Effective dose conversion factors in paediatric ... - BIR Publications

The British Journal of Radiology, 82 (2009), 748–755

Effective dose conversion factors in paediatric interventional cardiology 1

A KARAMBATSAKIDOU, 1 A FRANSSON, PhD

MSc,

2

B SAHLGREN,

MD, PhD,

3

B HANSSON,

MSc,

4

M LIDEGRAN,

MD, PhD

and

Departments of 1Medical Physics and 2Paediatric Cardiology, Karolinska University Hospital, S-171 76 Stockholm, Department of Hospital Physics, Danderyd Hospital, S-182 88 Stockholm, and 4Department of Paediatric Radiology, Karolinska University Hospital, S-171 76 Stockholm, Sweden 3

ABSTRACT. Conversion factors for effective dose (CFE 5 effective dose/dose–area product (mSv (Gy cm2)21) in paediatric interventional cardiology were estimated retrospectively for 249 patients using the dose–area product (DAP), irradiation geometry, exposure parameters and tissue-weighting factors (TWFs) from the International Commission on Radiological Protection (ICRP) 60. Two methods for estimating the conversion factors, which differed in the description of the irradiation geometry, were evaluated. The effective doses obtained with the two methods were almost identical. The results showed that irradiation geometry had no significant impact on the CFE, and a single factor was defined for both diagnostic and interventional examinations. In addition, the effect of the new tissue-weighting factor for breast tissue (TWFb) given in ICRP 103 on the effective dose was assessed. The CFE was 3.7¡0.2 mSv (Gy cm2)21 (neonate), 1.9¡0.2 mSv (Gy cm2)21 (1 year), 1.0¡0.1 mSv (Gy cm2)21 (5 years), 0.6¡0.1 mSv (Gy cm2)21 (10 years) and 0.4¡0.1 mSv (Gy cm2)21 (15 years). Applying these CFs to the individual DAP values of each patient yielded mean effective doses of 13.0 mSv (neonate), 8.6 mSv (1 year), 6.4 mSv (5 years), 8.6 mSv (10 years) and 12.7 mSv (15 years). The maximum estimated skin dose (15 patients) did not exceed 60 mGy. With the new ICRP value for TWFb, increases in the CFs in the order of 10–30%, and in the effective dose of 10–20%, were indicated. The results indicated that the effective dose in paediatric interventional cardiology is of much greater concern than the skin dose. Furthermore, age-dependent CFE values are required so as not to underestimate the doses to very young patients.

As the number of interventional radiological procedures increases, the need for accurate methods to determine organ and effective doses also increases [1]. It is well known that adult patients undergoing cardiac interventional radiological procedures could suffer skin injuries as a consequence of the intervention [2–4]. In children, the main concern is instead the increased risk of cancer induction [5]. As the radiation sensitivity in children is considered significantly higher than in adults, radiological interventions in children should always be performed using techniques that minimise the radiation exposure. It is well known that the maximum skin dose is highly dependent on radiation geometry and on the total dose–area product (DAP) [6–8]. The influence of these and other factors on the effective dose in children undergoing intervention is, however, not yet well established. This is most likely related to a lack of standardised techniques reflecting a multitude of paediatric congenital heart diseases, as well as large variations in body shape with age in this patient group.

Address correspondence to: A Karambatsakidou, Department of Medical Physics, Karolinska University Hospital, S-171 76 Stockholm, Sweden. E-mail: angeliki.karambatsakidou@karolinska. se

748

Received 19 May 2008 Revised 21 October 2008 Accepted 24 October 2008 DOI: 10.1259/bjr/57217783 ’ 2009 The British Institute of Radiology

Despite these difficulties, several studies have been performed to estimate the effective dose in children undergoing radiological interventions. The techniques used in these studies include thermoluminescent dosimeter (TLD) measurements on patients [9], phantom studies with TLDs [10], calculations applying conversion factors (effective dose/DAP) from the National Radiological Protection Board [11] and Monte Carlo simulations [12, 13]. Common to all these methods is their limited applicability to dose estimations in routine work, owing to the wide range of patient age and size and irradiation geometry encountered in clinical practice. The aim of this study was to establish conversion factors for effective dose (CFE 5 effective dose/DAP (mSv (Gy cm2)21)) in paediatric interventional cardiology, and to evaluate the influence of radiation geometry and age on these factors. The results reported are based on a total of 249 paediatric patient examinations performed on the same interventional X-ray equipment during a 6-year period. In addition, the maximum estimated skin dose (MESD) and typical effective doses from such procedures are reported. The effective doses are based on tissue-weighting factors (TWFs) specified in International Commission on Radiological Protection (ICRP) Report 60 [14]. Owing to the recent update in TWFs by ICRP in 2007 [15], an The British Journal of Radiology, September 2009

Effective doses in paediatric interventional cardiology

attempt to evaluate the impact of the new set of factors on the effective dose is also reported.

