In order for radiologists to generate the most value for their patients and referring providers, they need to be equipped with all the right tools during their workflow. That includes adequate context in the form of patient data.
However, that data often isn’t readily available to radiologists at the point of interpretation, forcing them to go tracking it down, or possibly just going without.
Cree M. Gaskin, MD, radiologist for the University of Virginia Health System in Charlottesville, says it’s a common problem to be left without important information, and some information could be reliably expected to be absent from an imaging order.
“No radiologist is going to be surprised that the referring physician didn’t indicate a 50-pack-year smoking history in the order because they don’t take the time to write it,” says Gaskin. “They’re busy and that data is in the electronic patient chart. Their view may be, ‘If you want the data, go to the medical record, don’t expect me to write it for you.’ And I would say that’s right.”
There are a number of other examples of commonly missing, but useful pieces of information. Gaskin also points to cases like an x-ray showing a joint effusion. A radiologist may assume that’s an injury, but it could also be caused by an infection, and so laboratory data like inflammatory markers would be valuable to have immediately at hand. A radiologist might not know what information is missing and thus won’t go looking.
How often do radiologists do without potentially useful patient data? A group including Piotr Obara, MD, and colleagues from the University of Chicago tried to answer that question in a 2014 paper published in the Journal of Digital Imaging. In a scan of 140 cases, a patient’s relevant chronic condition was not mentioned in the clinical indication for imaging exams nearly 60 percent of the time, with non-oncological conditions most commonly omitted.
“Our study demonstrates an alarming lack of communication of pertinent medical information to the radiologist, which may negatively impact interpretation quality,” wrote the authors. “Presenting automatically aggregated patient information to the radiologist may be a potential avenue for improving interpretation and adding value of the radiology department to the care chain.”
Adding patient data to the workflow
The question then becomes two-fold: what data do radiologists need to provide an accurate interpretation and how should that data be presented?
Gaskin says that radiologists would want different information based upon the type of study and clinical indication, but generally any anatomic information such as surgical history, and major medical conditions that could affect imaging are very important. Certain lab values would be useful in various situations, such as in the evaluation of a thyroid nodule. Family history suggesting potential familial conditions and social history reflecting tobacco, drug, and alcohol use can also be helpful.
Presentation of the data is nearly as important as deciding which data to include. Gaskin notes that radiologists read studies very quickly, in a matter of seconds for a basic chest x-ray, and so there’s not time for radiologists to log into the EMR and search a chart and stay equally productive. There are methods of streamlining this process, however.
“What’s most important is that radiologists have access to the patient data,” says Gaskin. “That’s the fundamental goal. Radiologists must have efficient access to relevant patient data that impacts their interpretation.”
For one method, Gaskin recently co-authored a paper on leveraging an EHR-driven workflow for the Journal of the American College of Radiology. This technique leverages a RIS-like module within an EHR to generate a work list, and then providing easy access to prescribed, boiled-down, relevant patient data from the medical record during image interpretation.
You could also take the same concept and use third-party software or workflow tools that can pull the same information, provided it’s correctly mapped to display the data a radiologist wants.
In the end, the goal is to minimize the time a radiologist spends searching for patient information so that they can concentrate their effort on reading and reporting on an image. No radiologist is going to say they don’t want this important information, but they may not realize the technology exists to provide it efficiently.
“Whether you do it through an EHR-driven workflow or another