Foetal Radiation Dose and Risk from Diagnostic Radiology Procedures: A Multinational Study

In diagnostic radiology examinations there is a benefit that the patient derives from the resulting diagnosis. Given that so many examinations are performed each year, it is inevitable that there will be occasions when an examination(s) may be inadvertently performed on pregnant patients or occasionally it may become clinically necessary to perform an examination(s) on a pregnant patient. In all these circumstances it is necessary to request an estimation of the foetal dose and risk. We initiated a study to investigate fetal doses from different countries. Exposure techniques on 367 foetuses from 414 examinations were collected and investigated. The FetDoseV4 program was used for all dose and risk estimations. The radiation doses received by the 367 foetuses ranges: <0.001–21.9 mGy depending on examination and technique. The associated probability of induced hereditary effect ranges: <1 in 200000000 (5 × 10−9) to 1 in 10000 (1 × 10−4) and the risk of childhood cancer ranges <1 in 12500000 (8 × 10−8) to 1 in 500 (2 × 10−3). The data indicates that foetal doses from properly conducted diagnostic radiology examinations will not result in any deterministic effect and a negligible risk of causing radiation induced hereditary effect in the descendants of the unborn child.


Introduction
In diagnostic radiology examinations, there is a bene�t that the patient derives from the resulting diagnosis, provided that they are fully �usti�ed. However, given that so many examinations are performed each year, it is inevitable that there will be occasions when an examination will be performed on a woman who subsequently discovers that she was pregnant at the time of her examination(s). It may also occasionally become clinically necessary to perform a radiological examination(s) on a woman who is known to be pregnant. In the later case, there must be rigorous �usti�cation of the examination and the procedure itself must be optimized to minimise the foetal dose [1][2][3]. In order to avoid the former, some special rules have been developed to apply to the exposure of potentially pregnant women (who are or who may be pregnant) in which radiological examinations of such women are restricted to a certain period following menstruation [4][5][6][7].
For the protection of the foetus from occupational exposure of the pregnant worker, the International Commission on Radiological Protection (ICRP) [1,2] considers that if a female worker has declared (i.e., noti�ed her employer) that she is pregnant, additional controls have to be considered to protect the embryo/foetus. It is the Commission's policy that the methods of protection at the workplace for women who are pregnant should provide a level of protection for the embryo/foetus, that is, broadly similar to that provided for members of the public. erefore, the working conditions of a pregnant worker, aer declaration of pregnancy, should be such as to ensure that the additional dose to the embryo/foetus would not exceed about 1 mSv during the remainder of the pregnancy [1,2]. Irrespective of how these protective measures are applied in practice, it is almost inevitable that inadvertent foetal radiation exposures do occur. In these circumstances, it is necessary to consult an expert in medical physics and request that the foetal dose and hence the projected risks be estimated. is may require the medical physicist to calculate the dose based on the knowledge of the technique factors used, or a simulation of the examination using a phantom loaded with dosimeters or review the data from published scienti�c literature [7][8][9][10][11][12][13][14][15][16][17][18][19][20].
Osei and Faulkner [11] estimated the foetal dose of 50 pregnant women in the north of England. Doses to the embryo/foetus varied between less than 0.01 Gy and 117 mGy, depending on the examination and the gestational ages ranged between 2 and 24 weeks. Angel et al. [18] used Monte Carlo simulations to investigate foetal radiation dose resulting from an abdominal and pelvic examination for a range of patient gestational age and maternal size. e normalized foetal dose estimates from the Monte Carlo simulations ranged from 7.3 to 14.3 mGy/100 mAs, with an average of 10.8 mGy/100 mAs. Doshi et al. [19] also investigated foetal radiation dose from CT pulmonary angiography in late pregnancy. Dose measurements were made on three helical CT scanners, with an anthropomorphic phantom representing the chest and abdomen in late gestation. e estimated foetal doses from the CT scans of the maternal chest were in the range of 60-230 mGy, however by modulating the mA, shielding with a lead coat and scanning 5 cm shorter lengths, foetal doses were reduced by 10%, 35% and 56%, respectively. e Health Protection Agency, UK [20] has recently published foetal doses for some common diagnostic medical exposures. Foetal doses were derived from doses to the uterus observed in recent UK surveys. e foetal doses ranged from about 0.001 mGy to 50 mGy depending on the examination.
An accurate approach to provide an estimate of foetal dose, either prospectively or retrospectively is to assess the dose for the individual patient using the technique parameters that were used for the patient's examination and taking into account both foetal depth and size. Simply looking up the average uterus dose for a speci�c examination may lead to underestimation or over estimation of the foetal dose [8][9][10][11]. Requests for foetal absorbed dose and associated radiation risk estimates are usually sent to medical physics departments where quali�ed medical physicists with expertise in foetal dose calculations will estimate the dose and risks associated with the exposure. In this study, we contacted several of these departments in different countries where requests for foetal dose estimate have been sent. Speci�c values for the following parameters were requested: projection and view (e.g., AP, PA) for each examination, beam quality (e.g., kVp, �ltration), source-to-imager receptor distance, machine outputs, and techniques parameters (kVp, tube current, exposure time, etc.) used for the examination(s). In all, we collected and studied exposure data of 367 foetuses and 414 examinations from �ve institutions in �ve countries. e purpose of this paper is to present the results of this study.

