The perils of CPOE

Abstract

Many commentators1Kaushal R Shojania KG Bates DW Effects of computerized physician order entry and clinical decision support systems on medication safety: a systematic review.Arch Intern Med. 2003; 163: 1409-1416Crossref PubMed Scopus (1070) Google Scholar, 2Kuperman GJ Gibson RF Computer physician order entry: benefits, costs, and issues.Ann Intern Med. 2003; 139: 31-39Crossref PubMed Scopus (343) Google Scholar are touting the use of computerised physician order entry (CPOE), whereby doctors enter prescription orders directly into a computer instead of handwriting them, as a solution to the “large” number of medication errors in hospitals that put patients at risk of serious injury. The gravity of medical errors prompted the California legislature to enact legislation requiring all hospitals to have an error-prevention programme in place by Jan 1, 2005. I have been a practising internist-endocrinologist for more than 30 years, with hospital privileges at three different institutions—a major non-profit teaching hospital and two small for-profit community hospitals. I have also been a hospital surveyor as part of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) for more than 15 years. I am a certified utilisation review specialist and have been a consultant to the JCAHO on medical audit and quality improvement and assurance for many years. Extensive experience affords me a different perspective on medication errors—a perspective that tells me that CPOE will probably exacerbate the problem, not correct it. Experience has shown that most hospital medication errors occur after the ordering pen hits the paper and after entry into the “system”. Having the second worst handwriting on a medical staff of 1700 physicians, I learned long ago that it was essential that I read my orders to the nurse or clerk responsible for system entry to avoid midnight phone calls to explain what I had written. In my years of medical practice, all major (and fatal) medical injuries have occurred at the interface of the hands-on caregiver and the patient. These errors are now so commonplace that I insist that all my inpatients, with few exceptions, have a companion in the room at all times. Such a precaution prevents the one major and half dozen minor errors per day per patient. Between the order page and the arrival of the medication on the nursing floor, there is one failsafe procedure in all three hospitals with which I am affiliated and in most of the hospitals I have surveyed. All medications are dispensed from the pharmacy under the supervision of a licensed pharmacist and staff who have instant access to several computer programs with alerts and contraindications as well as a complete list of all current and past medications for each patient. Information on dosage, drug interactions, timing, allergies, and previous reactions are at their fingertips. If there is any conflict, the medication is not dispensed until the ordering physician has been contacted. If the physician is unavailable, and there is an urgent need for the medication, the pharmacist makes the correction or substitution, with a note on the chart requiring the physician's signature. With such a system in place there is no need for CPOE. The pharmacist's salary is far less than the cost of CPOE, which usually runs into millions of dollars. I appreciate the need to put safeguards in place to prevent these errors, and I believe there are other options more effective and less expensive than CPOE. These include: the bar-coding and scanning of patients' identification bracelets, medications, treatments, and procedures; the requirement that all personnel with responsibilities for any aspect of patients' care should pass an English language competency test to include medical terminology; hands-on, step-by-step investigation of the system that allowed the error and initiation of appropriate safeguards immediately after the fact; provision of all involved in or affected by an error with a notice and corrections; and for teaching pro-grammes, constant vigilance by the supervising physician or algorithms available continually and in real time with the interns, residents, and fellows. Errors that occur in teaching programmes will not be corrected by CPOE. In conclusion, I find that CPOE actually slows down the process of patients' care management and inhibits the patient-physician interaction. Writing orders while sitting at the bedside affords more time for personal contact, allowing me to ask more questions and my patient to offer additional information. Taking time away from my patient to address a computer is not the best medical care.