Diagnosing the adoption of IT to make health care healthier

September 24, 2008

Last year, Dennis Quaid's anguished visage was splashed across the tabloids. Like many Hollywood stories, this one revolved around drugs. But it wasn't the usual A-list overdose or contraband possession. Instead, Quaid's newborn twins were mistakenly given two doses of Heparin rather than Hep-lock. The dosage was a thousand times stronger than it should have been, nearly killing the babies.

For anyone who's ever had a doctor handwrite a prescription and marveled at its illegibility, the issues that Quaid faced and which Michael F. Furukawa, an assistant professor in the School of Health Management and Policy at the W. P. Carey School of Business, addresses seem self-evident. But the problem is not just a pharmacist dispensing an incorrect drug or a nurse administering too much of the right one. The health system abounds with medical mistakes. Some have minor effects; some are fatal. Together, they add up to a very real problem.

When the report, "To Err Is Human," was released a decade ago by the Institute of Medicine, it revealed a shocking calculation: each year, 44,000-98,000 American patients die as the result of hospital errors. Those mistakes cost $17-29 billion according to the report. The Leapfrog Group, a coalition of some of the nation's largest employers, envisions cutting errors and saving $41.5 billion. Those costs eventually affect patients, hospitals, insurers, corporations and the government. In 2000, the Medical Error Reduction Act was introduced and four years later President George W. Bush appointed the nation's first National Coordinator for Health Information Technology. The coordinator was charged with fulfilling an Executive Order calling for the widespread deployment of health information technology to improve safety and efficiency.

"There was an expectation -- a hope -- that hospitals would jump on the bandwagon and immediately and rapidly adopt these technologies," says Furukawa. "It didn't happen. That's where my study comes in." Other countries such as England and Australia are much further along than the United States.

With so much at stake, it is no surprise that there is still a lot of talk in boardrooms and in stump speeches about improving patient safety and controlling costs. Yet, despite the interest, Furukawa discovered that no one really had the most basic facts about the use and impact of the most popular technologies meant to improve patient safety.

Technologies and errors

While a small number of previous studies have looked at health IT adoption, Furukawa's is the first to focus on IT for medication safety. He looked at eight basic IT implementations commonly discussed as answers to medical safety problems. They include electronic medical records and using bar-codes on patients' bracelets and on individual medications to ensure patients are getting the proper medication. Using data from a comprehensive, national survey, he sought to discover to what extent American hospitals had adopted health IT for the year 2006. His results, co-authored by fellow W. P. Carey School researchers Raghu Santanam, Trent J. Spaulding, and Ajay Vinze, were published in the May/June 2008 issue of Health Affairs. He found that hospitals varied widely in their adoption of the technologies depending on their organizational and financial characteristics, geographic location and regulatory factors.

Among the more challenging aspects of Furukawa's research was correlating adoption of health IT to actual reductions in hospital errors. This is, however, not a problem unique to his research. Hospitals generally do not track their medical error rates and don't have data to suggest what errors even cost. From the hospitals' perspective it's understandable because it means admitting that errors happen when patients' lives are on the line. That's bad for business. Unfortunately, that same closely guarded number is extremely important to researchers and, eventually, to policymakers. The rate allows hospitals to benchmark their own progress and, on a macro scale, it enables studies to determine if these new technologies actually work.

The federal government does not require hospitals to track and report medical errors, however, some states do have mandatory reporting systems. These include California, Massachusetts, New York and Texas.

Waiting for the tests; taking cultures

One of the biggest hindrances preventing the adoption of medical safety technologies is that there is no smoking gun proving that these systems actually improve care and reduce costs. Even if there was a mountain of evidence, there are two major hurdles to adoption: implementation cost and cultural resistance.

All of the proposed technologies can induce what doctors might diagnose as sticker shock. Electronic medical records, clinical decision support, and computerized physician order entry all require substantial hardware, software and IT integration. Bar-coding and medication dispensing machines require specialized hardware. The million-dollar price tag to install a special robot to fill prescriptions may give administrators headaches and chills. And all this at a time when hospitals are already feeling pinched financially.

"Managers have limited resources. They can either remodel their emergency room or they can buy these computer systems," says Furukawa, who underscores that many of today's hospitals were built as a result of the Hill-Burton Act of 1946 so many are now spending money to simply keep the buildings standing.

Even without these concerns, hospitals are still reticent to invest in immature technology. In this way, a 500-bed institution is much like an individual consumer: We know that new technology -- whether a CD player or an enterprise health application -- begins with a high price tag and limited features. After months or years, it will likely be substantially better and cheaper.