Methods and materials X-ray equipment At the Karolinska University Hospital, Solna, paediatric cardiac angiographic and interventional procedures are performed on a biplane Philips Integris H 5000C Xray unit (Philips, Eindhoven, the Netherlands). This digital cardiac imaging system has an integrated DAP meter (Diamentor; PTW-Freiburg, Germany) mounted on each X-ray tube housing. The DAP value that is reported corresponds to the sum of the exposures using the frontal and lateral planes. The integrated DAP meters are checked once a year against a reference DAP meter (Doseguard 100 (VacuTec 70157); RTI Electronics AB, Molndal, Sweden) calibrated at the secondary standard dosimetry laboratory in Sweden. The image intensifiers (IIs) have three magnification modes — 23 cm, 17 cm and 14 cm — and are equipped with antiscatter grids with a ratio of 10:1. For the three different II modes (14 cm, 17 cm and 23 cm), the air kerma rates at the entrance window of the II (without grid) are 0.40 mGy per frame, 0.28 mGy per frame and 0.14 mGy per frame, with digital image acquisition, and 0.39 mGy s–1, 0.30 mGy s–1 and 0.22 mGy s–1, respectively, in fluoroscopy mode. The unit uses automatic exposure control, which includes programmes that automatically select beam quality parameters based on the type of examination and patient size. Typical values for the tube peak voltage in paediatric examinations are between 60 kV and 85 kV. The angiographic X-ray unit is equipped with a spectral filter on the tube side, consisting of a sandwich of copper (Cu) and aluminium (Al). The total filtration of the tube for the majority of cardiology examinations is in digital image acquisition mode: 6 mm Al + 0.4 mm Cu (,4 years) or 6 mm Al + 0.2 mm Cu (.4 years and for slow procedures). In fluoroscopy mode, the total filtration was 6 mm Al + 0.4 mm Cu. The programme chosen for cardiology procedures uses a digital image acquisition of 25 or 50 frames per second (,4 years), 25 frames per second (.4 years) and 12.5 frames per second (slow procedures), and a matrix size of 1024 6 1024.

Estimation of conversion factors for effective dose Effective dose CFs for 249 paediatric patients between 0 years and 18 years old who underwent cardiac radiological examinations between 1999 and 2004 at the Karolinska University Hospital, Solna, were determined retrospectively. The cardiac procedures were divided

into interventional (including a diagnostic part; 110 patients) and pure diagnostic procedures (139 patients). The interventional procedures were transcatheter closure of the patent ductus arteriosus (PDA) or atrial septal defect (ASD) with a device, balloon dilatation of pulmonary valvar stenosis, peripheral pulmonary stenosis or aortic coarctation and, in a few cases, closure of collateral arteries or fistulas with coils. The diagnostic examinations were almost exclusively pre-operative evaluation of complex congenital cardiac anomalies such as tetralogy of Fallot or pulmonary atresia with ventricular septal defects (VSDs) and systemic to pulmonary collaterals (major aortopulmonary collateral arteries; MAPCAs) or univentricular heart before palliative surgery with bidirectional or total cavopulmonary connection (TCPC). Examination reports containing information on cine and fluoroscopy data acquisitions were retrieved for all patients. The patients were divided into five age groups (neonate (‘‘0’’), 1 year, 5 years, 10 years and 15 years old). The entrance radiation field size used during these procedures varies with age. For the purpose of this study, the field size was assumed to be constant for all patients in each age group (see Table 1). Two different methods (A and B) were used to estimate effective dose CFs. The fundamental difference between the methods is in the definition of the irradiation geometry.

Method A Conversion factors for all 249 paediatric patients were estimated using data published by Schmidt et al [12]. They reported CFs (effective dose/DAP value) for a wide range of projection angles and for children of age 0, 1 year, 5 years, 10 years and 15 years, assuming a fixed isocentre of 70 cm and a radiation quality of 65 kVp and 3 mm Al beam filtration. For radiation qualities differing from 65 kVp and 3 mm Al filtration, a set of correction factors for tube peak voltage and total beam filtration were reported. The starting point for using their method in our study was to calculate a single CF for effective dose for each acquisition plane (frontal, lateral). This was done by calculating the mean value of the CFs for the most frequent projection angles used at our hospital, and for a radiation quality of 65 kVp and 3 mm Al filtration (Table 2). The projections included were for the frontal plane — right anterior oblique (45 ˚, 30 ˚, 15 ˚), anteroposterior, left anterior oblique (15 ˚, 30 ˚, 45 ˚) (tilt: cranial 15 ˚, 0 ˚; caudal 15 ˚, 45 ˚) — and for the lateral plane — lateral (tilt 0 ˚). As the tube voltage and total filtration vary between patients, the technique provided by Schmidt et al [12] to adjust the CFs for tube voltage and total beam filtration was applied. Firstly, the CF (Table 2) for the

