Pediatric Diagnostic Reference Levels for Diagnostic and Therapeutic Cardiac Catheterization in Japan

To propose Japanese national DRLs for air-kerma at the reference point (K a,r ), air-kerma area product (P KA ), uoroscopy time (FT), and number of cine images (CI) for four age groups. We posted a nationwide questionnaire to 132 facilities. Questions focused on identifying the procedure, age, weight, height, K a,r , P KA , FT, and CI during diagnostic and therapeutic pediatric cardiac catheterization. For diagnostic cardiac angiography, the 75th percentile values were as follows; K a,r : 103, 127, 194, and 351 mGy; P KA : 7.0, 12.3, 14.3, and 47.2 Gy.cm 2 ; FT: 36.8, 30.7, 33.4, and 35.7 min; and CI: 2018, 2313, 2408, and 2016 images for less than one year, 1–5 years, 6–10 years, and 11–15 years respectively. For therapeutic cardiac angiography, the 75th percentile values were as follows: K a,r : 146, 209, 130, and 501 mGy; P KA : 7.54, 16.0, 8.35, and 46.0 Gy.cm 2 ; FT: 56.5, 52.0, 49.4, and 52.0 min; and CI: 4075, 4514, 3576, and 5984 images for less than one year, 1–5 years, 6–10 years, and 11–15 years respectively. Our survey of diagnostic and therapeutic cardiac catheterization in Japanese pediatric patients showed that all age-based Japanese 75th percentiles for the K a,r , P KA , FT, and CI were higher than in other surveys. Based on the result of our study, it is necessary to establish DRLs for pediatric cardiac catheterization examinations in Japan, in order to optimize the safety of pediatric protocols for diagnostic and therapeutic cardiac catheterization.


Introduction
Fluoroscopy-guided procedures have been used more frequently for pediatric patients in recent years [1,2]. However, some organs and tissues in children are more sensitive to ionizing radiation and have greater potential to develop malignancies associated with diagnostic medical radiation [3,4]. According to the UNSCEAR report, estimates of lifetime cancer risk for those exposed as children are uncertain and might be a factor of 2-3 times greater than estimates for a population exposed at all age [5]. The "as low as reasonably achievable" (ALARA) concept emphasizes radiation dose reduction while still maintaining acceptable image quality [6]. It is thus necessary to optimize programs for radiation safety, as well as characterizing dose, x-ray image quality, imaging systems and measurement of patient doses related to diagnostic and therapeutic cardiac catheterization in the pediatric population.
One possible solution for optimal determination of diagnostic and therapeutic radiation doses, establishment of a local diagnostic reference level (DRL), is currently under consideration in several countries [7][8][9][10]. A certain level of control is necessary for radiation exposure of patients, therefore, the International Commission on Radiological Protection (ICRP) has recommended the use of DRLs [11]. However, these have not yet been evaluated for pediatric cardiac catheterization in Japan. Furthermore, according to ICRP publication 135, there are no current national DRL values for pediatric interventional radiology or interventional cardiology anywhere in the world [12] .
The purpose of this study was to conduct a nationwide questionnaire survey concerning air-kerma at the reference point (K a,r ), air-kerma area product (P KA ), uoroscopy time (FT), and number of cine images (CI) during diagnostic and therapeutic cardiac catheterization in Japanese pediatric patients, and to propose national DRLs for four pediatric age groups-less than one year, 1-5 years, 6-10 years, and 11-15 years. This research may represent the rst national DRL report published anywhere in the world.

Materials And Methods
This retrospective study was approved by Tsuchiya General Hospital review board; in-formed consent was waived.
And institutional review board approval was not required for this retrospective nationwide questionnaire study. We sent questionnaires to 132 facilities using conventional post and the e-mail system of the Japanese Society of Radiological Technology. We subsequently received responses by post and e-mail. The survey period was from March 2017 to August 2017.
We also searched the NGC (National Guideline Clearinghouse) and NICE (National Institute for Health and Care Excellence) for clinical guidelines and performed exhaustive searches of PubMed and CiNii. In addition, we performed Internet searches for similar information. Keywords used within the search were "Japan", "Japanese", "diagnostic reference level" and "pediatric" (Fig. 1).

Survey items and data analysis
The quantities for the K a,r, P KA , FT, and CI were obtaind for the questionaire. K a,r is expressed in Gy. K a,r is the air kerma at a point in space located at a xed distance from the focal spot (see Patient entrance reference point) cumulated from a whole x-ray procedure. P KA is expressed in Gycm 2 . P KA is the integral of the air kerma free-in-air (i.e. in the absence of backscatter) over the area of the x-ray beam in a plane perpendicular to the beam axis. The accuracy of the displayed value is ±35% [13]. FT and CI were collected from the displayed values on the Angio device.
For each patient, the identi ed procedure, age, weight, height, body mass index (BMI), K a,r , P KA , FT, and CI were registered. Data were extracted from the patient dose reports produced by the dose reporting system at the end of each procedure. Data for all procedures were divided into two groups (diagnostic and therapeutic) and four age groups that ranged from less than one year, 1-5 years, 6-10 years, and 11-15 years, as was done in previous papers [7][8][9]. From these data, we extracted the minimum, 25th percentile, median, 75th percentile, and maximum value using the boxplot method. In addition, we compared the 75th percentile in the diagnostic and treatment groups using Mann-Whitney's U test, with P < 0.05 regarded as representing a signi cant difference. Statistical analyses used the free statistical software EZR, version 1.33 ( http://www.jichi.ac.jp/saitamasct/SaitamaHP. les/statmedEN.html) [14]. We also investigated uoroscopy and exposure rates used in diagnosis and treatment. In addition, we examined characteristics of hospitals, vendors, equipment age, and number of analog and digital units.

