The use of technology to minimize preventable medical errors was a central recommendation of the famous 1999 Institute of Medicine report, To Err is Human: Building a Safer Health System. One application with immediate implications for safer medical practice, computerized physician order entry, has met with strong resistance from its key constituency, doctors.
At first glance, it may be hard to see why. Inferential databases (those with relational "thinking") are used in almost every other type of business from automotive repair to travel agencies to track whether orders are correct and don't conflict with other orders or established procedures. Needless to say, the consequence of an error in a medical order can be far more severe than that of an error in an airline ticket purchase, so it is especially important for healthcare to use these kinds of databases.
With CPOE systems in place, there is very little room for ambiguity for the nurse who carries out the order or for the pharmacist who fills the prescription. Drug interactions and allergy alerts are automatically screened as the orders are written. Drug dosages outside the normal ranges are questioned with "pop-up" dialogue boxes on the computer screen before the physician electronically signs the order. And with CPOE, pharmacists can review and fill the prescription almost instantaneously. So the implementation of CPOE appears to be a no-brainer for the practice of medicine.
"No-brainer" takes on a completely different meaning when discussing physicians and CPOE, however. Physicians state that the use of inferential databases that cross-check and correct all aspects of their work leads to a "dumbing down" of the profession. Dominic Reilly, an attending physician at the University of Washington Medical Center in Seattle, says he fears residents will lose their ability to think critically when they are provided with computerized order-sets and cross-checked medication orders. "When these residents complete their programs and move to settings without computerized order sets, they will be at a disadvantage," Reilly says.
Not only does the CPOE system challenge the intellectual power of physicians; it then forces the physician to wait for the computer itself to "think" through the algorithms. The CPOE system implemented in Cedars-Sinai Medical Center in Los Angeles was reported to have eight-second delays between the three screens required to order a simple antibiotic for one patient. That system was shut down in January of 2003 as a result of physician complaints. There are reports of at least six other hospitals that have pulled CPOE systems as a result of physician resistance.
The medical culture is subtly threatened by CPOE in other ways as well. With CPOE, medical orders are clearly printed so that there can be little misunderstanding between the intent of the order and what was actually written. Anecdotally, some physicians have been heard to say that they prefer to write an ambiguous handwritten order or progress note, because they weren't sure of the spelling of a word, or to protect themselves should the chart be pulled for litigation. The ability to call into question what was actually written versus what was carried out lends the writer of the order some protection.
With its intellectual dampening effect, nearly 30-second delays between orders and unambiguous nature of typewritten text, it is no wonder CPOE is not widely endorsed by physicians forced to work with it. That is not to say however that physicians are not involved in implementing these CPOE systems -- they are.
Despite his concerns about the impact of CPOE on physicians in the learning environment, Reilly is one of the lead physicians in the On-line Record of Clinical Activity project being carried out at the University of Washington Medical Center, Harborview Medical Center and the Seattle Cancer Care Alliance. His clinical expertise and technical knowledge are essential to the custom design of the $40 million project.
Of course CPOE is not the first innovation to create such resistance in the healthcare setting. Hand washing's acceptance as a standard of practice in the medical community is a prime analogy. Ignac Semmelweis, a Hungarian obstetrician in the mid-19th century, instituted hand washing with chlorinated lime solution to prevent the spread of infectious disease in his hospital. Semmelweis' recommendation was not widely accepted by the medical community, in part because it forced physicians to admit that they were responsible for the spread of disease and the act of hand washing itself took time between each patient encounter. Fortunately, clinicians now have faster and safer ways to cleanse their hands between patient encounters. Now it takes 10 seconds with alcohol based cleansers to kill the bacteria on the hands as clinicians move between patients.
As has hand washing, the function of CPOE will improve with time and innovation. Physician resistance to CPOE as it currently exists will ultimately serve as a catalyst to improve it. However, the efficacy of CPOE as a modality for preventing medical errors will need to be assessed as these systems continue to be put into place in the healthcare setting.
Kyle Sisco is assistant nurse manager of the cardiothoracic intensive-care unit at the University of Washington, Seattle.