- © 2007 Canadian Medical Association or its licensors
The account by Liette Lapointe and Suzanne Rivard of the challenges of implementing electronic clinical information systems in hospitals1 vividly reminded me of my own experience with such an undertaking 20 years ago.2 Their report indicates, not surprisingly, that the success of an implementation is determined by the implementers' ability to cope with user expectations and hospital politics. In all 3 of their cases, the introduction of the module for order entry appears to have been the crucial challenge to ultimate success and was the step most likely to provoke physician resistance.
This finding demands further examination. The explicitness inherent in order-entry systems provides a degree of routine transparency and implied accountability to which physicians are not accustomed, and it can provoke physician resistance. I believe that the key to successful implementation lies not in trying to overpower this resistance but rather in circumventing it by exploiting a feature that meets with universal favour. The “problem list” might be such a feature, judging by its inclusion in virtually all of the electronic clinical information systems currently on the market, but presently it is an incidental or optional by-product with little apparent utility. This does not fit with Weed's original concept,3 in which the problem list was to be the backbone of the patient record.
After more than 30 years of effort, we are still far from having a health care system that is free from the crippling inefficiencies of paper records. A radical rethinking of implementation strategies for electronic clinical information systems is urgently needed.