Table 1. Entrance radiation field size for each age interval Age group (years)

Age interval (years)

Entrance radiation field (cm 6 cm)

0 1 5 10 15

0–0.5 0.51–2.5 2.51–7.5 7.51–12.5 12.51–18.0

767 868 9.569.5 11611 12612

The British Journal of Radiology, September 2009

749

A Karambatsakidou, B Sahlgren, B Hansson et al Table 2. Mean value of DAP to effective dose conversion factor (mSv (Gy cm2)21; Method A) for various ages and most frequent radiation angles (frontal X-ray tube: RAO (45 ˚, 30 ˚, 15 ˚), AP, LAO (15 ˚, 30 ˚, 45 ˚) (tilt: cranial 15 ˚, 0 ˚, caudal (15 ˚, 45 ˚)); and for lateral X-ray tube: LAT (tilt 0 ˚) Age (years)

Conversion factor for lateral X-ray tube (mSv (Gy cm2)21)

Conversion factor for frontal X-ray tube (mSv (Gy cm2)21)

0 1 5 10 15 Adult

2.34 1.16 0.64 0.38 0.22 0.16

1.97 0.87 0.45 0.27 0.16 0.12

AP, anteroposterior; DAP, dose–area product; LAO, left anterior oblique; LAT, lateral; RAO, right anterior oblique.

frontal plane was divided into fluoroscopy and cine fluorography as, with the equipment used at our hospital, different filtrations are used with the different modes. It was further assumed that all fluoroscopy had been performed using only the frontal plane. The median tube voltage value using this plane was retrieved from the examination report for each patient. With this information, a correction for the tube voltage was performed using the CFs published by Schmidt et al (see Tables 9 and 10 in [12]). The correction for total filtration was performed in a similar manner using the technique reported by Schmidt et al (see Table 12 in [12]) and extending the data to include the total filtration values used at our hospital. These calculations were performed using the software PCXMC (version 1.5; Radiation and Nuclear Safety Authority, Helsinki, Finland) [16], and were cross-checked with the filtration correction data provided by Schmidt et al [12]. The two data sets (from Schmidt et al [12] and ours) were identical for all radiation qualities included in the work by Schmidt et al [12]. The CF for the lateral plane was retrieved using the same procedure. In the next step, the effective doses from the different modes (fluoroscopy and cine fluorography) using the frontal plane were calculated by multiplying the CFs from fluoroscopy and cine fluorography with the corresponding DAP values provided in the examination reports. The effective dose from the lateral plane (only cine fluorography) was estimated in the same way. Finally, a single CF for the entire procedure (frontal and lateral planes) was estimated by adding the effective doses from each plane, and dividing by the total DAP value of the examination.

Method B For 52 of the 249 patients, examination reports containing comprehensive information on exposure parameters for the cine fluorography part (beam projection and cranio/caudal angles, tube voltage, tube filtration) could be retrieved. Combined with the information on patient age and DAP values, these data were fed into the X-ray dose calculation software PCXMC [16], and the effective dose corresponding to each projection angle of the cine fluorography procedure was calculated. For simplicity, for any given projection angle and for each 4 kV interval, a single tube voltage value equal to the mean tube voltage was used. In clinical practice, an isocentric irradiation geometry is used, such that the focus-to-patient distance will vary with patient size. In a previous study, and using 750

anthropomorphic paediatric phantoms, focus-to-patient distances were typically between 50 cm and 60 cm [10]. As in this work, no data on patient size were available and a focus-to-skin distance of 60 cm was assumed for all projection angles and patients. The field sizes used with Method B are reported in Table 1. The fluoroscopy mode data were incomplete in that they did not contain information on projection angle and tube voltage. Based on observations of patient examinations as they are performed today and on statements from staff, a frontal projection angle was assigned to the complete fluoroscopy part of the examination. The tube voltage used in fluoroscopy was set to the median tube voltage used in frontal projections in the cine procedure. Finally, the CF for the complete cardiac procedure was determined by adding the effective doses from all projections, and dividing by the total DAP value. In order to evaluate the impact on the CF of the radiation geometry during fluoroscopy, two different situations were assessed: one using only the frontal plane for fluoroscopy (100% frontal) and the other assuming that fluoroscopy was performed using both planes (50% frontal, 50% lateral). CFs using either geometry were estimated with Method A.

