AUA QI Summit 2017: Potentially Avoidable Harms from Radiation: Keynote Address

Linthicum, Maryland ( Dr. Rebecca Smith-Bindman from UCSF provided the keynote address at the AUA Quality Improvement Summit, discussing potentially avoidable harms from radiation. Dr. Smith-Bindman started by highlighting the spectacular improvements in ultrasound, CT and MRI, noting that these tests are important and beneficial to our patients. However, these imaging modalities come with tradeoffs, and it is important to consider potential harms of imaging in addition to potential benefits, since both need to be included in selecting the appropriate imaging choice.

Imaging has soared over the last several decades, with more than 80 million CTs done annually in the US at a cost of $100-200 billion (growth is 2x the rate of total health care costs). The drivers of medical imaging are multifactorial according to Dr. Smith-Bindman, including (i) improvement in technology, (ii) patient demand: no perceived disincentive to imaging, (iii) physician demand: testing is often easier than seeing the patient, (iv) fear of malpractice, (v) profitability, and (vi) increased capacity resulting from a proliferation of equipment. Dr. Smith-Bindman notes that evaluating how much imaging is unnecessary is a difficult question to answer since there are no good studies, although estimates are around 30%. Factors driving unnecessary imaging include a lack of evidence-based guidelines, strong incentives to use imaging, and few disincentives to use imaging. The benefit of imaging is rapid and accurate treatable diagnoses, however potential harms include: (i) false positives/incidental findings/over-diagnosis, (ii) radiation exposure, (iii) wasting time and missing the opportunity to treat the patients, and (iv) unsustainable costs.

False-positives/incidental findings/over-diagnosis are all different but have important similarities. Indeed, none of these help our patients and all lead to a cascade of testing, labelling, anxiety, and increased cost. Over-diagnosis is a large problem that is hard to fix or explain according to Dr. Smith-Bindman. This entity reflects a disease under that microscope (often cancer), but will not progress and is not what patients and physicians truly feel is a disease. Imaging leads to a profound increase in detection of cases, but by definition will not lead to a reduction in advanced disease. Importantly, it is difficult for a patient to differentiate over-diagnosed disease from real disease, and this must be thought of as a harm of imaging. False positives lead to “chasing imaging findings”: in some cases there is no way of deciding what is important, but in many cases it is possible to know that a finding is likely not important, albeit difficult to ignore.

Radiation exposure is a substantial potential harm of imaging. Ionizing radiation can remove tightly bound electrons from their orbits, leading to a break in chemical bonds and disruption of DNA. It is estimated that 75% of all radiation exposure is from CT imaging, considering that CTs are used frequently and doses per scan are high. Different models have suggested that 2-5% of all cancers are secondary to imaging radiation exposure, but because the risks are delayed they are harder for patients to comprehend. Several studies of Japanese atomic bomb survivors have estimated that the median dose received was 40 mSv and that 11% of solid cancers among these individuals were associated to their exposure. Furthermore, radiation doses of only ~10 mSv was associated with an increased in leukemia and solid organ cancers in select survivors. In a study of 178,604 children in the UK undergoing a CT between 1985-2002 who were followed until 2008, 74 subsequent leukemias and 135 brain cancers were diagnosed and attributed to radiation exposure from CT imaging [1]. Within 10 years of receiving a CT scan, children who received doses of 30-50 mSv tripled their risk of brain cancer and leukemia; importantly, 10-20% of children received these doses from only one CT scan.

In 2014, Dr. Smith-Bindman led a multi-center, pragmatic, comparative effectiveness trial to assess whether initial imaging for patients with suspected nephrolithiasis should be a CT scan or an ultrasound [2]. There were 2,759 patients randomized to point-of-care ultrasound (n=908), radiology ultrasound (n=893) or CT (n=958). The overall rate of high-risk diagnoses with complications was only 0.4% and did not vary by imaging method. Return emergency department visits, hospitalizations, and diagnostic accuracy did not differ between the groups (followed for 6 months), and overall radiation exposure over the 6-month follow-up was highest among patients undergoing initial CT imaging. When looking at a subset analysis, they found that among 1,582 CT scans performed, only 7.6% had doses <4 mSv with a median effective dose of 11 mSv. Dr. Smith-Bindman and her colleagues concluded that an ultrasound should be the first test in patients with suspected stones, however noting that this does not imply that these patients should only undergo ultrasound imaging but that this should be the initial diagnostic test with further imaging warranted at the physician’s discretion.

Dr. Smith-Bindman has several hypotheses for why CT doses are so variable: (i) there are no comprehensive standards or guidelines on CT imaging and there is a sense that everyone should be free to choose the dose, (ii) doses should be as low as reasonably achievable (ALARA), but there are few guidelines for what doses are reasonable OR achievable, (iii) in the absence of guidelines, practice variation introduces unnecessary harm from excessive radiation, and (iv) no organization is responsible for collecting dose data. Dr. Smith-Bindman is currently leading a 5-year study (the International CT DOSE Collaboration) at academic and non-academic medical centers in the US and abroad to expand on their previous work. This will create broader dose benchmarks, assess what works and what does not work to optimize CT dosing, learn from a broader group of institutions, and provide strategies for delivering feedback to the participating sites. Registry data is being collected at a rate of 4,000 CTs/day with over 4 million CTs registered thus far; the study is nearly complete and results will be forthcoming.

Dr. Smith-Bindman concluded with sobering stats, suggesting that the use of CT scanning in children will result in 4,870 future cancers. Furthermore, reducing the highest outlier doses to the median dose would prevent 44% of these cancers. This highlights the need for standardized dosing and continued evaluation of practice variation.

Speaker: Rebecca Smith-Bindman, University of California-San Francisco, San Francisco, CA, USA

Written by: Zachary Klaassen, MD, Urologic Oncology Fellow, University of Toronto, Princess Margaret Cancer Centre, Twitter: @zklaassen_md, at the AUA Quality Improvement Summit - October 21, 2017- Linthicum, Maryland

1. Pearce MS, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukemia and brain tumours: A retrospective cohort study. Lancet 2012;380(9840):499-505.
2. Smith-Bindman R, et al. Ultrasonography versus computed tomography for suspected nephrolithiasis. N Engl J Med 2014;371(12):1100-1110.