Radiation Dose Roundtable: Progress, Potential Upsides & Remaining Challenges

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
Managing CT radiation dose remains a critical issue for providers across the country. As studies attempt to estimate and the media too-often exaggerates the risks of radiation-induced cancers, referring physicians and patients are looking to the radiology community for answers.

Dose has become the differentiator in the CT market and a red hot topic of clinical conversations. An array of dose management software has hit the market, but there is no magic bullet to slash radiation dose. Dose reduction requires a series of incremental strategies including multiple IT tools, protocol optimization, operational changes and conversations among physicians and physicists within and among institutions. This month, Health Imaging & IT gathers leaders in the CT community for a status update. We found:
  • The top dose reduction strategy is eliminating unnecessary studies;
  • ED physicians, oft-criticized for overutilization, face particularly challenging legal constraints; and
  • Early dose tracking registries may be immature and misleading.


How has clinical understanding of radiation dose evolved in the last year?

McCollough: The heightened media focus has spread misinformation and raised anxiety among physicians and patients. However, the positive outcome of all the media attention is that it has put a spotlight on radiation dose. This has led to a focus on “right-sizing” dose for specific patients and applications, and it’s put pressure on manufacturers to spend research and development dollars to implement new technologies and other dose reduction strategies.

But bad outcomes have occurred at the clinical level. If the primary care physician accepts what is propagated in the media, the “overblown level of risk,” and makes a fundamental change in ordering patterns, patients may not get important imaging studies that could contribute to their medical care, or may get less accurate or more expensive studies.  

Multiple technologies—decision support, dose tracking and new algorithms—promise to address various elements of the dose issue.  Are we seeing wider availability and adoption of decision support systems to help cut dose?

Thrall: In some sense, the most harmful radiation is radiation from a study not indicated in the first place. No protocol optimization or dose reduction strategy even applies in this situation. This is sometimes overlooked in the rush to optimize protocols and buy new scanners with lower radiation profiles. The most dramatic way to reduce risk is to not do studies when they are not needed.

We are beginning to see substantial interest in and adoption of decision support systems. The best approach is to use a computer-based decision support system that can fit right into the workflow of a physician ordering an exam. That way, he can get feedback at the point of care on the appropriateness of the study and at the same time not be deflected from his interaction with the patient or burn up a lot of time reviewing indications.

Where do we stand as far as leveraging IT and implementing operational changes to reduce duplicate CT studies?

Thrall: The software has not been developed commercially to address the duplicate exam issue. Now it’s being done on an institution by institution basis.

At Massachusetts General Hospital, we are able to look at patients’ entire histories throughout Partners Health System, including Brigham and Women’s Hospital and our community hospitals. The physician is informed in the CPOE at the point of care if the patient has had or is scheduled to have a similar exam. This alone has eliminated 3 to 5 percent of exams.

What about new protocols to cut dose? Are there strategies that can be employed during CT scanning?

Lesser: Overall, our doses have dropped dramatically with newer scanners and new awareness of how to manage studies. With new equipment, doses are extraordinarily low. But staff has to pay a lot of attention on how the scan is set up.  Every small aspect is very important in achieving low dose.

For instance, we use a very high-pitch scan (a gated flash scan, which is a pitch of 3.4) on children under the age of two. If the provider pays appropriate attention to the timing of contrast, the scan can be acquired at 0.5 or 0.1 mSv, or less.

We’ve evaluated dose for coronary artery imaging in patients from ages 3 to 17 and found they remain less than 3 mSv, even for 17 year olds. But patients have to be beta-blocked, and we have to make the right protocol choices.

Mayo-Smith: Over the last five years, Rhode Island Hospital and other institutions have developed and revised protocols from a more generic one-size-fits-all model to a much more specific approach that can drop dose as much as 30 to 40 percent depending on the study. In the protocol development process, we dial back dose until image quality is markedly reduced.  

Dose exposure depends on three questions: the body part, the scanner and the indication, which is why we have so many protocols today. Rhode Island Hospital has 413 protocols, and we modify them all of the time.

Can you provide an example of how specific protocols impact dose?

Mayo-Smith: Ten years ago, most abdominal CT exams used the same dose. Now, we use a certain protocol for general abdominal pain. If a patient presents in a certain way and we are specifically looking for a kidney stone we can employ a very low dose protocol with exposure that is 30 to 40 percent less than a standard abdominal CT.   

How can smaller practices best approach low-dose CT protocols?

Thrall: I would urge smaller institutions to contact larger institutions to talk with them about protocol optimization. This turns out to be a non-trivial exercise. There are dozens of protocols and dozens of variations. A small group is going to be challenged to have the expertise in each organ system to optimize protocols. However, larger centers do this as part of the academic exercise. I hope that people will develop partnerships with flagship institutions to better disseminate low-dose protocols.

Where do we stand as far as development and availability of dose tracking tools? How mature are the systems?