Effective dose For all 249 patients and for both Methods (A and B), the effective doses were estimated by using TWFs from ICRP Report 60 [14]. With the recent update of such values by the ICRP, the impact on the effective dose from irradiation of breast tissue has increased significantly (weighting factor for breast tissue has increased from 0.05 to 0.12). In addition, other TWFs have changed, but these changes can be expected to have only a minor effect on the effective dose from cardiac procedures. To evaluate the importance of the new value for breast tissue on the effective dose, effective doses for the subset of 52 patients were calculated replacing the weighting factor for breast tissue from ICRP Report 60 with the new value in ICRP Report 103 [15]. In addition, an estimate of the effective dose by applying the complete set of new TWFs was attempted for five patients. These calculations were performed using Method B.

Skin dose For the evaluation of maximum skin dose in paediatric cardiac procedures, 15 patients undergoing PDA (5) or The British Journal of Radiology, September 2009


Figure 1. Mean dose–area product (DAP) values from diagnostic and interventional procedures, and the corresponding number of patients in the different age groups. Results for adult patients [6] are included for comparison.

ASD closure (4), pre-operative evaluation of pulmonary atresia and VSD with MAPCA (4) and univentricular heart before TCPC (2) were selected for film dosimetry. These procedures are all high-dose procedures and are also the most common interventional and diagnostic paediatric heart examinations performed at Karolinska University Hospital. In these patients, the MESD was estimated using slow radiographic film (EDR2, 35 cm 6 43 cm; Eastman Kodak Co, Rochester, NY). This method has been applied by our group in a previous study on adults [7], and the same methodology as that used for the adult patients was applied in this work. Based on observations of paediatric examinations performed at this unit, together with information from examination reports, it was concluded that the maximum skin dose was located on the patient’s back. The skin dose from the frontal plane (includes the region of the MESD on the patient’s back) was documented on a film positioned between the patient and the table, and the corresponding MESD was obtained from the maximum film density (Figure 3 in [7]). The film processor used was a Kodak RP X-OMAT Processor, Model M8, and the transmission densitometer was an X-rite 331 (X-rite Co., Grand Rapids, MI).

Statistical analysis The two-tailed Mann–Whitney test was applied to evaluate if there was a statistically significant difference between the results using Method A and those using Method B, between diagnostic and interventional procedures, and between using one (frontal) or two (frontal plus lateral) projections during fluoroscopy. To investigate correlations in scatter plots displaying linear or exponential behaviour, linear and exponential regressions were used, respectively. Linear regression was applied on effective dose vs DAP value for the different age groups. CF vs age for Method A and B and mean CF vs age for diagnostic and interventional procedures both showed exponential behaviour, and exponential regression was used in these cases.

Mean lifetime mortality risk The multiplicative model of ICRP 60 [14] was used to calculate the lifetime probability of death following irradiation. The mortality risk in this model is age- and gender-dependent, and the risk factors vary with age (0– 9 years 5 16%/Sv (girls) and 13%/Sv (boys); 10–19 years 5 9.5%/Sv (girls) and 7.5%/Sv (boys)).

Results

Figure 2. Comparison of the two methods for estimating mean conversion factors (CFs) for effective dose in 52 paediatric patients. The British Journal of Radiology, September 2009

The mean DAP values, including all procedures (diagnostic and interventional) in the different age groups, together with data on the number of patients in each group, are shown in Figure 1. For comparison, data for adult cardiac procedures (diagnostic, 73 Gy cm2; interventional, 120 Gy cm2) taken from a previous study are included [6]. The paediatric DAP values include fluoroscopy and cine fluorography from both the frontal and lateral planes. There is a clear trend for an increase in total DAP with increasing age. There is also a tendency for a slightly higher DAP value for diagnostic procedures (3.7¡2.6 Gy cm2 (neonate; ¡1 SD (standard deviation)), 6.0¡5.8 Gy cm2 (1 year; ¡1 SD), 7.6¡9.5 Gy cm2 (5 years; ¡1 SD), 15.9¡12.9 Gy cm2 (10 years; ¡1 SD), 37.9¡52.3 Gy cm2 (15 years; ¡1 SD)) compared 751

A Karambatsakidou, B Sahlgren, B Hansson et al Table 3. Contribution from fluoroscopy to the total DAP in diagnostic and interventional procedures for the various age groups Age (years)

(DAPF/DAPT)diagnostic (%)

(DAPF/DAPT)interventional (%)

0 1 5 10 15

31 50 64 58 51

49 66 79 76 59

DAPF, dose–area product from fluoroscopy; DAPT, dose–area product from the complete procedure.

