The article titled “High Rates of Adverse Drug Events in a Highly Computerized Hospital” in the May 23, 2005, issue of the ARCHIVES adds more fuel to a growing skepticism within the medical community's top journals about the value of medical automation technology.1
But in the sea change we are experiencing in clinical information systems, the risks and benefits of clinical information systems must be evaluated—not just the risks. Also, the issues surrounding quality of care and computerization are much larger than adverse drug events. Computerized physician order entry can vastly improve the quality of care only when it is part of an overall effort by a hospital to improve quality, and that involves tailoring the software to each hospital's specific needs and substantially investing in training of physicians, nurses, pharmacists, therapists, and other personnel in how to use the system. To assiduously seek and report on the failings of an automated system without also reporting the problems averted is misleading.