• radiation exposure;
  • Crohn's disease;
  • ulcerative colitis;
  • IBD;
  • children


  1. Top of page
  2. Abstract


Children with inflammatory bowel disease (IBD) undergo imaging using ionizing radiation and may be exposed to high cumulative radiation. We hypothesized that children with IBD have high exposure to radiation from medical imaging.


An Institutional Review Board (IRB)-approved retrospective chart review from 2002–2008 was performed on all patients with IBD. Radiographic studies performed were recorded and exposure for each study was estimated.


A total of 117 children with IBD (86 Crohn's disease [CD], 31 ulcerative colitis [UC]) were evaluated. The median current exposure was 15.1 mSv in CD and 7.2 mSv in UC (P = 0.005). Computed tomography (CT) scan and small bowel follow-through (SBFT) were responsible for 43% and 36% of all radiation exposures, respectively. The rate of radiation was higher in CD compared to UC (4.3 versus 2.2 mSv/yr). In CD, the rate of exposure was highest in the first 3 years of diagnosis (8.2 mSv/yr), and no different between the 3–5 year follow-up and 5+ year follow-up groups (3.8 versus 4.3 mSv/yr). Using the annual dose rate in those followed for more than 3 years, an estimated 47 out of 78 (60%) children (40 CD, 7 UC) would exceed 50 mSv by 35 years of age.


Radiation exposure from medical imaging is high in a subset of children diagnosed with IBD. Estimation of radiation exposure at age 35 suggests a significant portion of children with IBD will have high radiation exposure in their lifetime. Nonionizing imaging such as magnetic resonance imaging (MRI) and ultrasound should be offered to children with IBD as an alternative to current imaging that employs radiation. Inflamm Bowel Dis 2011

The use of diagnostic imaging modalities continues to improve the clinical treatment of many chronic diseases such as inflammatory bowel disease (IBD), which includes Crohn's disease (CD) and ulcerative colitis (UC). The risk of ionizing radiation has been acknowledged for years, yet the number of computed tomography (CT) scans obtained in the United States has increased exponentially.1 Furthermore, several scientific publications indicating high radiation exposure due to medical imaging and a recent series of articles in the New York Times examining errors in diagnostic and therapeutic medical radiation have increased awareness of these issues.2–4

Exposure to ionizing radiation from medical imaging is increasing. The use of diagnostic imaging modalities continues to increase around the world, driven by multiple factors that include improved technology, patient demand, ease of access, the medicolegal environment, and a significant decrease in time required for imaging such as CT scans. Sedation is rarely necessary to perform CT scanning in children today. CT examination increased exponentially in the United States over the past 25 years, with more than 62 million CT scans obtained in 2006 compared to ≈3 million CT scans obtained in 1980.1 Furthermore, the medical radiation effective dose has more than doubled in the past decade.5 Chronic diseases often require repeated imaging which results in increased ionizing radiation exposure. Thirty-three percent of over 30,000 adult patients in a large cohort study underwent five or more CT examinations and 15% of this cohort received cumulative radiation doses in excess of 100 mSv. Many of these patients had recurrent CT examinations due to chronic disease or management of acute exacerbations of a chronic condition.6

Cancer risks increase with exposure to ionizing radiation. Radiation effective dose is used to estimate the cancer risk in human populations and it is based on the radiosensitivity of a key organ's exposure by a source of radiation such as an imaging study. Much of the current literature examining cancer risk due to radiation is based on survivor cohorts of atomic bombs dropped on Japan. Data from these cohorts suggest that even at low to moderate radiation doses (<100 mSv), survivors had a small but significant increased rate of cancers. Higher radiation doses and earlier age of exposure were both risks for increased cancer incidence.7–9 A large cohort of nuclear industry workers confirms this and also suggests that less than a 50 mSv exposure can increase the risk for some cancers.10

Children have more biologically active tissue and thus are more vulnerable to ionizing radiation which typically result in genomic damage.11 Children thus are more sensitive to radiation than adults and have a longer duration of life in which to develop cancer after exposure. Data from the atomic bomb survivor cohort clearly shows an increased lifetime risk of solid cancer deaths when exposed at a younger age.7–9 This cohort remains the largest cohort study to confirm that children are more sensitive to radiation due to its significant radiation exposure and greater than 60-year duration of follow-up. However, modeling from the Hiroshima data suggests that an abdominal CT of a 1-year-old results in a lifetime cancer mortality risk of 0.18%, more than 10 times the risk in adults.12

Children and adults with IBD are exposed to high radiation effective doses due to medical imaging in multiple cohorts.13–15 However, no studies have evaluated medical imaging utilization and estimation of radiation exposure with extrapolation based on rate of exposure, which may identify a much larger proportion of IBD patients at high exposure to medical imaging radiation.