with interventional procedures (3.2¡4.1 Gy cm2 (neonate; ¡1 SD), 2.6¡5.1 Gy cm2 (1 year; ¡1 SD), 7.8¡11.8 Gy cm2 (5 years; ¡1 SD), 10.0¡9.7 Gy cm2 (10 years; ¡1 SD), 34.2¡38.9 Gy cm2 (15 years; ¡1 SD)) in the paediatric age groups. The contribution to the total DAP from fluoroscopy is displayed in Table 3. The percentage share of fluoroscopy for diagnostic and interventional procedures is lowest for neonates (31% and 49%), reaches a peak at the age of 5 years (64% and 79%), and finally decreases to values of 51% and 59%, respectively, at the age of 15 years. CFs for effective dose using Methods A and B for the subset of 52 patients are shown in Figure 2. There was no significant difference (a>0.05; Mann–Whitney test) in the results using the two methods, showing a mean deviation of 5%. The difference in CFs using Method A compared with Method B was always less than 10%, except in four patients (216%, 15%, 20% and 25% difference). Detailed inspection of the corresponding patient data revealed that, for three of these patients, a program containing a lower tube filtration than normal had been applied whereas, in the fourth patient, a radiation quality using a 25% higher tube voltage than for patients in the same age group had been used. Recalculation of the CFs taking these differences into account re-established the limit of 10% difference between the two methods. The data could be fitted with an exponential function: (CFE 5 2.7e20.15*age; age 0– 15 years; r50.97 for both Methods A and B).

Based on the very good agreement of the CFs obtained using either of the two methods (A and B) in the subset of 52 patients, the simplest technique (Method A) was applied to the whole patient sample (249 patients) and the results used to evaluate the extent to which CFs from diagnostic and interventional procedures differed. The results showed almost identical factors for both diagnostic and interventional procedures within each age group (a>0.05; Mann–Whitney test; Figure 3), and could be fitted with exponential equations (diagnostic, CFE 5 2.6e20.14*age; interventional, CFE 5 2.6e20.15*age; r50.96, age 0–15 years). The corresponding data for adult patients, as reported in a previous study [6], are included for comparison. Table 4 shows the influence of radiation geometry during the fluoroscopy part of the examination on the CF. There was no significant difference using the two geometries for neonates (a>0.05; Mann–Whitney test), and 1-year-old (a,0.05; Mann–Whitney test) and 15year-old patients (a>0.05; Mann–Whitney test). In the 5 year and 10 year age groups, there was a minor yet significant difference (a,0.01 and a,0.001, respectively; Mann–Whitney test). The MESD for the subset of 15 patients did not exceed 60 mGv. The mean value of MESD for patients undergoing ASD closure (monoplane, 4 patients) was 12 mGy whereas, in patients undergoing PDA closure, TCPC or MAPCA (biplane; 11 patients), the mean MESD was 20 mGv. The relationship between age and effective doses as a function of DAP values is illustrated in Figure 4, showing a linear correlation for all age groups (R250.96–0.99). Table 5 shows the impact of the new TWFb (tissueweighting factor for breast tissue) from ICRP Report 103 [15] on the effective dose. This recalculation (Method B) resulted in an upward adjustment of the mean effective dose by 13% (newborn), 16% (1 year), 14% (5 years), 10% (10 years) and 11% (15 years). In five patients, the effective dose was estimated taking into account all new TWFs from ICRP [15]. There was only a small difference (,5%) compared with results obtained applying only the new TWFb. The effect of the updated TWFb on the contribution to effective dose from the different projections (posteroanterior and lateral) is also reported in Table 5. The effective dose from posteroanterior and lateral projections in the different age groups increased by up to 12% and 32%, respectively. As can be observed from these data, there is a significant increase in the effective dose from the Table 4. Conversion factors obtained using two different radiation geometries for the fluoroscopy part of the procedures (Method A)

Figure 3. Mean conversion factors (CFs) (Method A) for 249 paediatric patients undergoing diagnostic and interventional procedures.

752

Age (years)

CFE1 (mSv (Gy cm2)21) (¡1 SD)

CFE2(mSv (Gy cm2)21) (¡1 SD)

0 1 5 10 15

3.7¡0.2 1.9¡0.2 1.0¡0.1 0.6¡0.1 0.4¡0.1

3.7¡0.4 1.9¡0.1 1.0¡0.1 0.7¡0.1 0.4¡0.1

CFE1, conversion factor from 100% fluoroscopy using only frontal plane; CFE2, conversion factor obtained from using both planes during fluoroscopy (50% frontal and 50% lateral plane); SD, standard deviation.



