They say the devil is in the details. That could be the case for computerized physician order entry, or CPOE. Great idea, great concept, but if improperly executed, potential trouble.
In December, stuff hit the fan when a group of researchers from the University of Pittsburgh School of Medicine and Children’s Hospital of Pittsburgh reported an “unexpected increase in mortality coincident with CPOE implementation.” (“Unexpected Increased Mortality After Implementation of a Commercially
Sold Computerized Physician Order Entry System,” Pediatrics Vol. 116, No. 6, December 2005, pp. 1506-1512.) In other words, kids appeared to have died at Children’s because of computer systems designed to ensure their safety. Who would have thought? After all, computerized order entry has been hailed as a panacea for many of healthcare’s ills.
CPOE allows physicians to order medications for their patients via computer terminals. Gone are the mistakes due to sloppy handwriting or a lack of information. Instead, these systems integrate doctors’ orders with patients’ electronic medical records, preserving all prescriptions. The best CPOE systems interface with clinical knowledge systems to help doctors assess whether the medications being prescribed are indeed the best ones for the patient.
Yes, order-entry systems can be expensive. The Leapfrog Group, a coalition of 170 major employers, says that adoption of CPOE among hospitals has been extremely slow due to the high cost of these systems- anywhere from $500,000 to $14 million, depending on the size of the hospital.
But the potential benefits are huge, according to some. The Agency for Healthcare Research and Quality (part of the Department of Health and Human Services) estimates that anywhere from 28 percent to 95 percent of adverse drug events can be prevented by reducing medication errors through computerized monitoring systems. What’s more, computerized medication order entry has the potential to prevent an estimated 84 percent of dose, frequency, and route errors, according to AHRQ.
These numbers are significant. More than 770,000 people are injured or die each year in hospitals from adverse drug events, which can cost up to $5.6 million each year per hospital, depending on hospital size, estimates AHRQ. Patients who experienced adverse drug events were hospitalized an average of eight to 12 days longer than patients who did not, and their hospitalization cost $16,000 to $24,000 more.
President Bush himself has made implementation of electronic medical records a priority. And CPOE is a vital part of the initiative. No wonder the Pittsburgh researchers’ study prompted controversy!
Specifically, the researchers collected clinical and mortality data on kids who were admitted to Children’s via interfacility transport during an 18-month period (13 months before implementation of CPOE and five months after it). Among 1,942 kids referred and admitted for specialized care during that period, 75 died, accounting for an overall mortality rate of 3.86 percent. Analysis revealed that mortality rates significantly increased from 2.8 percent (39 of 1,394) before CPOE implementation to 6.57 percent (36 of 548) after it.
To be sure, a number of questions have been raised about the validity of the study. For example, it turns out that at the same time Children’s was implementing CPOE in October 2002, it was also centralizing all medication into one pharmacy. That process by itself could have resulted in administrative mixups, according to detractors.
Other questions raised:
Is five months a long-enough period to conduct the post-CPOE portion of the study?
Should the researchers have studied 13 months of post-CPOE data to balance the pre-CPOE data?
Is it legitimate to study only those patients who were transported to the hospital for specialized care (which reportedly comprise fewer than 10 percent of all the patients admitted during the same time period)?
But it would be too convenient to blame everything on flawed research. For example, the researchers reported that immediately following implementation of the order entry system, doctors found it difficult to enter medication orders and tests; nurses were pulled away from the bedside to work on terminals, and system crashes froze terminal screens and delayed the delivery of some medications.
Was it the automated system itself that led to the childrens’ deaths, or was it the manner in which it was implemented? Indeed, one doctor writing to Pediatrics noted that some automated systems are “abysmal in design,” and are “slow, cumbersome and unintuitive.” If it is true, as the Children’s researchers reported, that order entry took one to two minutes, “this is simply unacceptable in a busy ICU setting,” he wrote.
On the other hand, perhaps Children’s Hospital was guilty of poor execution. “The speed of the implementation was unusual – it is not good practice to implement [computerized physician order entry] in seven days, as this does not leave time for mid- course corrections,” wrote a group of doctors and Ph.D.s in another letter to the journal.
But in that same letter, the authors make a key point when they write, “The real lesson from this study is that there can be unintended adverse effects if hospitals don’t carefully plan for and implement major clinical transformations like CPOE.”
It’s true that contracting professionals will not decide what order-entry system is brought into their IDNs. However, they have an important role to play. For example, they can help the decision-makers prepare thorough, well-thought-out RFPs; they can insist that the key decision-makers make site visits to current installations; and they can lend assistance when it comes to negotiating pricing and terms with the vendor of choice.
The Pittsburgh study – and all the fallout – show that even the noblest of causes (including something as popular as computerized physician order entry) can go down in flames because of inadequate planning and poor execution. Journal of Healthcare Contracting readers can play a role in making sure that doesn’t happen.