Foetal Dose Calculations.
We contacted several departments in different countries where requests for foetal dose estimate are usually sent. Speci�c values for the exposure parameters for simple examinations (e.g., chest, etc.), complex (�uoroscopic) examinations (e.g., barium meal, etc.), and CT scans were requested using (see the tables (1A-1C) in Appendix 1 in supplementary material available online at http://dx.doi.org/10.5402/2013/318425). All foetal doses and risks were calculated using the most recent version of FetDose [8] (FetDose V4). FetDose is a computer program developed for the estimation of the radiation dose to the foetus. It calculates the absorbed dose and risk to the foetus from conventional radiography, �uoroscopy, computed tomography, and radiation therapy procedures performed on the pregnant patient. It also calculates the dose to the foetus from occupational exposure of the pregnant worker in the radiological department. e normalized uterus doses (NUD) used for foetal dose calculations are taken from NRPB SR262 [12] and NRPB-SR250 [17]. A more detailed information on the methods used for foetal absorbed dose calculation and hence risk estimations using the FetDose program can be found elsewhere [8], however a short summary is given here. e foetal absorbed dose from a series of radiographic examinations, , when the quantity supplied is the entrance surface dose (ESD rad ) for each radiograph is calculated as [ is the number of radiographs, NUD ESD is the uterus dose at a mean foetal depth normalized to free-in-air entrance surface dose, and SF is the foetal size factor (i.e., uterus to foetus dose conversion factor). For complete examinations, including �uoroscopy (e.g., barium enema, barium meal, etc.), involving both spot �lms and screening procedures of different areas of the body, the total foetal dose is calculated by summing the contributions from both the spot �lms and screening procedures as If the quantity provided is the dose-area product per examination, in Gy-cm 2 , then the foetal absorbed dose from a series of radiographic examinations is calculated from where NUD DAP is the uterus dose at a mean foetal depth normalized to the dose-area product, and DAP is the dosearea product for each examination, . e foetal dose from a series of computed tomography scans is calculated from = NUD * CTDI so tissue * (mAs/100) Pitch (6) where CTDI so tissue (mGy/100 mAs) is the CTDI air to ICRU muscle (CTDI so tissue = CTDI air * 1.07) used as an approximation to the dose to so tissue within the body, NUD is the sum of normalized doses for all 5 mm slabs lying within the scan volume. e National Radiological Protection Board (NRPB) report SR250 [17] provides normalised organ dose data of 23 series of Monte Carlo calculations modelling the conditions of exposure relevant to 27 common models of scanners. Each set of calculation involved 5 mm transverse section (slab) of a mathematical phantom. Estimates of normalised dose to 27 organs or regions of the phantom are calculated for 208 contiguous slabs of the phantom. e contribution of the normalised dose to the uterus is used in FetDose for the estimation of foetal doses.

Consequence of Foetal
Irradiation. e FetDose program presents the consequence of foetal irradiation in four different ways: (1) risks (the probability that a consequence has to be expected), (2) safety (the probability that a consequence is not expected), (3) equivalent number of chest X-rays, and (4) equivalent period of exposure to natural background radiation. In this paper, the consequence is expressed only as the probability that a consequence is expected (risk). Foetal radiation risk ( ) is calculated as where RC is the risk coefficient for the consequence of interest [8]. Radiation effects considered include childhood cancer induction (risk coefficient used is 8.0 × 10 −5 per mGy [4][5][6]), hereditary effects (risk coefficient used is 0.5 × 10 −5 per mGy [4][5][6]), decline in IQ (risk coefficient used is 25 × 10 −3 IQ points per mGy [21]) and severe mental retardation (risk coefficient used is 43 × 10 −5 per mGy [22]). ese risk coefficients have been mainly derived from high doses and doserates, and the extrapolation to low doses is far from veri�ed. Nevertheless, most experts accept that the linear nothreshold (�NT) model best �ts available data and should remain the foundation of radiation protection.

Results and Discussion
e categorization of the total number of foetuses exposed in utero into various gestational age (weeks) intervals is shown in Figure 1. Figure 2 shows the categorization of the total number of foetus exposed in utero into various absorbed dose intervals. Table 1 shows a summary of the maternal and foetal parameters. Estimated mean and range of entrance surface doses to female adult patients from diagnostic radiology examinations is shown in Table 2. Tables 3 and 4  summary of the mean and range of foetal absorbed dose per examination from some simple and special types of diagnostic radiology examinations, respectively. e foetal radiation risks of childhood cancer and hereditary effects per examination are shown in Table 5. Table 6 also shows the risks of childhood cancer and hereditary effects for the 367 foetuses investigated in this work.

