Radiation risk issues in recurrent imaging

Abstract

Millions of patients benefit from medical imaging every single day. However, we have entered an unprecedented era in imaging practices wherein 1 out of 125 patients can be exposed to effective dose >50 mSv from a single CT exam and 3 out of 10,000 patients undergoing CT exams could potentially receive cumulative effective doses > 100 mSv in a single day. Recurrent imaging with CT, fluoroscopically guided interventions, and hybrid imaging modalities such as positron emission tomography/computed tomography (PET/CT) is more prevalent today than ever before. Presently, we do not know the cumulative doses that patients may be receiving across all imaging modalities combined. Furthermore, patients with diseases with longer life expectancies are being exposed to high doses of radiation enabling radiation effects to manifest over a longer time period. The emphasis in the past on improving justification of imaging and optimization of technique and practice has proved useful. While that must continue, the current situation requires imaging device manufacturers to urgently develop imaging technologies that are safer for patients as high doses have been observed in patients where imaging has been justified through clinical decision-support and optimized by keeping doses below the national benchmark doses. There is a need to have a critical look at the fundamental principles of radiation protection as cumulative doses are likely to increase in the coming years.

Introduction

In current clinical practice, it is not rare for patients to undergo recurrent high-dose imaging exams in a lifetime.^1,2 Recent reports have shed light on the increasing frequency of CT, positron emission tomography/computed tomography (PET/CT), and fluoroscopically guided interventional procedures. For example, a recent study comprising 2.5 million patients who underwent 4.8 million CT exams found that patients underwent a median of 6 CT exams in a year and that some patients received up to 109 exams over 5 years.³ As the radiation dose in any imaging exam may vary widely, it is more appropriate to discuss this topic in terms of the radiation dose involved rather than the number of imaging exams performed.

It is recommended to use measured radiation dose quantities whenever possible for the sake of maintaining accuracy. For CT, the volume weighted CT dose index (CTDI_vol) is a quantity measured in a phantom. However, these measured quantities, which represent radiation output from the machine and patient exposure, become less applicable to real-world clinical settings where patient dose rather than scanner output is needed. Typically, several types of measured dose quantities are involved, and many organs are exposed in patients undergoing recurrent imaging of different body regions. Effective dose (E) is a dose quantity defined by the International Commission on Radiological Protection (ICRP) representative of the stochastic risk of radiation.⁴ Although it is inappropriate to utilize effective dose to determine the individual risk of radiation; it remains the best way to discuss radiation exposure to patients (or representative phantoms, to be precise) in a cumulative manner.⁴ In the setting of recurrent imaging, as per current practice, organ doses cannot be meaningfully used and thus effective dose remains a meaningful metric for discussing radiation doses.

In the cohort study mentioned above, patients who received a cumulative effective dose (CED) ≥ 100 mSv had a median CED of 130.3 mSv and a maximum CED of 1185 mSv over a period ranging from 1 to 5 years.³ Roughly, 1 out of every 100 patients who underwent a CT exam received a CED ≥ 100 mSv over 1 to 5 years. Furthermore, it is estimated that an additional 0.9 million patients get added to this CED ≥ 100 mSv cohort every year globally^2,3,5 and approximately, 2.5 million patients reach a CED level ≥100 mSv in a 5-year period across 35 member countries of the Organization for Economic Cooperation and Development (OECD).⁶ These recent studies focused on cohorts of patients with CEDs ≥ 100 mSv not because 100 mSv is a threshold for radiation effects, but rather because CED ≥ 100 mSv correlates with many organs receiving a dose >100 mGy at which there is a high degree of confidence on radiation effects among both international and national organizations.^3,7

The purpose of this paper is to review the existing literature on recurrent imaging studies that involve X-rays and/or radiopharmaceuticals resulting in high cumulative radiation doses to patients and the associated potential radiation risks.

