How Much Is Too Much? Tracking Radiation Dose through an EMR
When word came out that more than 200 patients at Cedars-Sinai Medical Center in Los Angeles had been accidentally exposed to extraordinarily high doses of radiation during CT brain scans, it spotlighted an issue that has roiled radiology for several years now—how much radiation exposure is too much, what is as low as reasonably achievable (ALARA) and how can it be tracked?
It’s an issue that has made the jump from professional journals to the popular press, so in that sense, “the timing [of the Cedars’ experience] couldn’t have been better,” says Kevin McEnery, MD, a diagnostic radiologist at MD Anderson Cancer Center in Houston, who adds that his facility is seeing more and more patients making inquiries about their cumulative dose.
Patients are talking about the issue, and institutions are in the process of trying to deal with it, says McEnery, and incidents like the one at Cedars-Sinai “will only facilitate those efforts.” (Since the situation at Cedars-Sinai was discovered, the FDA also has been investigating reports of excess CT scan radiation cases at two other Los Angeles-area hospitals as well as hospitals in other states.)
One hospital that’s decided to try to tackle the issue head-on is Brigham and Women’s Hospital (BWH) in Boston.
In a study published in the April issue of Radiology, Aaron D. Sodickson, MD, PhD, assistant director of emergency radiology at BWH and a team of researchers from its Center for Evidence-Based Imaging (CEBI), found that patients who undergo numerous CT scans over their lifetime may be at increased risk for cancer.
The authors studied 31,462 adult patients who had diagnostic CT scans at BWH or at the Dana-Farber Cancer Center in 2007 and had undergone a total of 190,712 CT exams over the previous 22 years. They determined that 7 percent of the patients in the study had enough recurrent CT imaging to raise their cancer risk by 1 percent or more above baseline levels. They estimated cumulative radiation dose and associated radiation-induced cancer risks for individual patients by extracting each patient’s CT history from the EMR and applying standard risk-estimation models.
Now the informatics team at Center for Evidence-Based Imaging is using EMRs to develop a decision support tool that, according to Sodickson, would provide point-of-care radiation risk assessments of individual patients for providers ordering studies.
When a referring physician places an order for a CT scan in the computerized physician order entry (CPOE) system, he or she also can extract a real-time image history from the database for that patient, according to Sodickson. “Using that history and our best estimate about dose for each of those exams, we try to come up with a cumulative radiation dose and, along with some demographic information from the patient, try to back out to a best estimate of radiation risk to the patient.”
The informatics challenge for the BWH-CEBI team is that while all of this information is available in BWH’s radiology information system (RIS), says Sodickson, the RIS—as well as other data warehouses—are structured around an exam or an encounter. “We built sort of a patient-level repository for this information, so that for each patient you could record the full dose risk history over time,” he says. “Then, you need this system to interact with your order entry system so that you have a real-time alert or notification system set up. And none of these are terribly well-established connections.”
Nevertheless, Sodickson says he hopes to have the system implemented at BWH within a year and that BWH will be looking to expand it to its whole hospital group.
Radiation Safety Alert
When radiologist Steven Birnbaum, MD, began compiling dosage data at the two hospitals he worked at in New Hampshire several years ago, he discovered a number of patients who had received a “phenomenal” number of studies, he says. Birnbaum began educating himself to determine how much of a cumulative radiation dose patients could receive, and eventually came up with a dose tracking program. To help with this process, Birnbaum enlisted hospital CT technologists to review imaging histories.
Once high-dose patients had been identified, Birnbaum reviewed their histories, and if appropriate, decided they would be flagged in the RIS.
Birnbaum replaced the allergy field of the RIS with a “radiation safety alert” that would pop up when a study was requisitioned and would initiate a consultation between the referring physician and the radiology department.
Originally, Birnbaum’s radiation safety program was put together manually, but now information on individual patient imaging studies and radiation dose is compiled through data mining.
Other institutions are making efforts to tackle the cumulative radiation dose issue. The Rhode Island health system Lifespan, which includes five partner hospitals, keeps a running record of CT scans by patient for the entire system and makes that information available through its electronic data repository. And the National Institutes of Health (NIH) Clinical Center has taken steps to begin tracking radiation dose in its EMR. New CT and PET/CT equipment purchased by the Clinical Center—NIH’s research hospital in Bethesda, Md.—is required to routinely record radiation dose exposure in a patient’s hospital-based EMR.
The Integrating the Healthcare Enterprise (IHE) initiative has developed the Radiation Exposure Monitoring (REM) profile, which specifies in detail how imaging equipment should support a dose-monitoring program like that of the NIH.
The profile is an implementation guide for vendors—designed by vendors—that documents how to report dosage using current Digital Imaging and Communications in Medicine (DICOM) standards to facilitate collection and distribution of patient radiation exposure information.
Although the information collected could be distributed to an EMR, work on the REM profile has really been aimed at a “variant” of an EMR, says David Mendelson, chief of clinical informatics, Mount Sinai Medical Center in New York City and co-chair of the IHE subcommittee of the RSNA Radiology Informatics Committee.
Mendelson says that a possible way of aggregating cumulative dose information is through a registry. “While the EMR is a good aggregation point within any institution, you wouldn’t be aware of exams outside of that institution,” says Mendelson. “So the better model might be a resource your EMR could talk to that is available universally and would tell you what [a patient] has been exposed to lifelong.”
One of the possibilities for such a registry could be the American College of Radiology, says Mendelson. He also points out that there are private companies developing registries in parallel with the development of dose monitoring systems.
“But there’s no reason—the way the profile is written and if you’ve provided a field in your local EMR—why that information can’t be ported to your own EMR,” Mendelson says.
So, could radiation dose be tracked universally? “Part of me says that it should be relatively straightforward and easy to do,” says McEnery. “But then, look how long it took DICOM to get established as an imaging standard—we’re still trying to figure out how to get images transmitted between institutions.” Questions about security and privacy concerns also will complicate matters, McEnery adds.
“Obviously, it would be a lot easier if we had a national EMR,” says Birnbaum. “But systems have to talk to each other—ERs need to talk to each other. But I don’t even talk to the other hospital across town.”
At this point, he says, it’s really up to individual institutions—and patients—to monitor radiation exposure, adding that if these institutions don’t realize there’s a problem, they soon will.
“I will challenge any radiologist in this country to look at his or her patients, look at the data, and then tell me he or she don’t have any of these patients [who have received high levels of radiation exposure through imaging],” says Birnbaum. “All of us do. The idea is to prevent it as we go into the future.”