CPOE: not the first step toward patient safety

Health Management Technology, Jan, 2005 by Richard Kremsdorf

"To Err is Human," the landmark Institute of Medicine (IOM) report from 1999, successfully highlighted the enormous problem of patient safety. Unfortunately, since its publication, not much has been accomplished.

A recent analysis of adverse clinical events--the AHRQ Patient Safety Indicators--measured their incidence in the Medicare population and confirmed that a large number of people in the U.S. die from potentially preventable, in-hospital medical errors. The study found that "failure to rescue" accounted for a majority of these patient safety-related deaths.

Recent "failure to rescue" statistics not only are troubling, but they also show that whatever the industry has been doing to try to improve patient safety hasn't shown its benefit yet. Why have we made so little progress with patient safety? Simply put: We are focused on the wrong problems and are not giving the right tools to healthcare professionals with urgent needs.

Errors of Omission, Not Commission

Unfortunately, the original IOM study put us on the wrong track. The study was useful in highlighting systematic problems with the way healthcare is delivered. However, its implicit emphasis on medication safety led the industry to rush to implement computerized physician order entry (CPOE) and bar code medication administration tools as the solution.

Not only haven't CPOE and bar code medication administration tools solved the problem, but they have also become a distraction from solving the real systemic issues at the heart of "failure to rescue."

How did this happen? The IOM report prompted a focus on "errors of commission." These are the visible, often blatant mistakes usually associated with an individual caregiver, such as a physician writing a defective order for a drug or a nurse giving the wrong pill.

Yet, at the core of the patient safety problem are the often invisible, unrecognized mistakes that happen, called "errors of omission." Errors of omission occur when an appropriate step in a process doesn't happen or when a needed action is not performed. One example is suboptimal patient care resulting from the inadequate availability of information for decision-making. Another more blatant example is one in which a patient's vital signs indicate decompensation, but the clinical response is inadequate or delayed.

As a pulmonary and critical care practitioner for more than 20 years, I saw the latter happen time and time again. I was called in to save patients when they were falling apart. After looking at the chart, I could see that in the preceding four to six hours there had been several signs indicating decompensation. If I had been alerted earlier, I could have intervened in a much less invasive way, averting many complications and deaths. These defects in care are often rationalized away as though they were unavoidable, rather than systematic failures in the care process.

Walk Before You Run

CPOE will not solve the core patient safety problem; it is difficult and expensive to implement. More importantly, focusing on CPOE has prevented organizations from funding other efforts that would have a larger beneficial impact. We need a shift in focus in the types of electronic tools that are provided to address the "failure to rescue" problem.

First, hands-on caregivers need electronic tools to help them carry out their many tasks. This begins with the capture and review of point-of-care data that defines a patient's clinical condition.

Practitioners need to have access to important patient data in electronic form to support ongoing patient assessment. The result would be improved clinical decision-making and responsive reformulation of the plan of care. Right now, a clipboard with the patient's handwritten, paper flow sheet is still our primary data source for important information such as vital signs. Unfortunately, it is often hard to locate, illegible or incomplete.

Secondly, we must provide clinical surveillance tools to create an effective safety net. While CPOE focuses on creating a perfect order at the initiation of therapy, the patient's course is often not what was expected or another problem develops. While CPOE has its place, the industry has a greater short-term need for easier-to-implement tools to identify how the patient is responding to prescribed treatments.

Proactive Clinical Surveillance

We need clinical surveillance tools that caregivers and supervisors can use to monitor patients and identify which patients are at greatest risk.

The patient's nurse is on the front lines and is easily overwhelmed when the clinical condition of one or more assigned patients worsens. The charge nurse and a variety of patient care supervisors are supposed to provide backup and a safety net, but their methods for identifying such circumstances are limited by the available tools and are consequently fairly rudimentary: walk around and look for signs of trouble, ask caregivers if they need help, or respond to a page or phone call that says, "I'm in trouble."

The most important aspect of effective clinical surveillance is that information needs to be presented in a way that conveys the big picture. Users should be able to see a hospitalwide view and a unit view. At a glance, a supervisor should be able to see all patients on the floor, the number of alerts that each has and triggering values for each patient. This is actionable data showing important patient vital signs and care history that can lead to decisive, quick and informed clinical decisions.