Historical aspects

In the past, unsatisfactory image quality was a driving factor for repeat imaging studies. Such imaging practices were typically discussed in terms of repeat or reject rates and significant emphasis was placed on reject rate analysis.⁸ The term recurrent imaging is more appropriate for situations where multiple imaging studies are performed for clinical reasons. Tracking radiation doses was not routinely implemented as there were no mechanisms to facilitate tracking such as fixed patient IDs, dose displays on imaging machines, structured dose reports, networks to transmit dose information, or automatic methods of detecting cases with repeat studies. Therefore, the focus was on quantifying radiation risk from avoidable repeats rather than the cumulative dose to which individual patients were exposed.⁸

Also, prior to 2000, CT imaging using single detector scanners was slow requiring several minutes to scan a given body region. As such, recurrent imaging on the same patient was not common except in the follow-up of malignancy (although still at a much lower rate than today in this patient population). Aside from the speed of the exam and radiation risks, cost and availability were the primary factors limiting frequent and recurrent use of CT imaging. However, with the advent of helical CT imaging, scanners became faster and led to the erroneous perception that radiation doses might decrease with shorter scan times.⁹ Increased affordability and availability of CT scanners led to drastic increases in the use of CT imaging as well as the adoption of CT imaging covering larger body areas (e.g. abdomen/pelvis or even chest/abdomen/pelvis). As a result, many radiographic examinations were replaced with CT imaging. Given the increasing use of CT imaging, research in the early 2000s began to focus on radiation risks of CT imaging and, in particular, radiation risks in children.^10,11 As cancer risks from CT imaging gained more attention in mainstream media, radiation dose reduction and, to some extent, dose optimization movements began to gain momentum.^12–16 With increasing emphasis on cancer risks, industry responded by prioritizing radiation dose reduction and began competing with each other on this metric.¹⁷

The concept of tracking cumulative doses appeared in the latter half of the 2000s.^18–21 During the first decade of this century, the implementation of electronic medical records, availability of picture archiving and communication system (PACS), and standardization of radiation units in diagnostic radiology enabled dose tracking systems.^22–24 The push by the International Atomic Energy Agency (IAEA) Smart Card/SmartRadTrack project and its position statement issued jointly with other organizations such as the European Society of Radiology (ESR), Food and Drug Administration (FDA), International Organization for Medical Physics (IOMP), International Society of Radiographers and Radiological Technologists (ISRRT), World Health Organization (WHO), and the Conference of Radiation Control Program Directors (CRCPD) was integral to the commercial development of dose management systems.²⁵ A series of papers in the early 2010s discussed various aspects of dose management systems such as patient IDs,²⁶ prototypes of smart cards,²⁷ tracking of examinations and doses,²⁸ templates and models for patient exposure tracking,²³ and worldwide surveys to identify dose tracking and management needs.²⁴ A number of publications of the European Commission, IAEA and a position statement by Heads of European Radiological Protection Competent Authorities (HERCA) focussed on the issue of justification.^29–33 The positive dose tracking experience from a pediatric hospital in Finland demonstrated strengthening the process of justification and optimization through tracking of procedures and doses.³⁴

While we progressed from an emphasis on repeat imaging and reject rates, the concept of recurrent imaging was consolidated with a focus on identifying patients who receive high cumulative doses – for instance, ≥100 mSv – over a period of several years.^3,35,36

Imaging involving X-rays and considerations of radiation risks

It is appropriate to consider various imaging modalities separately to better understand the radiation risks involved.

Radiography

Radiographic studies like plain radiography, mammography, and dental exams are recurrent, but still, they involve relatively low doses. A single plain radiograph of the chest, head, abdomen, or pelvis involves an effective dose between 0.02 and 2 mSv. Thus, even 10 chest radiographs result only in a CED of approximately 1 mSv, which will, of course, depend upon the views taken. Similarly, 10 radiographs of abdomen or lumbar spine result only in a CED of <10 mSv. Tables that provide radiation doses for individual radiographic examinations, including dental radiographs, are widely available from different publications, e.g. NCRP publication 184³⁷ in the United States and other publications in the United Kingdom.^38,39 For plain radiography, a publication on extremely low birth weight infants indicated a mean of 31 radiographs performed over the first few months of life, including an average of 17 chest radiographs, 5 babygrams, and 9 abdominal radiographs.⁴⁰ The majority of chest radiographs and babygrams are performed in the first month of life, whereas the frequency of abdominal radiographs increased during the second month. Another study found that patients in a neonatal intensive care unit (NICU) underwent an average of 4.2 (range 1–21) radiographs during the course of their admission.⁴¹ The value of routine X-rays in intensive care unit settings has been questioned, both in NICUs and adult intensive care units.^42–44 Recurrent imaging is also common in dental radiological practices and some studies have drawn attention to potential adverse effects.^45–47 It is prudent that all exams are properly justified and imaging pathways should be appropriate in terms of modality and frequency irrespective of the doses involved.

