Skip to main content
Thank you Drs. Walkowiak for looking at the interrater variability in ICD code usage. Your findings verify what most of us intuitively have experienced but been unable to put data to: namely that ICD codes are a poor source of data for epidemiology, planning, surveillance, research and public health.
There is wide variability in the effort and therefore accuracy of these codes between physicians and patients. Physicians spend relatively little time hunting thru the enormous list of ICD codes to make sure they are as precise as possible. On the other hand, physicians do actually spend time on this task even though it isn't of meaningful value to the patients we are seeing.
It would seem in light of this excellent data that we should move past ICD codes and gather data from the EMR directly and that it be a standard function of EMRs to generate this data from notes, orders, and test results that could be accomplished without any wasted physician work and time. Let's use the information we have to improve the data and the lives of physicians.