Radiology: E-messaging system accurately shares radiology findings
In July 2006, Beth Israel began using an electronic messaging system to communicate important, but not emergent, radiology results to referring physicians. Radiologists input into the web-based system level 3 (or yellow) abnormal findings, which indicate conditions such as a lung nodule or solid renal mass that could result in considerable morbidity if they are not appropriately treated, but are not immediately life-threatening.
The input process took approximately one minute, and information was transmitted to two communication facilitators, who would report the abnormal findings to the referring physician via his preferred mode of contact (email, phone, fax or pager). All findings communicated by email required both an automated confirmation that the email had been read and reply from the referring physician; faxes were followed up by telephone. No other costs were associated with the system aside from the two full-time facilitators.
A total of 10,510 abnormal findings were communicated through the system between July 2006 and June 2009. Ronald L. Eisenberg, MD, and co-authors from Beth Israel assessed 500 of the communications, selecting every 20th communication for evaluation (to avoid bias with experienced use of the system). Eisenberg and colleagues assessed the rate of successful communication of findings, the frequency with which the findings were communicated within the 48-hour institutional goal and the results of a satisfaction survey regarding the electronic system.
All 500 findings were successfully communicated to referring physicians, with 82 percent of the results communicated within the hospital's 48-hour goal. Of the 89 communications that took more than 48 hours, 73 percent were submitted between 3 p.m. on Thursday through Sunday.
Sixty-five percent of abnormal results were sent to referring physicians by email, while 34 percent were communicated by phone.
Surveys were sent to all 438 of the referring physicians who had utilized the communication system, to which 119 physicians responded. Overall, 94 percent of physicians expressed satisfaction with the system. Five percent of physicians responded that they were dissatisfied with the system. The authors pointed out that of these six dissatisfied physicians, five had used the system on fewer than five occasions.
Abdominal, musculoskeletal and thoracic imaging sections accounted for just over 89 percent of the electronic submissions. Sixty-two percent of these transmissions communicated findings from either conventional radiographs or CT scans, with 60 percent of radiologists in these cases reporting intermediate findings with unclear clinical importance.
A note of caution
Radiologists recommended additional imaging in 65 percent of the 500 abnormal findings submissions. These recommendations met with some disdain from referring physicians, several of whom objected to being told which examination to order, while others reported that following radiologists' recommendations at times led to studies of questionable clinical value, according to the authors.
Several referring physicians responded that, because patients had access to the reports, they "feel that they are not getting thoroughly evaluated without further testing. It makes us clinicians nervous from a legal standpoint, and the wording should be carefully considered." Eisenberg and colleagues were only able to track physician compliance with radiology recommendations in 50 percent of patients' electronic medical records.
"[F]ailure to appropriately communicate important abnormalities to referring physicians in a timely manner can adversely affect patient care. It is a common cause of medical errors and adverse events in the outpatient setting and a relatively common cause of malpractice liability claims," the authors explained. "In any communication, the messaging system was associated with a 100 percent communication rate and spared radiologists from the frustration of wasting substantial time trying to contact the referring physician and sometimes even being unable to do so."
The authors reported a weakness to their study in not performing a controlled assessment of the previous phone communication system against which to evaluate the electronic system. Eisenberg and colleagues also considered self-selection bias in their survey's 27 percent response rate as well as the inability of physicians to respond with specificity about the system's usefulness and their satisfaction with it as potential limitations to their study.
"Nevertheless," the authors assessed, "the overall results appear to indicate general satisfaction among a substantial majority of respondents, especially those with the most familiarity with the electronic messaging system ... [T]he electronic messaging system communicated important, but not emergent, abnormal radiology results to referring physicians in a timely, accurate, and relatively inexpensive manner."