In summary, radiation due to medical imaging has increased rapidly over the past two decades. Cancer risk may be increased with low radiation effective doses (<100 mSv) and children are at highest risk for ionizing radiation-induced cancers. The purpose of this study was to determine the utilization rate of IBD imaging and its radiation dose in a tertiary care cohort of children with IBD and extrapolate the yearly rate of exposure to determine a total lifetime exposure at age 35.


  1. Top of page
  2. Abstract

An Institutional Review Board (IRB)-approved retrospective chart review of children with IBD from 2002–2008 was performed. Children were included that were under the age of 18 years. Demographic data, date of diagnosis, date of first radiographic imaging, and all ionizing radiation imaging studies performed at Children's Healthcare of Atlanta were recorded, including abdominal-pelvic CTs, abdominal x-rays (AXR), chest x-rays (CXR), bone age x-rays, dual energy x-ray absorptiometry (DEXA) scans, and small bowel follow-through (SBFT) studies. Frequencies of these imaging studies were summed and utilization rates per year were calculated and compared for children with CD and UC. The total radiation effective dose was estimated based on typical radiation doses for each study based on age and compared in those children with CD and UC. A total radiation effective dose (mSv) and radiation effective dose rate (mSv/yr) was calculated from first imaging at this institution if prior to diagnosis or at diagnosis no previous imaging had been obtained, yielding the most conservative rate of exposure. Imaging that may have been performed outside of this institution was not captured in this analysis. While our tertiary-care hospital does receive some referrals from surrounding areas, those referred specifically for a second opinion were not included in this study. All children in the cohort continue to be followed by our institution and a majority of the patients in this study were diagnosed by physicians at our institution.

Radiation effective doses were estimated based on previously reported literature values as below and based on age for SBFT and abdominal/pelvis CT. Bone age x-rays and DEXA examinations were determined to have minimal radiation exposure and thus were not included in the radiation estimate. Radiation effective doses vary greatly based on patient weight, radiographic imaging settings, and number of sequences for each study. While some published studies estimate an abdomen/pelvis CT to expose patients up to 20 mSv of effective dose, published literature estimates that single-pass CT scan provides an average of 10 mSv effective dose.16 We chose this more conservative estimate due to both improved CT technology that reduces radiation as well as recent efforts to limit radiation exposure in radiologic imaging such as the Image Gently Campaign. It is possible and likely that the actual effective dose is higher than our estimates; however, we chose to be conservative in our estimates. CT scans performed early in our cohort likely exposed children to significantly higher amounts of radiation while more recent CT scans approach our estimates of exposure (Table 1).

Table 1. Radiation Exposure Estimate for Radiographic Studies
Radiographic StudyRadiation Exposure (mSv)
Chest x-ray (CXR)-1 view0.1
Abdominal x-ray (AXR)1
Small bowel follow-through (SBFT)(16) 
 <10 years of age3
 10-15 years of age5
 15+ years of age6
Abdominal/pelvis CT 
 <15 years of age5
 15+ years of age10
Bone ageMinimal

We considered multiple prediction models to estimate radiation exposure at an arbitrary age of 35 taking into account that sensitivity to radiation decreases with age. These models included total average or median effective dose for the entire cohort and individual average yearly effective dose. The average or median yearly effective dose for the entire cohort are not likely an appropriate measure, as those with more severe disease have a higher likelihood of radiation exposure. Therefore, we chose the radiation effective dose rate (mSv/yr) for each patient to predict total radiation effective dose at age 35 for only those who had been followed for greater than 3 years. Given that radiation exposure is not linear, we also used a more conservative prediction of an arbitrary half the effective dose rate to calculate a conservative estimate at age 35.

The imaging and associated radiation doses include only those exams at this one institution and only those exams ordered for evaluation of IBD issues in these children. Other imaging ordered at different institutions was not included in this analysis and therefore likely underestimate average radiation exposures to these children.

Statistical Analysis

Proportions were computed to summarize categorical data, while means and medians were calculated for continuous measures. When the distributional assumption was not valid, medians along with the range were calculated. Univariate differences between diagnosis groups (CD or UC) were assessed for the categorical data using the chi-square test of independence and Fisher's exact test. Due to the skewed, nonnormally distributed continuous data, our comparisons between disease groups were made using the Wilcoxon Rank-sums test. P-values of less than or equal to 0.05 were considered significant. Calculations were performed using Stata v. 10 (College Station, TX).