Figure 4. Effective dose estimated using conversion factors from Method A and individual dose–area product (DAP) values for all 249 patients. A linear fit of the patient data within each age group yielded: 0 year, effective dose 5 3.796DAP–0.19 (R 2 50.99); 1 year, effective dose 5 1.966DAP–0.40 (R 2 50.97); 5 years, effective dose 5 0.986DAP+0.01 (R 250.99); 10 years, effective dose 5 0.566DAP+0.85 (R 250.96); 15 years, effective dose 5 0.286DAP+1.60 (R250.97).

lateral projection using the new TWFb. As the contribution to effective dose from the lateral projection is relatively high, the change in the TWFb therefore also changes the relative impact on effective dose from the posteroanterior vs lateral projections. The estimates of effective dose (Method A, 249 patients), together with the mean lifetime risk factors of death following irradiation as defined by ICRP Report 60 [14], yielded increases in mortality risk of 0.08–0.21% for girls and 0.06–0.17% for boys (using ICRP Report 60 TWFs), with the highest risk estimates corresponding to neonates. The risk factors have not been changed in ICRP Report 103 [15]. Taking into account the change in TWF for breast tissue in the new report, the resulting increase in effective dose is reflected by a corresponding increase in the mortality risk.

Discussion Patients undergoing cardiac radiological procedures can receive radiation doses that exceed threshold levels Table 5. Mean effective doses (Method B; 52 patients) using tissue-weighting factors from ICRP Report 60 (E60), and including the new tissue-weighting factor for breast from ICRP 103 (E103). Ratios (E103/E60)PA,LAT for the contribution to effective dose from the posteroanterior (PA) and lateral (LAT) projections, respectively, are included Age (years)

E60 (mSv)

E103 (mSv)

(E103/ E60)PA

(E103/E60)LAT

0 1 5 10 15

10.42 9.96 8.42 10.08 21.13

11.78 11.55 9.62 11.11 23.42

1.12 1.11 1.11 1.08 1.05

1.18 1.32 1.32 1.24 1.31

ICRP, International Commission on Radiological Protection.


for deterministic effects (skin erythema, >2 Sv). However, in paediatric cardiac radiological examinations, the skin dose is usually well below the 2 Sv limit [13, 17, 18], and in the present study a MESD of 60 mSv was registered for an MAPCA (high-dose diagnostic examination). In the younger patients, the concern is that the effective dose, owing to the smaller patient size, can be significantly greater than in adults. This is clearly demonstrated in Figure 4, which shows clinical data on the dependence of age on the effective dose. Comparing the mean lifetime risk of death related to irradiation [14] (ICRP Report 60, multiplicative model) of newborns (using data from the present study and TWFs from ICRP Report 60) and adults (data taken from the article by Hansson and Karambatsakidou [6]), the increased mortality risk for adults from cardiac procedures was 0.16% (intervention) and 0.08% (diagnostic), whereas the corresponding figures for newborns were 0.21% (girls) and 0.17% (boys) for all procedures. The results presented in this work are based on clinical data, and could be useful as reference data for sites performing radiological procedures in children with congenital heart disease in order to estimate the effective dose from such interventions. As the younger children are expected to have an increased risk for radiationinduced cancer [5, 14, 19], it is of the utmost importance to monitor the doses to these patients. This becomes especially important in patients who undergo multiple radiological procedures. Paediatric cardiac patients represent a relatively inhomogeneous group in the sense that there exist many different types of congenital heart disease in children. Consequently, the radiological interventions do not, in general, follow a standardised scheme. This contributes to the difficulty in comparing DAP values from this study with values published by other groups. An attempt to do so indicates that the total DAP values reported for paediatric cardiac patients by other groups are almost double those reported in this work [20, 21]. The significance of this fact is, however, doubtful in the light of the variations in age and in the type and severity of disease that these different data represent. In addition, differences in radiation quality between different radiological equipment will influence the total DAP (or rather the DAP rate) required to produce images of sufficient quality. The mean percentage share of fluoroscopy for all 249 paediatric patients in this study was close to 60%, which is similar to the results presented by Schueler et al [20]. For comparison, a 20–70% contribution to the mean DAP value from fluoroscopy has been reported in cardiac procedures on adult patients [6, 7, 22–24]. The CFs for effective dose were almost identical using Methods A and B, as evaluated for a subset of 52 patients (Figure 2). The two methods differ in the description of the irradiation geometry, with Method A being less exact than Method B. The results show that the dependence of the irradiation geometry (number of projections) on the effective dose is of minor importance, and that a relatively rough estimation of the geometry could allow for a calculation of effective dose from these types of procedures on children. This conclusion is furthermore supported by the results presented in Table 4 on the effect of irradiation geometry during fluoroscopy on CFE 753

A Karambatsakidou, B Sahlgren, B Hansson et al Table 6. Comparison of conversion factors (mSv (Gy cm2)21) obtained in this study with previously published values Study

CFE0 (mSv (Gy cm2)21)

CFE1 (mSv (Gy cm2)21) CFE5 (mSv (Gy cm2)21) CFE10 (mSv (Gy cm2)21)

CFE15 (mSv (Gy cm2)21)