Computed tomography

CT imaging has been the focus of most research on recurrent imaging. In addition to the drivers listed above, the relative ease of calculating cumulative effective doses with widespread availability of dose information in patient files (e.g. CTDI_vol and dose–length product [DLP]) has facilitated dose tracking in CT. Of course, increasing reliance on CT imaging and relatively higher doses compared to radiography also contribute to the increased scrutiny of recurrent CT imaging. Previously, the conversion of DLP to effective dose required additional computer software that could not directly interface with radiation dose structured reports (RDSRs) from imaging machines. Since commercial dose management systems (DMS) became available in recent years, it has become easier to track the radiation doses patients receive, obtain data on patients with defined cumulative doses (such as 100 mSv), and create alerts when patients receive doses above a certain threshold. Lack of availability of DMSs in certain regions of the world contributes to a lack of understanding of cumulative doses to patients globally.

There has also been some resistance to tracking cumulative doses because of the fear that such information might be used to prevent a clinically indicated imaging study.^48,49 However, cumulative dose data has been available via thousands of DMSs installed in Europe and the United States for many years without any evidence suggesting information on cumulative doses prevents necessary clinical imaging. Referring physicians play the most significant role in referring patients for CT imaging in most countries with the exception of a handful countries where radiologists hold official responsibility for accepting a CT referral.^50,51 International standards assign joint responsibility to radiologists and referring physicians for justifying a given imaging study.^52,53 Referring physicians are often unaware of radiation risks involved with CEDs greater than 50 or 100 mSv and that patients are not subject to dose limits.⁵⁴ Thus, in routine clinical practice, the decision to pursue an imaging study is more often based on benefits rather than consideration of risks of radiation weighed against potential benefits.⁵¹ Readers are encouraged to review the dialogs between physician and medical physicist presented in a recent paper to gain better understanding of dilemmas faced by physicians in the process in real-world imaging prescribing practices.⁵¹ Therefore, the fears discussed above are often held by individuals who are rarely involved in the process of referring patients for CT imaging. On the contrary, resistance to measuring cumulative doses has culminated in missed opportunities to detect millions of patients in radiation risk zones.^2,3,6,32,35 Cumulative dose research has been integral in identifying patients who receive CEDs ≥ 100 mSv in a single day.^3,36,55 Recent data from a study of nearly 4 million patients from 279 hospitals demonstrated that 0.8% of patients (1 out of 125) received ≥ 50 mSv and 0.03% (3 out of 10,000) received ≥100 mSv in a single day from CT imaging.⁵⁵

Fluoroscopically guided interventional procedures (FGIs)

FGI procedures often replace surgical procedures that involve higher risks as compared to radiation risks associated with FGI. In view of their utility, FGIs are employed outside radiology in interventional cardiology, electrophysiology, vascular surgery, orthopedic surgery, urology, gastroenterology, and in many other clinical specialities.^56,57 Some FGIs are well known to impart high patient doses resulting in skin injuries.^56,58–63 However, a recent study reviewing interventional radiology procedures at a major hospital over 9 years found that 4% of patients had CEDs ≥ 100 mSv with a median CED of 177 mSv.⁶⁴ Further, the majority (~90%) of patients had all of their procedures within 12 months and 10.7% were under the age of 40. Patients whose age at first procedure was 40 years or younger most commonly had a chronic disease of the torso (54.6%) and the percentage of cancer was low (11.1%) among this patient population.