  1. Top of page
  2. Abstract

Patient Characteristics

A total of 117 children diagnosed with IBD between 2002 and 2008 who had been followed for at least 1 year were identified. As expected, the number of children diagnosed with CD (86) was significantly higher than children diagnosed with UC (31), with rates similar to other epidemiology studies.17 The average age of diagnosis for CD and UC was 12.4 years and 10.8 years, respectively (P < 0.05). The average years of follow-up for CD and UC were 3.5 years (SD 1.8 years) and 3.3 years (SD 1.4 years), respectively. The population was 64% Caucasian (n = 72), 31% African-American (n = 36), 2% Hispanic (n = 2), and 4% Declined (n = 4). Among the patients who are African-American, 94.4% were diagnosed with CD, whereas 63.9% of the Caucasians were diagnosed with CD. Other races were not considered in this statistical test due to the small numbers. The number of males and females was similar (61 males, 56 females). For CD there were 46 males diagnosed and 40 females diagnosed, similar to other reports suggesting a slight male preponderance in CD. For UC, there were 15 males and 16 females who were diagnosed, similar to other reports suggesting an equal gender distribution in UC17 (Table 2).

Table 2. Patient Baseline Characteristics
 Crohn's Disease (CD)Ulcerative Colitis (UC)
Number of pts (%)86 (73.5)31 (26.5)
Age at Dx (mean and SD)12.4 (3.3)10.8 (3.6)
Years of follow-up from diagnosis mean (SD)3.5 (1.8)3.3 (1.4)
 Range (years)1.0 - 7.01.4 - 5.7
 Male (%)46 (53.5)15 (48.4)
 Female (%)40 (46.5)16 (51.6)

Utilization of Imaging

For both CD and UC, Table 3 describes the median and range of all imaging tests performed. There were no statistically significant differences in the number of AXR, CXR, SBFT, or bone age studies when comparing CD with UC. Patients diagnosed with CD underwent statistically significant more CT (Abd/pelvis) and DEXA radiographic images (P < 0.0001 and P < 0.001, respectively). The median number of CT of the abdomen and pelvis was 1 with 75.6% of CD and 41.9% of UC patients undergoing at least one CT scan (Table 3).

Table 3. Ionizing Radiation Imaging Utilization: Number of Studies per Patient
 Crohn's DiseaseUlcerative ColitisP-value
  1. More than 75% of children with CD underwent at least one CT scan of their abdomen. Children with CD were more likely to have a CT scan and DEXA scan than children with UC. CD = Crohn's disease; UC = ulcerative colitis; AXR = abdominal radiograph; CXR = chest radiograph; DEXA = bone age study of lumbar spine; SBFT = small bowel follow-through fluoroscopic study; bone age = left hand and wrist single radiograph.

AXR range0–400–250.93
CXR range0–280–170.61
CT abd/pelvis range (% at least 1)0–5 (75.6%)0–3 (41.9%)<.0001
SBFT range (% at least 1)0–9 (64.0%)0–2 (64.5%)0.14
Bone age range (% at least 1)0–6 (38.3%)0–2 (19.4%)0.06
DEXA range (% at least 1)0–4 (57.0%)0–2 (29.0%)<0.01

Current Radiation Effective Dose and Rate of Exposure

Radiation exposure was estimated for each patient and the median radiation exposure was 15.6 mSv for CD and 7.2 mSv for UC (P < 0.001). A total of 6 (4%) children were already exposed to over 50 mSv of radiation and all six were diagnosed with CD. The rate of exposure was significantly higher in children diagnosed with CD than UC (P < 0.01) (Table 4).

Table 4. Current Radiation Effective Dose Exposure and Rate
 Crohn's DiseaseUlcerative ColitisP-value
 Median (Range)Median (Range) 
  1. Children with CD had significantly higher current exposure and rate of exposure than children with UC.

Rate (mSv/yr)4.32.2<0.01
Current total exposure (mSv)15.1 (0–72)7.2 (0–34)<0.001
Current effective dose   
High (>100mSv) 
Moderate - high (50-100msv)6 (7.0)00.21
Moderate (30-50mSv)8 (9.3)1 (3.2) 
Low (<30mSv)72 (83.7)30 (96.8) 

Rate of Radiation Exposure Stratified by Follow-up

When considering children stratified by duration of follow-up, those followed 1–3 years had a median exposure of 8.2 mSv per year, those followed for 3–5 years had a median exposure of 3.8 mSv per year, and those followed greater than 5 years were exposed to a median 4.3 mSv per year. There was a statistically significant difference between the 3–5 year and 5+ year follow-up groups when compared to the 1–3 year follow-up group but no statistical difference between the 3–5 year and 5+ year follow-up groups. In those who were followed more than 3 years, ≈20% (10 of 51) continued to have significant radiation exposure of at least 6 mSv per year (Table 5).