Money aside, the technology faces what may be the most important hurdle: culture. In hospitals, as with any organization -- especially large ones -- culture change does not come easily. And as one might expect in a life-and-death business with lots of stakeholders and a conservative, show-me-the-research mindset, large changes are especially difficult.

Unequal care

Knowing why hospitals haven't jumped to implement IT projects for medication safety, it's somewhat easier to understand Furukawa's results. The overall adoption was quite low: an average of 2.24 of 8 applications per hospital. Specific findings showed that "hospitals in rural locations have about one-third the adoption rates of those in metropolitan areas, and hospitals with Joint Commission accreditation have three times higher health IT adoption than their unaccredited counterparts." Further, large hospitals (200 beds or more) have three to four times greater adoption rates than those of small hospitals (1–49 beds).

The research also showed that East Coast hospitals have higher health IT adoption rates than those in the Western and Mountain regions. The leading states (Rhode Island, DC, Delaware, Maryland, New Jersey, Massachusetts and Connecticut) had an average of three or more applications per hospital compared to less than one for South Dakota, Montana, North Dakota, Wyoming and Kansas. Somewhat ironically, for-profit hospitals counted fewer of the applications than non-profit hospitals. Furukawa says that while this may seem counterintuitive, the for-profit institutions are more likely to scrutinize the buying decision.

"It's limited capital for competing projects and when you do a traditional return on investment and you have trouble quantifying the actual financial benefit from the error reduction, then it's easy to say, 'Oh, I'll put it off,'" says Furukawa.

State and local government hospitals fared quite poorly, having only half the applications of private not-for-profit hospitals. In keeping with his findings, Furukawa writes that, "teaching hospitals and hospitals belonging to a multihospital system have about two times the rate of adoption that nonteaching and independent hospitals have."

Taking the Medicine

Perhaps because he has seen how error-prone prescriptions can be, Furukawa says his paper is not prescriptive but rather is meant to be a "snapshot" of the state of health safety IT. However, policymakers may draw a few conclusions.

First, hospitals need more evidence before they're willing to adopt IT for health safety. This requires studying whether these technologies actually reduce errors, improve medical care and decrease costs. This comprises Furukawa's forthcoming research effort. Better efforts to obtain medical error rates are needed in order to look at these facets effectively.

Patient safety centers, generally set up as information clearinghouses at the state level, also offer great promise. They help hospitals adopt the new technology by providing lists of appropriate IT vendors, coaching hospitals through RFPs and sharing best practices.

Furukawa's research is conclusive in showing that "hospitals in states with patient safety centers have the highest rates of adoption; the relative difference ranges from 38 percent to 118 percent higher probability of adoption." Similarly, hospitals in states with error reporting systems had 19-54 percent higher IT adoption rates than those in states without the initiatives -- although Furukawa points out that these relationships are not necessarily causal. Still, it's possible that if more states create reporting systems and set up safety centers, more hospitals in those states will adopt technologies with the promise of decreasing medical safety errors.

Finally, policymakers might also see the value in subsidizing hospitals' costs to adopt these new technologies. Furukawa says the system likely will improve more in the aggregate if funding is applied in a to-each-according-to-its-need manner, because there is such dire need for IT improvement in smaller, community hospitals. Even if these smaller facilities do nothing else -- hire more experienced practitioners, add staff, buy new medical equipment, etc. -- simply plucking this low-hanging fruit and fixing preventable mistakes could yield huge dividends.

"We should look at the hospitals that really need it first -- the ones that are really struggling -- and put more money there, rather than just roll out a uniform national program," says Furukawa.

Bottom Line:

  • Patient safety isn't just a concern for doctors and hospitals but affects everyone from individual patients to companies that face rising health care costs to insurers and the government.
  • A decade-old estimate figures that 44,000-98,000 American patients die each year as the result of hospitals' medical errors, with those mistakes costing $17-29 billion.
  • Since the introduction of the Medical Error Reduction Act in 2000, reducing medical mistakes has appeared to be a national priority but little progress has been made.
  • Adoption of IT systems for medication safety is slowed by hospitals' reticence to benchmark their error rates as well as a lack of data showing effective ROI when reducing errors.
  • Hospitals -- many of which were built in the 1950s -- have limited funds to spend on IT projects and much of their capital must go to maintaining the buildings. Also, there is a hesitation to invest in immature systems when there may be more bang for the buck in a few years.
  • There is significant variation in which hospitals adopt IT for medication safety. Generally bigger institutions in cities along the East Coast have the highest rates while those in rural areas in the West and Mountain regions have the lowest adoption. Non-profits generally have more applications than for-profit institutions.
  • Patient safety centers and error reporting systems may be an important ingredient in prompting hospitals to implement IT applications for medication safety.