Nuclear medicine and hybrid imaging

While much attention has been focused on radiation from CT imaging, nuclear imaging studies such as myocardial perfusion imaging (MPI), PET/CT, and SPECT/CT involve high doses in individual exams. In recent years, there has been greater realization that hybrid imaging is being used in a recurrent manner resulting in high cumulative doses. For instance, in a study it was demonstrated that MPI contributed 22% of CED from medical sources in 2010 and >10% of the entire CED to the American population from all sources.^65,66 Interest is growing in assessing CEDs from recurrent hybrid imaging and readers should refer to articles published in this issue of the journal for additional information.

Clinical conditions with high rates of recurrent imaging

Recurrent advanced imaging with consequent radiation exposure is widespread across various clinical conditions. In particular, CT imaging and nuclear imaging were identified in 2009 as the major drivers of ionizing radiation exposure in patients, accounting for 75.4% of CED in the United States.²¹ Furthermore, CT imaging utilization across various clinical settings has been increasing over the past 40 years with concomitant increases in health-care costs but unclear corresponding improvements in clinical care or outcomes.^20,67–69 Patients with chronic conditions and recurrent disease are more likely to undergo recurrent imaging resulting in high CEDs.^5,20 In particular, patients with active or past malignancy are frequently exposed to recurrent imaging to both assess for disease progression and monitor for recurrence. In a recent study analyzing clinical indications for patients who received a CED ≥ 100 mSv over a 5-year period, approximately 90% of patients had a history of malignancy.³⁵

Aside from malignancy, various other non-malignant chronic and recurrent diseases have been identified as major drivers of recurrent imaging. A recent review identified cardiac disease, end stage renal disease (ESRD), Crohn’s disease, and patients who had undergone endovascular aortic repair (EVAR) as patient populations who were more likely to be exposed to high CEDs.⁵ Other research has identified patients with abdominal pain of any cause and urolithiasis as patient populations with higher rates of repeat imaging.^20,68,70 Supporting this observation, abdominal pain and flank pain are the presenting complaints with the highest likelihood of receiving a CT scan in the emergency department.⁶⁸

With regard to abdominal pain, the use of CT imaging has increased dramatically in the emergency department (ED) setting with associated increases in costs and length of stay but has not resulted in lower admission rates or fewer cases of missed surgical illness.^71,72 In a recent study of patients who received a CED ≥ 100 mSv over a 5-year period, abdominal pain and related complaints were the most frequent clinical reasons for CT imaging in patients without a history of malignancy.³⁵ Patients with inflammatory bowel disease, in particular Crohn’s disease, are likely to undergo recurrent imaging with CEDs exceeding 50 mSv in 10–30% of patients who underwent imaging.⁵ Small bowel obstruction (SBO) is another recurrent disease that presents with abdominal pain for which patients are subjected to recurrent CT imaging with recurrences rates for SBO from adhesions ranging from 15 to 50% at 10 years.⁷³ CT imaging remains the gold-standard for the diagnosis of SBO despite ultrasound demonstrating similar diagnostic test characteristics.⁷⁴

Urolithiasis is another highly prevalent disease with high rates of recurrent CT imaging.^70,75–77 Current research suggests that 30–50% of patients experience recurrence of urolithiasis within 10 years of their initial episode.^76,78–80 A large-scale retrospective cohort study from 2017 found that 82.6% of patients presenting to the ED with renal colic receive an abdominal CT scan.⁸⁰ In a 2007 retrospective study of 356 patients presenting with suspected urolithiasis to a tertiary care ED, 79% of patients received ≥2 abdominal CT scans over the 10-month study and 15% of patients received ≥4 CT scans.⁷⁰ Another retrospective study from 2006 found that 4% of patients with suspected renal colic had undergone three or more CT examinations with estimated effective doses ranging from 19.5 to 153.7 mSv.⁷⁵ Despite such high rates of CT imaging for recurrent urolithiasis, research suggests that only 10% of CT scans reveal alternative pathology (e.g. adnexal masses, pyelonephritis, colonic pathology, etc.) and that repeat CT scans rarely change clinical management.^75,81 A multicenter comparative effectiveness trial demonstrated that ultrasound for suspected urolithiasis is noninferior to CT imaging for the initial diagnosis of urolithiasis and is associated with lower cumulative radiation exposure for patients.⁸² This study was crucial in prompting the development of the American College of Emergency Physicians Choosing Wisely recommendation to avoid CT imaging in patients age <50 years with a known history of urolithiasis presenting to the ED with symptoms consistent with uncomplicated renal colic.⁸³

