditions studied, increased automation of notes and records was associated with a 15% decrease in the odds of in-hospital death. At hospitals with higher order entry scores, those patients with myocardial infarction had 9% lower odds of death and those undergoing coronary artery bypass graft had 55% lower odds of death.

Though Ruben Amarasingham, MD, assistant professor of internal medicine at UT Southwestern and associate chief of medicine at Parkland Health & Hospital System in Dallas, says he was encouraged by the study's outcomes, he cautions that hospitals should take the time to come up with a strategic, collaborative deployment plan before rushing out to spend their share of the stimulus money.

I recently had a chance to speak with Amarasingham, who shared his thoughts on the stimulus plan, the UT Southwestern study, and what's been holding up EMR adoption.

Kathryn Mackenzie: What effect do you think the $19 billion in the stimulus bill dedicated to healthcare information technology will have on EMR adoption?

Ruben Amarasingham, MD: It's hard to know how much money is needed. There have been some studies that have modeled this theoretically that would suggest it might be between $150 to $200 billion to really automate the entire country from head to toe. If that were the case, this would be at the minimum a down payment. These systems take a long time to implement and implementation has to be done very thoughtfully. One concern I would have is that hospitals that are somewhat naïve to the process would rapidly try to implement these systems and they may not have some of these cultural aspects worked out. Overall, I think it is the right move and is a very wise investment, but I would be concerned about the unintended consequences of doing it too rapidly and not taking the appropriate non-technological steps such as building the right culture and socio-technical environment.

I would also caution that hospitals need to carefully vet the various information technology products within this multi-billion dollar industry. There are some wonderful products out there, but there are also products that are sub-optimal. Smaller systems especially will have to have a good vetting process for that.

KM: What is a socio-technical environment?

RA: One of things we talk about in the paper is an emerging view of the clinical workplace called the socio-technical environment. Basically it's the idea that you can't divorce the people in the environment and the processes and culture and organizational workflows in the environment from the technology and vice versa. If you just put a technology in place and don't consider all of these other factors, then in most cases our feeling is that the technology won't succeed.

KM: Do you think that lack of recognition is one of the things holding up adoption?

RA: There are a lot of things factored into that. One, it's extraordinarily expensive to do. In big hospital systems it could approach $75 to $100 million just for implementation. That is not including ongoing maintenance costs or the hidden cost of trying to engage your staff, which is vital. Also, in the U.S. our overall experience implementing these systems has not been extensive. A lot of the big studies showing benefits from information technology systems occurred at major academic hospitals that have been working on those systems for 30 years. That has left some of the smaller hospitals wondering if they can really purchase a commercial system and get the same results.

Our results would suggest they could. We looked at a lot of hospital types for this study, many were not academic, and we found if they were able to meet the criteria I talked about before, the patients in those hospitals had improved odds of some of those outcomes.

KM: What was the purpose of the study?

RA: Usually the studies that have tried to look at outcomes have either looked at just a single institution or they've looked at multiple institutions, but simply asked the hospital whether they had an electronic medical record or not. There are a lot of important components that are missed in a blunt analysis like that. So we spent three or four years developing a physician-based instrument that would assess how effective an information technology system was and then we used that to address the questions we were interested in looking at.

KM: What questions were you trying to answer and how did the instrument you created help with that?

RA: It's called the Clinical Information Technology Assessment Tool. We designed an instrument that asked physicians for the routine activities that they do on a daily basis, like ordering tests, for example. Then we determined if that process is available to do electronically. If it is, does physician or nurse know how to activate that electronic process? And third, even if it's available electronically and they know how to activate it, do they choose to do that over other processes? By creating those three criteria, I think that we set the bar very high for the usability, the maturity, as well as the functionality of an information system.

Just spending money and getting a product is not the same as achieving a high score on this instrument. For all the activities we assessed at each of the hospitals we studied, we found a very strong association between high performance on this instrument and clinical outcomes. So the big distinction is that we not only measured the presence of information technology, we measured its use and functionality.

KM: What do you hope people will take away from the results of this study?

RA: These results are very promising, but I think additional studies are needed. Certainly, a variety of perspectives are needed as we enter this $19 billion stimulus phase, in terms of specifically measuring outcomes. I would encourage the government as well as others to really carefully think about how we would measure how well systems are doing and, particularly, how would we translate success stories from certain institutions to others that may be struggling. We are at a wonderful point right now in American healthcare to be able to see information technology come in. I think we need to exercise respect for potentially how difficult it is to do and how much caution we should exercise in implementing this.

The notion that a healthcare system is a combination of technologies and its people is not new, but it is often overlooked. If CIOs don't take the extra steps to make sure the hospital staff is not divorced from the technology, there is the potential to spend a great deal of money without changing a thing.