A prototype system developed by the University of Maryland was able, for a time at least, to dramatically boost and document the communication of critical results findings, according to a presentation Tuesday. It's since been sent back to the drawing board for more work, but points to a solution to a problem that vexes radiologists nationwide.

Documenting the communication of critical findings is required by the Joint Commission on Accreditation of Healthcare Organizations, but is always problematic for radiology departments. Checking with the physician who referred the patient for imaging frequently finds him or her gone, or no longer in charge of the patient, said Paul Nagy, Ph.D., who presented the paper. Often, trying to find the right person proves to be an exhausting process.

The university's radiology department tried a commercially available notification system that meets JCAHO standards, but it wasn't popular with radiologists or referring physicians, Nagy said. Radiologists, in particular, wanted to be sure that their hot finding was communicated to and understood by the physician in a position to act on it and the commercial system didn't give them that assurance.

"They couldn't stand it if they had a serious finding," Nagy said. "They needed to get hold of somebody."

Initially, the system was used in about 10 (out of 25,000) cases at the hospital per month but eventually declined to about two per month, Nagy said.

Looking for a better solution, the department developed its own system based on a web tool that mines the hospital's electronic medical record to learn where a patient is currently located, who has been working with him or her recently, who has been placing orders, and what service team is providing care. Names and contact information of identified clinicians are provided to the radiologist.

Physicians who might need to be told about findings may include those who are looking at the images and those who have been placing lab orders for the patient. Once the potential clinicians are identified, a blogging system records attempts by the radiologist to contact the ones who might need to know about the critical finding.

The university's radiology department implemented the system in April and in the first month it was used to communicate 90 critical findings. The clinical IT support team trained the chest and body sections first. In the first two weeks there were a total of 38 critical findings; 21 were delivered within the first 60 minutes.

"This was very exciting," Nagy said. "We opened our champagne bottles and wrote our abstract for the RSNA."

But a short time later use of the system declined dramatically. One big problem was that a new set of residents and fellows came in and the integrated contact database was not updated promptly, Nagy said. That probably reduced the numbers by 10% to 20%.

Another problem was data overload. Some patients are touched by 200 clinicians over the course of 24 hours and weeding through that list was too difficult for the radiologist, Nagy said. Another problem was that the radiologist would move on to other work while waiting for a callback and the system design made it difficult to return to log the result.

"We're back into the rebuild phase," Nagy said.