Beyond these clinical conditions, patients with cardiac disease, particularly patients admitted to a hospital for acute myocardial infarction or who require heart transplants, often undergo recurrent imaging studies and are exposed to high CEDs.^5,65 For instance, Einstein et al⁶⁵ retrospectively reviewed procedures involving ionizing radiation in patients who had undergone MPI and found that this subset of patients underwent a median of 16 procedures involving radiation exposure over 30 years. Furthermore, multiple MPIs were performed in 39% of patients and 31% of patients received CEDs > 100 mSv from all medical sources.⁶⁵

Patients who undergo EVAR (e.g. for abdominal aortic aneurysm repair) are another patient population for whom recurrent imaging is common with such patients often being exposed to CEDs > 100 mSv in the first year following EVAR with substantial ongoing radiation exposure thereafter.⁵ Moreover, in many cases CED estimates may not include imaging that occurs prior to EVAR suggesting lifetime CEDs may be larger.^84,85 Finally, approximately one-third of patients with ESRD on hemodialysis accrue CEDs > 50–100 mSv over only 3 to 4 years.⁵

Total dose from recurrent imaging exams

There is a paucity of research on total cumulative doses patients receive from different imaging modalities. A large fraction of the current DMSs do not track dose information from FGI, nuclear medicine, and hybrid imaging studies. Even if one ignores radiography exams that contribute small doses, it is imperative that that DMS tracking capabilities expand to provide cumulative radiation dose data for individual patients across all high dose imaging modalities. In spite of this limitation, existing research on cumulative doses for a variety of diseases is available and summarized in Table 1 (reproduced with permission from Brambilla et al).⁵ There is further need for patient-based research to identify the contribution of different imaging modalities to CEDs among high-dose patient cohorts.

Table 1.

Cumulative radiation exposure and patients with CED >100 mSv⁵

Author	Condition	No. Pts	X-ray Procedures	Age (years) Mean or Median§	Patients with CED >50 mSv	Patients with CED >100 mSv	Follow-up (years)
Chen et al66	Pts with cardiac imaging	9,0121	Only cardiac procedures	51.1	3173 (3.5%)b	75 (0.08%)d	3
Einstein et al65	Pts with myocardial perfusion scan	1097	All medical imaging procedures	62.2		344 (31.4%)	20
Stein et al86	Cardiac disease	8656	All medical imaging procedures	65.9	533 (6.2%)		3
Kaul et al87	Acute myocardial infarction	6,4071	All medical imaging procedures	64.9f	1060 (1.7%)e		---
Eisenberg et al88	Acute myocardial infarction	8,2861	Only cardiac procedures	63.2f	1,5090 (18%)a		1
Lawler et al89	Acute myocardial infarction	1,1427	Only cardiac procedures	68.0f	825 (7.2%)a,e		1
Kinsella et al90	Haemodialysis	100	All medical imaging procedures	58.9	26 (26%)	13 (13%)c	3.4 median
De Mauri et al91	Haemodialysis	106	All medical imaging procedures	65.3		17 (16%)	3.0 median
Coyle et al92	Haemodialysis	244	All medical imaging procedures	52.7	56 (23%)		4.0 median
	Kidney transplant	150	All medical imaging procedures	45.7	12 (8%)
De Mauri et al93	Kidney transplant	92	All medical imaging procedures	52.4	26 (28%)	11 (12%)	4.1 median
Desmond et al94	Crohn’s	354	All medical imaging procedures	32		55 (16%)c	15
Levi et al95	Crohn’s Ulcerative colitis	199 125	All medical imaging procedures (no interventional)	39	23 (7%)		5.5 5.0
Kroeker et al96	Crohn’s	371	All medical imaging procedures	40	27 (7%)	12 (3%)c	5
Butcher et al97	Crohn’s	127	All medical imaging procedures	45	8 (6%)		11.2
Estay et al98	Crohn’s	82	All medical imaging procedures	36	16 (20%)		9.6
Chatu et al99	Crohn’s	217	All medical imaging procedures	31	29 (13%)		8.3
Jung et al100	Crohn’s	777	All medical imaging procedures	29	249 (35%)		15
Fuchs et al101	Crohn’s	171	All medical imaging procedures	11 (paediatric)	14 (8%)		5.3
Sauer et al102	Crohn’s	86	All medical imaging procedures	12 (pediatric)	6 (7%)		3.5
Huang et al103	Crohn’s Ulcerative colitis Indeterminate colitis	61 32 12	All medical imaging procedures	11§ (pediatric)	6 (6%)		5
Brambilla et al85	EVAR	71	All medical imaging procedures	74	71 (100%)	66 (93%)	1.8

