November 12, 2024 | Adam Rothschild
A patient recently asked me, his family physician, what follow-up testing was needed from his CT scan that was performed several weeks prior to screen for lung cancer. This was a curious question because I had previously followed our usual office procedure: I annotated and forwarded the CT scan report to a special account that was responsible for tracking and following up on results from screening studies. I double checked the report: Lung-RADS 2, a benign finding, accompanied by the standard radiologist recommendation to continue annual screening.
When I initially forwarded the report, I assumed my job was done. Upon re-review, I found that I overlooked the incidental finding of a thyroid nodule for which the radiologist recommended a follow-up ultrasound. As I had ordered the CT scan solely to screen for lung cancer, this piece of information is all that I had initially sought out in the report. The incidental finding and recommended follow up went unnoticed in a long page of text containing otherwise low-value information.
I mention this anecdote because it supports my belief that the use of text-based reports in electronic health records (EHRs) is another example of paving the cow path. As I described in a previous post, “paving the cow path” is the act of computerizing a previously manual business process without regard for whether or not the process is effective or efficient in solving the problem at hand. The CT scan report that I read was delivered to my EHR digitally, but it might as well have been on paper. The EHR software that I used to review the report did nothing special with the important information that could have emphasized incidental findings or nudged me to order follow-up studies.
Rather than just deliver a text-based document, it is reasonable to believe that my EHR, or some provider-facing EHR-adjacent software, could have identified the important incidental finding and/or recommendation for a follow-up study, which would have nudged me instead of relying on my fallible brain. Let’s recognize the value of software that solves problems and doesn’t simply pave the cow path by digitizing paper.
Adam Rothschild is a clinical informaticist at Solventum.
Get blog and podcast insights from our Solventum experts at the intersection of data science and healthcare.