In Reply: Dr Wilkinson makes 2 points: that our emphasis on explicit algorithms warrants refinement, and that we should not dismiss the importance of subjective input. We agree that health care delivery is tremendously complex, and as a result physicians often find themselves in uncharted waters. Indeed, even when prior evidence exists and has been incorporated into an explicit treatment algorithm, it is still necessary to decide whether the algorithm truly applies to the individual patient. We therefore view all algorithms as tools that transform past information into predictions.1- 3 Since any attempt to predict the future is flawed, we agree with Wilkinson that all algorithms require continual analysis and refinement.
Duncan JR, Evens RG. Human Information Processing, Health Information Technology, and Medical Outcomes—Reply. JAMA. 2009;302(13):1417-1418. doi:10.1001/jama.2009.1418