This study

3.7 PA: 3.65 LAT: 3.74 PA: 1.03–1.18 LAT: 2.14–2.48

1.9 PA: 1.80 LAT: 1.97 PA: 0.47 LAT: 1.02

1.0 PA: 0.94 LAT: 0.98 PA: 0.35 LAT: 0.71

0.6 PA: 0.62 LAT: 0.66 PA: 0.26 LAT: 0.53

0.4 PA: 0.33 LAT: 0.34 PA: 0.16 LAT: 0.32

PA: 1.8 LAT: 1.4 PA: 0.52–1.18 LAT: 0.70–1.16 1.59b/1.31c

PA: 0.9 LAT: 0.7 PA: 0.26–0.64 LAT: 0.35–0.64 1.59b/1.31c 0.35

PA: 0.17–0.38 LAT: 0.18–0.38 1.59b/1.31c

PA: 0.07–0.21 LAT: 0.09–0.22

Rassow et al [11]

Axelsson et al [10] Schmidt et al [12] PA: 1.36–2.49 LAT: 1.71–2.39 Bacher et al [13]a 1.59b/1.31c Schultz et al [25] 1.67 0

1

0.13

CFE , conversion factor for age group 0; CFE , conversion factor for age group 1; CFE , conversion factor for age group 5; CFE10, conversion factor for age group 10; CFE15, conversion factor for age group 15; PA, posteroanterior; LAT, lateral. a Children ,10 years. b Tube filtration 5 1.5 mm Al + 0.2 mm Cu. c Tube filtration 5 1.5 mm Al + 0.4 mm Cu.

and by results from phantom studies reported by Axelsson et al [10]. Statistical evaluation (Mann– Whitney U test) of the significance of the irradiation geometry during fluoroscopy (data as in Table 4) was negative for age groups 0, 1 year and 15 years, whereas there was a weak correlation for the 5 year and 10 year age groups. There is no clear explanation for this, although weak irradiation geometry dependence of the fluoroscopy part on the effective dose has been found for these two age groups. In addition, there was no significant difference in the CF between interventional (including a diagnostic part) and purely diagnostic procedures for a given age group. In conclusion, these results indicate that a single CF for effective dose can be applied to each age group, independent of the type of procedure. The CFs estimated in the present work have been compared with factors published by other groups (see Table 6). Except for the results given by Bacher et al [13] (no age grouping), all studies conclude that CFs for effective dose decrease with age [10, 12, 20, 25]. It should be noted that the CFE values from the present study are the highest of all. This is believed to reflect differences in radiation quality used at different hospitals/equipments, indicating that the higher tube filtration and tube voltage used at our hospital yields a higher effective dose per DAP. Conversely, owing to the lower total DAP values resulting from the use of a more penetrating beam, the effective doses in this study are lower than those reported by Schultz et al [25], except in the newborn for which the values in our study are approximately 20% higher. A similar trend with relatively higher effective dose estimates for newborns was given by Rassow et al [11] (approximately 30% higher). The high value of CFE for neonates most likely reflects the difficulties in collimating properly the radiation field, with the result that larger body regions will be irradiated in very young children. This has also been shown by Rassow et al [11] and Li et al [26]. The TWFs from ICRP Report 60 were used throughout the whole work. In addition, the impact of the new TWFb on the effective dose from paediatric cardiac procedures was assessed. These recalculations showed, as expected, a relative increase in the contribution to effective dose from the lateral plane compared with the frontal plane. There are most likely to be several factors contributing to 754

5

such a change, with the most important possibly being a larger contribution to effective dose from irradiation of breast tissue in the lateral projection than in the posteroanterior projection. The results show that CFs based on previous ICRP TWFs need to be adjusted upwards, and initial calculations indicate an adjustment up to +12% for posteroanterior projections and +32% for lateral projections (see Table 5). A detailed analysis of the impact of the new set of TWF values on effective dose for cardiac procedures is ongoing and will be presented in a future publication.

Conclusions A technique to define CFs for the effective dose to paediatric cardiac patients has been presented. Using ICRP Report 60 TWF, it has been shown that a single CF can be used within a given age group, independent of the type of procedure. The technique was applied to 249 paediatric examinations, yielding effective doses in the range of 0.2–77.2 mSv. Initial results applying updated TWFs from ICRP Report 103 indicate an increase in the CFs of approximately 10% for posteroanterior projections and 30% for lateral projections, and an increase in the effective dose of 10–20%, depending on age. In sites in which similar beam qualities as those presented in our study are used, the CFs reported here could be applied directly to radiological procedures in children with congenital heart disease in order to estimate the effective dose from such interventions.