Foetal Radiation Dose.
In this study, exposure parameters from 367 pregnant patients from �ve institutions in �ve counties were investigated to estimate the radiation dose and risk to the foetuses. e radiation doses received by the 367 foetuses ranges from less than 0.001 mGy from chest X-ray examination of the pregnant mother to 21.9 mGy from CT of lumbar spine examination (Tables 3 and 4). Some of the women had multiple examinations (2 to 4 different examinations performed on the same patient) and others had repeat examinations (2 to 3 repeated examinations on the same patient), thus contributing to increased foetal doses. According to Osei and Faulkner [11], foetal doses are in�uenced by the maternal size and foetal depth and size. As a consequence, simply looking up the average uterus dose for a speci�c examination may lead to underestimation or over estimation of the foetal dose. Foetal dose should therefore be estimated based on the actual technique parameters used for the examination and taking into account the foetal depth and size. However, if the AP thickness of the patient and the foetal depth are not known, an average value of 25 cm and 9 cm, respectively, are good approximations to use [9]. Data in Tables 3 and 4 shows that, for the same examination, the accrual foetal dose can vary signi�cantly from hospital to hospital and from patient to patient depending on several factors including the imaging equipment, technique parameters, patient size, and foetal depth and size. About 78% (285) of the foetus investigated received doses less than 1 mGy, and about 4% (16) of them received radiation doses greater than 10 mGy.

Risks of Death, Malformation, and Mental Retardation.
e principal deterministic effects of exposure of the developing embryo or foetus to ionising radiation are death, malformation, growth retardation, and mental impairment. e ICRP [1] has reviewed the risks of harmful tissue reactions and malformations aer prenatal irradiation and concluded that no deterministic effects of practical signi�cance would be expected to occur in humans below a dose of at least 100 mGy. e threshold dose [1,6] for the induction of these deterministic effects following in utero exposure of the 367 foetuses all lie very well above the estimated foetal doses, of 21.9 mGy in this study will carry a risk of hereditary effect of about 1 in 10000 (1 × 10 −4 ). e risks values are very low compared to the natural incidence. e natural frequency of genetic disease manifested at birth in human populations has been estimated to be in the range 1 in 100 (1 × 10 −2 ) to about 1 in 17 (6 × 10 −2 ) if minor congenital abnormalities are included [6,11,20]. us, the increased genetic risk of 1 in 10000 (1 × 10 −4 ) for an individual foetus associated with the highest dose in this study is very small compared with the natural risk of genetic disease.

Risk of Childhood
Cancer. e risk of childhood cancer resulting from exposure of the embryo or foetus is considered to be directly proportional to the radiation dose and fairly independent of the stage of pregnancy aer the �rst three to four weeks of gestation [20]. On the basis of the doses received by the foetuses, the risk of childhood cancer ranges from less than 1 in 12500000 (8 × 10 −8 ) to about 1 in 500 (2 × 10 −3 ). e risk of 1 in 500 (2 × 10 −3 ) associated with the maximum absorbed dose of 21.9 mGy is very comparable to the natural baseline risk of childhood cancer (1 in 500 (2 × 10 −3 )) in the UK [20]. Foetal doses up to about 50 mGy ( Table 6) could result in an approximate doubling of the natural baseline risk of childhood cancer. erefore, those examinations resulting in foetal doses of some tens of mGy during pregnancy should be avoided unless the health of the mother (and indirectly that of the unborn child) would be compromised by delaying the examination until aer the birth of the baby [20].

Conclusion
Exposure of the embryo or foetus to high dose ionising radiation can potentially lead to some adverse health effects. However, the radiation dose to the embryo or foetus that is likely to result from any diagnostic procedures should present no risk of causing any deterministic effects such as foetal death, malformation, growth retardation, or mental retardation. Data from this study indicate that foetal doses from properly conducted diagnostic radiology examination will not result in any deterministic effects. Furthermore, the foetal doses presented in this study shows a negligible risk of causing radiation induced hereditary effect in the descendants of the unborn child compared to the natural frequency of genetic disease manifested at birth in human populations. Despite that, all radiological examinations should be clinically justi�ed and the foetal dose kept to a minimum consistent with the diagnostic requirements. It should be noted that doses from some diagnostic radiology examinations may present a very low risk of childhood cancer compared to the natural baseline risk but are certainly not sufficient to justify termination of pregnancy taking into account the health of the mother. It is therefore recommended that [20], any radiological examinations that can result in relatively high foetal doses of some tens of mGy should be avoided on pregnant women, if that can be achieved without serious detrimental effects to their health.