McCollough: The American College of Radiology (ACR) registry tracks a number that reflects the radiation output of a CT scanner, called CT Dose Index, or CTDI. The CTDI is now an official DICOM field, so it will be easier to record and analyze going forward. This lets us look for trends in how much dose is used and ask questions about how dose management strategies are performing. But there are a lot of installed systems where the CTDI is not placed into the DICOM header, so manual data entry will be required to track doses from those systems. And there are two big caveats about dose-tracking tools that are essential to recognize.  

First, CTDI data are reported into a single database. The site can calculate the minimum, maximum and median CTDI values. They can look at the standard deviation, which, in many cases, will be quite large. The institutional response might be “we are not doing a very good job with dose management because we have such a big range of dose values.” But scanner output needs to be adjusted to the patient size. If the practice is scanning correctly and adjusting for patient size (which varies a lot), it will see a big variation in CTDI.  

When a registry takes the scanner output and throws all the data together to analyze, without correcting for patient size, the end result is an inaccurate picture. Current dose-tracking tools are not mature in terms of adjusting for patient size. But we recognize that issue, and are working to have a patient size metric put into the DICOM header.

The other challenge is the lack of consistent series description names. The ACR registries are receiving all these data that are associated with many different types of abdominal scans. What one institution might call “a low dose renal stone scan” might not have the same name at another institution. The result is a significant spread in doses that looks like they are for the same clinical indication, but aren’t. Some studies, such as for pancreatic cancer, need more dose; some studies, such as for renal stones, need less dose.  

Until we size adjust the scanner output and make sure we are sorting our dose data by the same diagnostic indications, some early dose-tracking reports are going to be pretty misleading.

One CT hot spot is the emergency department. What are the unique challenges in the ED? What strategies need to be employed to promote appropriate use and reduce duplicate studies in the ED?

Thrall: ED physicians are really in a box when it comes to utilization. Last spring, emergency physicians were invited to a joint seminar organized by the National Council on Radiation Protection and the ACR to discuss overutilization. The ED physicians pointed out that they have to work under the Emergency Medical Treatment and Active Labor Act (EMTALA), which states that emergency physicians who do not fully evaluate a patient can be fined $50,000 per occurrence.
They also pointed out that no appropriateness criteria, including the ACR Appropriateness Criteria, have legal standing either legislatively or in case law that would protect them. As much as many of us feel that at least some of them overutilize imaging we don’t have legal protections to change this behavior right now. So the strategy is to fully evaluate patients. If they are going to err, they will err on the side of a more complete versus a less complete evaluation.

What about special populations—pediatric patients, screening of asymptomatic patients? What progress has been achieved?

Thrall: The Image Gently campaign has really caused people to totally rethink the approach to pediatric imaging. One principle of Image Gently is image once. In the past, it was very common to do imaging with and without contrast. Now we use a strategy of multi-phase contrast injection so we only have to image once, which reduces dose by one half.

Mayo-Smith: In terms of screening patients, screening doses are significantly less than diagnostic doses. Doses for CT colonography exams are significantly lower than an abdominal CT scan. Likewise, for lung cancer screening, the dose is lower than a traditional chest CT. For chest CT screening, it is important to note that the preliminary studies are not in the general population but a select subset of high-risk patients with a heavy smoking history.

One frequently recommended strategy is to employ alternate, non-ionizing radiation imaging modalities where appropriate and available. Is there a shift away from CT and toward other modalities? What factors need to be taken into account as facilities consider this strategy?

McCollough: We are seeing a drop in CT volume in many centers around the country, with a shift toward different modalities.

We have to look at cost and access as well as safety. The radiation risk from CT is extremely low, but some people suggest MR now instead of CT because of potential risk.  But there is risk associated with gadolinium [contrast], and some MRI scans may require anesthesia, which carries risk. There are burns and other accidents that can happen in MRI, though the numbers are extremely low and everyone takes precautions to prevent them. MRI should not be considered a panacea, as if there were no risk associated with it. Furthermore, it’s not equivalent to CT in cost or access.  

Are there additional strategies to consider in the debate?

McCollough: I would really like to bring into the debate, and I don’t have the solution for it, a way to consider the risk of not doing the appropriate scan. On one hand, everyone is saying don’t overutilize imaging. But on the other hand, doctors and patients are afraid of missing something subtle. So we have to accept some uncertainty if we are going to image less. On top of proper utilization issues is the radiation concern. But if a patient turns down a recommended scan, or the doctor does not order it because of fears about radiation, there is a risk of not getting the right exam. That issue is not very well quantified or considered, but I do hear enough anecdotal stories and talk to enough patients to know that some patients are refusing important imaging exams.

Online Resources

www.imagewisely.org, a collaborative program focused on adult dose optimization, features equipment resources and information for referring providers and patients.
www.imagegently.org provides resources and practical strategies for all stakeholders to raise awareness of opportunities to lower pediatric radiation dose.
www.medicalimaging.org/policy-and-positions/radiation-dose-safety/ offers information about principles to reduce exposure and the CT Dose Check Initiative, which emphasizes new features to ensure safe CT scanning.