CED, cumulative effective dose; EVAR, endovascular aortic repair.

^aCED > 30 mSv.

^bCED > 60 mSv.

^cCED > 75.

^dCED > 150 mSv.

^ePer admission after acute myocardial infarction.

^fMedian.

Another patient population with documented high cumulative doses are patients with hereditary hemorrhagic telangiectasia (HHT) who have pulmonary arteriovenous malformations.¹⁰⁴ In a study of this patient population, CEDs ranged from 0.2 to 307.6 mSv with a mean of 51.7 mSv and the dose exceeded 100 mSv in 11% of patients. Interventional procedures and CT exams were the greatest contributors to radiation exposure accounting for 51% and 46% of the total CED, respectively. Factors associated with high cumulative exposure in this patient population were epistaxis and HHT-related gastrointestinal bleeding. Additionally, the number of patient-years was significantly associated with higher CEDs, given the continued need for imaging for the duration of patients’ lifetimes.

Critically ill patients are another patient population with high cumulative dose. For example, Kim et al studied cumulative radiation exposure in critically ill trauma patients.¹⁰⁵ They report the number of studies per patient (mean ± SD) across various modalities including plain film radiography (70.1 ± 29.0), CT imaging (7.8 ± 4.1), fluoroscopy (2.5 ± 2.6), and nuclear medicine (0.065 ± 0.33). The mean CED was 106 ± 59 mSv per patient (range 11–289 mSv; median 104 mSv). Furthermore, the authors found that age, mechanism of injury, injury severity score, and length of stay were not statistically significant predictors of high CEDs. Another study of critically ill patients demonstrated that 6.8% of such patients had CEDs exceeding 50 mSv.¹⁰⁶

A study of patients with skeletally immature idiopathic scoliosis treated with bracing or spinal fusion underwent mean of 20.9 (range 8–43) radiographs.¹⁰⁷ Patients who underwent surgical treatment underwent significantly more X-rays than those who were braced. Another study of patients with cystic fibrosis demonstrated that 24.3%, 15.6%, and 0.43% of patients had CEDs between 5–20 mSv, 20–75 mSv, and >75 mSv, respectively, over two decades.¹⁰⁸ Thoracic imaging accounted for 46.9% of the total CED and abdominopelvic imaging accounted for 42.9% of the total CED. Other research demonstrated that imaging-related cumulative post-transplantation radiation doses exceed 100 mSv over 5 years in 37% of patients with cystic fibrosis who underwent lung transplantation.¹⁰⁹ Finally, research on recurrent imaging in patients with pediatric malignancies found that this patient population undergoes an average of 3.2 PET/CT studies per child (range 1–14). Patients received an average effective dose of 24.8 mSv (range 6.2–60.7) per PET/CT exam and an average cumulative radiation dose of 78.9 mSv (range 6.2–399) from PET/CT imaging alone.¹¹⁰ These doses are on the higher side for imaging in children and emphasise the need for further imaging optimization in accordance with ALARA (as low as reasonably achievable) principle.

