High error rate creates new urgency for CPOE

High error rate creates new urgency for CPOE

July 2008

Published in the July 2008 issue of Today’s Hospitalist

WHEN A REPORT released earlier this year found that one in every 10 patients at six community hospitals in Massachusetts suffered an adverse drug event, computerized physician order entry (CPOE) technology finally received some of the attention it deserves.

The report found, for example, that CPOE systems could have prevented as many 70% of the events tracked in the study. Perhaps just as importantly, the report also dispelled some of the myths and misconceptions that have hounded the technology for years.

Take the all-too-common view that CPOE is simply too expensive for hospitals, particularly smaller, community-based facilities. The report found that the technology is not only affordable, but that it typically pays for itself in 26 months “and saves $2.7 million a year thereafter.

Wendy Everett, ScD, president of the New England Healthcare Institute, which co-sponsored the “Saving Lives, Saving Money” report, says that hospitalists have long championed CPOE at their hospitals. But many others in health care “including physicians who would be end users of CPOE systems “have a long way to go to embrace the technology.

Dr. Everett is optimistic that the data in her organization’s report, along with a number of nascent trends, may help change that. Payers and state legislators, for example, are finding new ways to steer hospitals and physicians toward CPOE.

In her home state of Massachusetts, for instance, health plans and the legislature are creating both positive and negative incentives for hospitals to put CPOE in place by 2012. Dr. Everett talked to Today’s Hospitalist about the study results and what they mean for hospitalists.

Were you surprised by the high rate of errors?
We had anticipated that the medication error rate in community hospitals would be at worst the same as that in a teaching hospital with a very complicated staffing pattern. I imagined that having a simpler decision-making process would decrease the potential for errors.

But after a team headed by Dr. David Bates of Brigham and Women’s Hospital abstracted 4,200 charts, it turned out that the rate was close to double. That was both stunning and heartbreaking.

You included only events with “serious outcomes.” Can you give an example?
The results were very conservative because we didn’t count rashes or minor reactions that could have been prevented but were only a discomfort to patients. For a serious outcome, think of an elderly patient whose kidney function is compromised.

A good CPOE decision support system would contain lab data about creatinine levels, so if you order a dose of a drug that shouldn’t be given to this patient, the system will say that it is contraindicated. Without CPOE, physicians have to page through the chart for the last creatinine level, which may or may not be there, do the calculations, and look up the issues with that drug. As our report found out, that just doesn’t happen.

Physicians are often reluctant to learn a new computer system. What are the major holdups?
An enormous number of physicians still see CPOE as an imposition. It does take more time to enter data, and there is a learning curve. And data entry can never be done as quickly as yelling down the hall to a nurse. But it’s an issue of patient safety if the order is illegible and therefore filled incorrectly. Yes, it takes a while to learn, yes it decreases some efficiency, but not for long and not forever.

But isn’t CPOE too expensive for all hospitals to adopt?
That’s a myth, because we didn’t really understand how much CPOE costs. As part of this study, we calculated that to purchase, install, train staff and implement a program would cost a hospital $2.1 million, with annual operating costs of $435,914. But a hospital can save $2.7 million each year, and payers could save $900,000 per year per hospital.

For the first time, we were able to determine not just how much money we could save, but to whom those savings would accrue. In the greater scheme of things, spending $2.1 million to save $2.7 million a year is a bit of a no-brainer.

Your study also says that CPOE saves costs in other areas. What are they?
Hospitals can save $154,800 annually by moving to less expensive drugs; $47,900 on average per hospital by changing from IV to oral drugs; and $11,800 by eliminating redundant lab tests. CPOE can help achieve these savings.

The test repeated most often was a white blood cell count done manually. Because the marginal costs were so low and the test doesn’t cause harm, it didn’t pop to the top as a big issue.

Hospitalists can get involved by building a formulary or putting suggestions into a clinical support system. For example, the doctor may not know when a patient goes from a liquid diet to a solid one. But you can make a link in the system that issues an alert saying, “I notice that this patient has eaten. Wouldn’t you like to stop the $1,000 per day IV and give him a $3 pill?”

Do hospitalists have a unique advantage in implementing CPOE systems?
We tracked whether the hospitalist movement affected adoption of CPOE, and we’re looking at those data now. Hospitalists may not have the same cultural barriers as community physicians. It could be that physicians coming out of training now started using computers in high school, if not grammar school, and they see those tools as central to clinical care. But it may be more about willingness to use evidence-based guidelines.

Is Massachusetts ahead of the curve on CPOE?
Massachusetts may be a little more tech-savvy because of the concentration of academic medical centers with sophisticated computer systems and young folks coming out of training programs. I don’t think that the state is very different from the rest of the country.

Still, Blue Cross/Blue Shield here has said if hospitals don’t have CPOE by 2012, they won’t be eligible to participate in a 10% pay-for-performance initiative. Other private health plans are following suit.

Although only 10 of the state’s 73 hospitals had fully implemented CPOE systems when the study came out, almost all have said they are going to move forward and meet this deadline. Also, there’s language in proposed state legislation that says that hospitals that don’t have CPOE may not be eligible for continued licensure. That’s not an outcome we had anticipated.

Are there options for hospitals that just don’t want to or can’t put out the money for CPOE
You don’t need the Rolls Royce of CPOE systems, but you do need the Volkswagen to make sure you have clinical decision support that is integrated into lab and pharmacy systems. Almost every hospital in the country can find a way to borrow the $2 million.

It’s a question of how important this is to the individual institution. In the future, if I’m a smart hospital CEO, I’m looking at the movement to deny payment for “never” events. Medication errors are not currently in that category, but it won’t be long before they will be.

Paula S. Katz is a freelance health care writer based in Vernon Hills, Ill.