Table 5. Radiation Exposure Rate Stratified by Follow-up Period
 CD Patients (%)Median RateP-value
  • Radiation rate was highest in the first few years after diagnosis but remained stable after 3 years of follow-up.

  • a

    Compared to follow-up 1–3 years.

  • b

    No statistical difference between the follow-up 3–5 years and the follow-up 5+ years groups.

Follow-up 1–3 years (n = 35)   
MEDIAN 8.2 mSv/yr 
 0–3 mSv/yr7 (23%)  
 3–6 mSv/yr8 (20%)  
 6–10 mSv/yr10 (29%)  
 >10 mSv/yr10 (29%)  
Follow-up 3–5 years (n = 29)   
MEDIAN 3.8 mSv/yr<0.05a
 0–3 mSv/yr11 (38%)  
 3–6 mSv/yr14 (48%)  
 6–10 mSv/yr3 (10%)  
 >10 mSv/yr1 (3%)  
Follow-up 5+ years (n = 22)   
MEDIAN 4.3 mSv/yr<0.05b
 0–3 mSv/yr11 (50%)  
 3–6 mSv/yr5 (23%)  
 6–10 mSv/yr4 (18%)  
 >10 mSv/yr2 (9%)  

Estimation of Radiation Effective Dose at Age 35

Using the rate of radiation exposure since first imaging or date of diagnosis (whichever was lower), the predicted radiation exposure at age 35 was calculated in only those patients who were followed for more than 3 years. The median estimated radiation exposure at age 35 for CD and UC was 81.2 mSv and 40.3 mSv, respectively, with significantly higher median estimated exposure in CD compared to UC (P < 0.05). Likewise, using a more conservative method, the median estimated radiation exposure at age 35 for CD and UC was 41.1 mSv and 20.1 mSv, respectively (Table 6).

Table 6. Estimation of Radiation Effective Dose at Age 35 in Patients Followed 3 or More Years (60 CD, 18 UC)
 CD PatientsUC PatientsP-value
  1. Patients with CD had significantly higher estimated radiation effective dose than those with UC. Standard method = individual rate of radiation exposure. Conservative method = half the rate of radiation exposure.

Median exposure at age 35 in mSv (range) – standard method81.2 (0–317)40.3 (0–276)<0.05
Median exposure at age 35 in mSv (range) – conservative method41.1 (0–159)20.1 (0–138)<0.05
Estimated effective dose at age 35 – standard method
 High (>100mSv)222<0.05
 Moderate - High (50–100msv)185 
 Moderate (30–50mSv)116 
 Low (<30mSv)95 
Estimated effective dose at age 35 – conservative method
 High (>100mSv)31<0.05
 Moderate - High (50–100msv)191 
 Moderate (30–50mSv)142 
 Low (<30mSv)2414 

Using rate of radiation exposure, 40 out of 60 CD patients (67%) and 7 out of 18 UC patients (39%) would have received cumulative effective doses over 50 mSv by the time they were 35 years of age from IBD imaging alone. Using a more conservative model, 21 out of 60 CD patients (35%) and 2 out of 18 UC patients (11%) would have received cumulative effective doses over 50 mSv by the time they were 35 years of age (Table 6).

For children diagnosed with CD, cross-sectional evaluation suggests a small but significant number are already exposed to moderate doses of medical radiation. More important, using the rate or half the rate of exposure in those followed for more than 3 years, a significant proportion of patients will have at least moderate radiation exposure at age 35, with some children being exposed to very high amounts of radiation (Fig. 1).

thumbnail image

Figure 1. Radiation exposure: current and estimated in CD patients followed for more than 3 years.

Download figure to PowerPoint


  1. Top of page
  2. Abstract

We report a significant radiation exposure due to medical imaging in a cohort of children with IBD, which may increase their lifetime cancer risk. CD patients displayed a significantly higher imaging utilization and radiation dose compared to UC children. Keep in mind that we report radiation exposures for one disease only—IBD—and do not consider the added doses from ionizing radiation imaging from other medical conditions or accidents.

The rate of IBD radiation exposure was highest in the first few years after diagnosis; however, the rate of radiation exposure was similar in the 3–5 years of follow-up and the 5+ years of follow-up. This is not unexpected, as many children undergo radiological imaging shortly after diagnosis to determine the extent of disease. Furthermore, many children begin to experience complications after 3–5 years of follow-up, which often results in more imaging studies. It should be noted that our maximum length of follow-up was 7 years and actual radiation exposure likely varies with follow-up and complication rate. Interestingly, some children followed for more than 5 years continued to have very high yearly radiation exposure (>10 mSv/yr).