References 1. ICRU. Patient dosimetry for X rays used in medical imaging. Report 74. Oxford, UK: Oxford University Press, 2005. 2. Barnea Y, Amir A, Sharif R, Weiss J, Gur E. Chronic radiodermatitis injury after cardiac catheterization. Ann Plast Surg 2002;49:668–72. 3. Dehen L, Vilmer C, Humiliere C, Cordos T, Pentousis D, Ollivaud L, et al. Chronic radiodermatitis following cardiac catheterisation: a report of two cases and a brief review of the literature. Heart 1999;81:308–12. 4. Aerts A, Decraene T, van den Oord JJ, Dens J, Janssens S, Guelinckx P, et al. Chronic radiodermatitis following percutaneous coronary interventions: a report of two cases. J Eur Acad Dermatol Venereol 2003;17:340–3.


Effective doses in paediatric interventional cardiology 5. ICRP 85. Avoidance of Radiation Injuries from Medical Interventional Procedures. Oxford, UK: Pergamon Press, 2001. 6. Hansson B, Karambatsakidou A. Relationships between entrance skin dose, effective dose and dose area product for patients in diagnostic and interventional cardiac procedures. Radiat Prot Dosim 2000;90:141–4. 7. Karambatsakidou A, Tornvall P, Saleh N, Chouliaras T, Lo¨fberg P-O, Fransson A. Skin dose alarm levels in cardiac angiography procedures: is a single DAP value sufficient? Br J Radiol 2005;78:803–9. 8. Van de Putte S, Verhaegen F, Taeymans Y, Thierens H. Correlation of patient skin doses in cardiac interventional radiology with dose-area product. Br J Radiol 2000;73:504–13. 9. Faulkner K, Love HG, Sweeney JK, Bardsley RA. Radiation doses and somatic risk to patients during cardiac radiological procedures. Br J Radiol 1986;59:359–63. 10. Axelsson B, Khalil C, Lidegran M, Schuwert P, Mortensson W. Estimating the effective dose to children undergoing heart investigations-a phantom study. Br J Radiol 1999;72:378–83. 11. Rassow J, Schmaltz AA, Hentrich F, Streffer C. Effective doses to patients from pediatric cardiac catheterization. Br J Radiol 2000;73:172–83. 12. Schmidt PWE, Dance DR, Skinner CL, Castellano Smith IA, McNeill JG. Conversion factors for the estimation of effective dose in paediatric cardiac angiography. Phys Med Biol 2000;45:3095–107. 13. Bacher K, Bogaert E, Lapere R, De Wolf D, Thierens H. Patient-specific dose and radiation risk estimation in pediatric cardiac catheterization. Circulation 2005;111: 83–9. 14. ICRP. Recommendations of the International Commission on Radiological Protection. Publication 60. Oxford, UK: Pergamon Press; 1991. 15. ICRP 103. The 2007 Recommendations of the International Commission on Radiological Protection. Elsevier Ltd, 2007.


16. PCXMC v1.5, user’s guide. Available from: http://www. stuk.fi/sateilyn_kaytto/ohjelmat/PCXMC/en_GB/pcxmc/ _files/12222632510018057/default/pcxmc15_usersguide.pdf 17. Boothroyd A, McDonald E, Moores BM, Sluming V, Carty H. Radiation exposure to children during cardiac catheterization. Br J Radiol 1997;70:180–5. 18. Moore JD, Shim D, Sweet JL, Arheart KH. Beekman III R. Radiation exposure to children during coil occlusion of the patent ductus arteriosus. Catheter Cardiovasc Interv 1999;47:449–54. 19. Darby SC. Epidemiological evaluation of radiation risk using population exposed at high doses. Health Phys 1986;51:269–81. 20. Schueler BA, Jubrud PR, Gray JE, Stears JG, Wu KY. Radiation exposure and efficacy of exposure-reduction techniques during cardiac catheterisation in children. AJR Am J Roentgenol 1994;162:173–7. 21. Leibovic SJ, Fellows KE. Patient radiation exposure during pediatric cardiac catheterization. Cardiovasc Intervent Radiol 1983;6:150–3. 22. Tsapaki V, Patsilinakos S, Voudris V, Magginas A, Pavlidis S, Maounis T, et al. Level of patient and operator dose in the largest cardiac centre in Greece. Radiat Prot Dosimetry 2008;129:71–3. 23. Kuon E, Glaser C, Dahm JB. Effective techniques for reduction of radiation dosage to patients undergoing invasive cardiac procedures. Br J Radiol 2003;76:406–13. 24. Livingstone RS, Chandy S, Peace TB, George PV, John B, Pati P. Audit of radiation dose to patients during coronary angiography. Indian J Med Sci 2007;61:83–90. 25. Schultz F W, Geleijns J, Spoelstra FM, Zoetelief J. Monte Carlo calculations for assessment of radiation dose to patients with congenital heart defect and to staff during cardiac catheterizations. Br J Radiol 2003;76:638–47. 26. Li LB, Kai M, Kusama T. Radiation exposure to patients during paediatric cardiac catheterization. Radiat Prot Dosimetry 2001;94:323–7.

755