Entering Misleading Information to “Get Scans Faster” Puts Patients at Risk

Research at a large tertiary referral hospital shows that a high level of electronic radiology requests contained incorrect or misleading information


Twomey
Twomey

Research conducted in Ireland indicated that up to 45 percent of electronic radiology requests contained incorrect or misleading information about patients' biochemical or hematological status.

“The high level of erroneous clinical and laboratory information is concerning,” said RSNA 2014 presenter Maria Twomey, M.B.Ch.B., a fellow of the Royal College of Surgeons in Ireland. “The primary concern is the effect this can and does have on how the radiologist protocols, prioritizes the study and reports the study.”

Dr. Twomey said she and her team saw the need for a formal study when they heard anecdotal evidence of incorrect clinical information on requests. “There was some suspicion that colleagues were entering erroneous information to get scans faster,” she said.

The study was performed at Cork University Hospital, a large tertiary referral hospital that receives requests via an electronic radiology information system. Information submitted by the referring physician was compared to the reported levels on their own institution's biochemical and hematology reporting system.

Of the 250 requests included in the study, researchers found that up to 45 percent contained erroneous information about creatinine, hemoglobin, white cell count and C-reactive protein levels. Fifteen percent of requests for CT pulmonary angiography, for example, reported an abnormal D-dimer result when the actual reported result was normal. Twenty-five percent had reported hypoxia when the lab-reported blood oxygen level was normal. Elevated C-reactive protein and/or white blood cell count was reported in 70 percent of acute abdominopelvic CT requests, but 20 percent of the formal results in that subgroup were normal.

Daily Bulletin coverage of RSNA 2014 is available at RSNA.org/bulletin.

“Significantly higher incidences of erroneous parameters were supplied by medical physician referrals than by surgeons,” Dr. Twomey said.

Study Request Errors Can Lead to Misinterpretation of Results

Errors in an imaging request can result in selecting an inappropriate imaging procedure, Dr. Twomey said. “It may lead to the incorrect modality or study protocol being performed or inappropriate radiation dose. It could cause a delay in other patients being scanned and may lead to misinterpretation of radiological findings.

“Ideally, all biochemical and hematological results would be checked with the laboratory system,” she said. Acknowledging that this could be laborious, Dr. Twomey's team instead recommends the implementation of an electronic ordering system linked directly to patients' laboratory reports and, ideally, the electronic patient chart. “The software exists and is in use, but it is not available in our and many other institutions," Dr. Twomey said. “Budgetary constraints are prevalent throughout radiology; however, these findings would support capital input into this software.”

Accurate clinical information is essential for radiologists to make informed judgments on patient exposure to radiation, Dr. Twomey emphasized. “Hopefully a software system connected to the lab results would make our colleagues think twice when they make that request,” she said. “Solutions based on data are the most effective. Don't do a D-dimer if you're going to ignore the results.”

Evonne Acevedo Johnson is a Goliad, Texas-based freelance writer specializing in healthcare issues.