Using current radiation exposure, a small proportion of children have high medical radiation exposure increasing their lifetime risk for radiation-induced cancer. Estimating radiation exposure using a conservative method suggests a significant number of children would have moderate radiation exposure (>50 mSv) for the management of their IBD by the time they are 35 years of age, with some CD patients estimated to have high exposure at that time. This subset of patients may have more severe disease, thus being exposed to higher radiation due to repeat CT imaging. Our results show that children with CD have higher imaging rate and radiation doses compared to UC children. In addition, our results suggest that there is a subset of CD children at highest risk for radiation exposure, likely those with more severe disease or persistently active disease, although this specifically was not examined in this study. These children had repeated abdominal-pelvic CT scans accounting for the majority of their radiation exposure.

The current study supports literature that suggests that children and adults with IBD exposed to high levels of medical radiation exposure. A recent adult study from Australia revealed an at-risk radiation exposure (>50 mSv) in 11 out of 100 patients with IBD. There were no predictive factors for high radiation doses but the dosage was higher in CD when compared with UC.14 Another adult study demonstrated ≈15.5% of a cohort of 409 patients with CD were exposed to high doses of radiation (>75 mSv). Age less than 17 at diagnosis and severity of disease were associated with increased doses of radiation.13 A recent study in children revealed 34% of CD subjects and 23% of UC subjects were exposed to moderate diagnostic radiation. However, moderate radiation dose was defined as one CT in a 2-year observational period, which may not be indicative of true exposure.15

This study confirms and extends earlier work by extrapolating our data to estimate the risk at age 35 and includes a longer duration of follow-up averaging over 3 years. Our estimation of radiation exposure at age 35 further highlights the importance of perspective of a child's lifetime with regard to medical imaging and radiation exposure. Perspective is especially important in treating children with lifelong diseases, where long-term sequelae of disease such as cancer are uncommon in childhood, but where management during childhood could place these children at risk for morbidity as adults. Furthermore, the management of adult patients should be altered, with avoidance of unnecessary ionizing radiation imaging especially if these adults have been exposed to significant radiation as children.

The study is limited by its retrospective design and its tertiary-care study population. The radiographic imaging was only captured at Children's Healthcare of Atlanta and it is likely that some proportion of these children underwent medical imaging at other institutions, including referring hospitals and emergency departments, which was not captured. Therefore, our current radiation exposure results are likely underestimated. Our prediction model using both current rate of radiation exposure and a more conservative half the rate can be criticized, as a true linear exposure rate is unlikely. However, we only included those children who had been followed for greater than 3 years, as the radiation rate of exposure was no different in the group followed for 3–5 years than those followed for more than 5 years. Using a composite rate of exposure would not likely improve the model, as some IBD patients will not need any medical imaging while others may require multiple imaging studies. We acknowledge that past radiation exposure does not necessarily suggest future exposure, especially as providers become more aware of significant exposure and risk to medical radiation. Our prediction model is used in this instance to highlight and further underscore the importance that continued radiation exposure contributes to significant lifetime exposure in chronic diseases, specifically IBD. Current health systems lack adequate information systems to track cumulative doses across different medical specialists and from childhood into adulthood to make providers aware of this risk.

This study demonstrates significant exposure to medical imaging radiation in children with IBD, particularly but not limited to CD. Specifically, there seems to be a subset of patients with IBD and predominantly CD that required multiple imaging studies, while others required few imaging studies. It is often difficult for physicians to exhibit a long-term perspective; however, a lifetime perspective is necessary when treating chronic diseases, especially in children. Identifying high radiation exposure in a cohort of children with IBD and extrapolating the data to age 35 highlights the risk of multiple imaging studies that use ionizing radiation on the total lifetime radiation exposure, which may place patients at increased risk for cancer. Identifying radiation exposure as a potential problem is essential in finding a solution. Lee et al18 suggested that only 9% of Emergency Department physicians believed that there was an increased risk of cancer from a CT scan and even fewer members of the public.

Magnetic resonance imaging (MRI) is feasible and obtains excellent images of the small intestine and can identify abscesses along with other abdominal findings common in IBD, with no radiation exposure. The selection of MRI for imaging the small bowel in CD may be warranted, given its feasibility and excellent images, although the availability of MRI at institutions will need to increase to meet this need. Other nonradiation imaging modalities such as ultrasound should also be studied further for the monitoring and management of IBD. The careful use of medical imaging is necessary with a lifetime perspective when treating children and adults with IBD.


  1. Top of page
  2. Abstract