Radiation risks in perspective

Radiation risks have been classified by ICRP in two categories: (a) tissue reactions (deterministic effects) and (b) stochastic effects. Tissue injuries have been reported with interventional procedures since the early 1990s and continue to occur today.¹¹¹ Despite rotation of the beam to spread doses across larger areas of skin and adopting optimization strategies, there continue to be patients with difficult and prolonged fluoroscopy resulting in skin injuries of varying severity with more severe injuries requiring months to years to heal.^58,59,62,63 Skin injuries are best avoided through adoption of suitable technique, particularly in patients with high body mass index. Most injuries occur in patients with recurrent procedures involving the same region of skin.¹¹² In neuroradiological procedures, skin injury presents with hair loss.¹¹¹ There are a number of publications directed at the avoidance of skin injury^111,112 and the frequency of skin injuries ranges from high of 1 in 7000 to a low of 1 in 100 000 interventional procedures.¹¹³ However, the risk of not performing an FGI may outweigh the potential risk of skin injury associated with FGI. Notwithstanding utilisation of all available features in equipment for optimization, this highlights the need for the development of safer imaging technologies to enable safer imaging of patients who are at risk of adverse radiation effects. Similar calls for action have been voiced in recent articles discussing high cumulative doses from CT^3,35and by the IAEA.⁵ Emphasis must be placed on prospective planning of imaging strategies for chronic conditions as well as optimization actions in acute imaging procedures.

While skin injuries are avoidable and every precaution should be taken to avoid such injuries, increasing reports of CEDs > 100 mSv over a short period suggests that focus be devoted to stochastic risks which occur with a higher frequency of 0.5% per 100 mSv (i.e. stochastic risks occur for every 1 in 200 patients who receive a CED ≥ 100 mSv). Thus, the stochastic risk is about 35 times the risk of tissue injury. Based on a median dose of 177 mSv in fluoroscopically guided interventions as reported in a recent study, the risk of stochastic injury could be approximately 62 times that of tissue injury.¹¹⁴ It may be noted that the risk estimates for stochastic effects are for the general population and without consideration of age or sex.

The common teaching to date has been that tissue injuries are of greater concern in interventional procedures than stochastic risks except in the case of children. However, the risk frequencies estimated above suggest otherwise. The stochastic risks need to be seen in light of a latent period of many years and reduced probability of effects at higher ages as a majority of patients are in higher age brackets. On the other hand, minor tissue injuries such as transient erythema and hair loss are short-lived but more severe tissue injuries can be debilitating and significantly affect quality of life. As such, high radiation exposure in patients who are younger (e.g. age <40 years) and/or with diseases with normal life expectancies warrant greater consideration of cancer risks. Furthermore, the above risk figures apply to both CT imaging and interventional procedures. Thus, the best way to understand the stochastic risks is by considering the example of smoking, which also has latent period of many years before causing cancer. Effective actions across various domains of public health to limit smoking control have culminated in major reductions in cancer risks over the past two decades. In the case of recurrent imaging, the greatest intervention to limit radiation exposure likely will come from advances in technology that reduce radiation risk for necessary exams while we continue to focus on efforts to limit unnecessary recurrent imaging.¹¹⁵

Overuse and inappropriate use of imaging has been widely documented and many reasons have been identified for overuse.^13,16,116 Using imaging appropriateness criteria has been emphasized by several organizations.^16,29 But, imaging guidelines and appropriateness use criteria are not available for many conditions.³⁵ Wherever available, they are indicated for initial work-up and diagnosis and there is a lack of guidance on serial CT imaging and when different imaging modalities are involved.³⁵ Incorporating cumulative dose considerations into clinical decision-making regarding imaging practices remains controversial – both in terms of whether it should be considered and how to use it.^{34,48,54,117,118} There has been a recent attempt to challenge professionals on making headway in this direction by taking the example of drug prescriptions.⁵⁰ High dose imaging should be considered somewhat similar to “controlled drugs” with risk-stratification and associated framework for prescribers (referring physicians) and deliverers (imaging facilities). This issue warrants further research and discussion among international organizations.