A team led by Dr. Allen Grady, a radiology resident at the University of North Carolina, reviewed incident and accompanying radiology reports, and found that the reports were completed as suggested by the American College of Radiology (ACR) recommendations less than half the time.
But after electronic records were instituted and, later, new hospital tracking policies, the complete reporting of the extravasation incidents improved, Grady said in his presentation.
"Subcutaneous extravasation of intravenous contrast is a potentially serious complication in performing imaging studies," Grady said. "It is reported in the literature to occur from 0.03% to 0.9% of the time. There is a trend toward increasing frequency with greater use of automated mechanical bolus injectors."
Adverse events caused by extravasation include swelling, pain, erythema, compartment syndrome tissue necrosis, and skin ulceration, he said, while treatment can include elevating the extremity, warm or cold compression, and even surgery.
Grady and his colleagues conducted a retrospective study that looked for incidents of extravasation. They reviewed handwritten records in 2006 and then the electronic records that were kept after January 2007 through May 2011.
A complete report on an extravasation incident was considered to be one that relied on ACR recommendations: "All extravasation events and their treatment should be documented in the medical record, especially in the dictated imaging report of the obtained study, and the referring physician should be notified," states the ACR Manual on Contrast Media.
Grady and his team looked for five elements of a complete report:
If the report had fewer than all five elements, it was considered "incomplete." Failure to report extravasation in the record was listed as "not documented."
In the initial 31-month review, 60 events were reported out of 55,320 studies -- an extravasation rate of 0.11%. Of those incident reports, 28 (46%) were complete, 16 (27%) were incomplete, and 16 were not documented (27%). Of the corresponding radiology reports, the incident was complete 4% of the time, incomplete 28% of the time, and not reported 68% of the time.
Grady said reporting improved when electronic records were implemented. Complete reporting was measured at about 22% in the handwritten era, but it rose to 60% with electronic records. Reports with no documentation occurred 62% of the time in the handwritten period but dropped to 5% in the electronic era.
In 2009, the hospital created a new institutional policy that explicitly required that technologists call a radiologist if extravasation occurred; radiologists were required to examine all of these patients, and radiology reports were required to completely document the results.
In the 23-month follow-up, 73 events were reported out of 40,878 studies -- a rate of 0.18%. Following the intervention policy, the number of complete reports increased from 46% to 75%; incomplete reports fell from 28% to 19%, and not documented reports dropped from 27% to 4%, Grady said.
The corresponding radiology reports showed similar changes: complete reports increased from 4% to 27%, incomplete reports increased from 28% to 38%, and not documented reports dropped from 68% to 35%.
"Incident report quality improved with implementation of an electronic reporting system and policy and educational initiatives," Grady said. "Documentation of extravasation in radiology reports improved after policy and educational initiatives."
In response to questions from the audience, Grady said that follow-up is continuing with new residents and radiologists to maintain the improvements seen in reporting.