Results
Among the 132 facilities to which the questionnaires were sent, 43 facilities (33%) responded. Hospital characteristics were as follows: national hospitals (21 facilities), public hospitals (14 facilities) and others (8 facilities). X-ray imaging system vendors were Philips Medical Systems (116 units), Canon Medical Systems Corporation (111 units), Siemens Medical Systems (105 units) and Shimadzu Corporation (1 unit). The total number of units was 333. And equipment age was as follows: less than 5 years (126 units), 5 to less than ten years (176 units), and10 years or greater (31 units).All imaging units were at panel detectors for a digital system.
At each facility, we collected up to 10 cases. In total, 333 patients were available for data analysis for both diagnostic cardiac catheterization (n=276), at ages less than one year (n=62), 1-5 years (n=137), 6-10 years (n=45) and 11-15 years (n=32); and therapeutic procedures (n = 57) at ages less than one year (n=7), 1-5 years (n=31), 6-10 years (n=11) and 11-15 years (n=8). We also accepted blank K a,r (n=4), P KA (n=63) and CI (n=39) values because some angiographic devices cannot display some of these values. Therefore, blank values were Page 4/10 excluded at the time of analysis. The diagnostic and treatment populations were also treated as separate populations.

Patient demographics
The 75th percentiles for diagnostic angiography are shown in Table 2 for those aged less than one year, 1-5 years, 6-10 years and 11-15 years, respectively. Table 3 shows the minimum, 25th percentile, median, 75th percentile, and maximum values for K a,r , P KA , FT and CI with therapeutic angiography.
The 75th percentiles for therapeutic angiography are shown in Table 3 for those aged less than one year, 1-5 years, 6-10 years, and 11-15 years. The diagnosis showed that K a,r and P KA tended to increase with age increasing, but no treatment was observed. In addition, FT and CI were age-independent in both diagnosis and treatment. Table 4 contains Mann-Whitney U test comparisons of the 75th percentiles for diagnosis and treatment. The diagnostic and therapeutic populations were similarly distributed and comparable. There were no signi cant differences for K a,r and P KA between the diagnostic and therapeutic procedures. However, the FT of therapeutic procedures (55 minutes) was 1.7 times greater than that for diagnostic procedures (32.7 minutes), and the CI of therapeutic procedure (4132 images) was 1.8 times that of diagnostic procedures (2265 images); there were also signi cant differences in the FT and CI on comparison of diagnostic and therapeutic procedures (P<0.05).

Fluoroscopy Rate and Exposure Rate for Diagnostic and Therapeutic Procedures
The most commonly used uoroscopy rate for each institution where diagnostic procedures were performed was 7.5 pulses/sec followed by 15, 10, 6, 6.25 and 3 pulses/sec. For diagnostic procedures, the commonest exposure rate was 30 frames/sec, followed by 15, 60 and 25 frames/sec (Fig. 2a-b).
For therapeutic procedures, the uoroscopy rate at most institutions was 7.5 pulses/sec followed by 15, 10 and 6 pulses/sec. The most commonly used exposure rate during therapeutic interventions was 30 frames/sec, followed by 15, 25, 7.5, 60 and 1 frames/sec (Fig. 2c-d).

Discussion
A great bene t of recent advances in non-invasive examination using modalities like echocardiography, cardiac CT and MRI is that the number of indications for diagnostic cardiac angiography has decreased in comparison with the past. Nevertheless, reliable and detailed measurement of blood ow, vascular resistance and blood pressure still depends on invasive angiography.
According to ICRP publication 135, there are no current national DRL values for pediatric interventional radiology or interventional cardiology. Several countries have tried to establish local pediatric DRL values for interventional procedures [12]. There are however no studies that attempt any form of qualitative synthesis (Fig. 1). Therefore, our study may be the rst instance in the literature of large-scale, national DRL research applied to pediatric interventional cardiology.
The present nationwide survey found that the radiation dose used in pediatric diagnostic and therapeutic cardiac catheterization was much higher than reported in previous literature [7][8][9][10]. As compared with previous reports, radiation dose was remarkably high: from 2 to 4 times greater in those aged 1 to 5 years and 1.2 to 4.5 times greater in those aged 6 to 15 years (Fig. 3, 4). There are several possible reasons for such obvious differences. First, Ubeda's et al., who investigated DRL in Chile and Costa Rica and determined 75th percentile values for diagnostic and interventional cardiac catheterization, based their ndings on results from just a single center in both countries.
In this study, all equipment for Chile, Costa Rica, and Spain was from the same manufacturer. In contrast, our DRL data originate from equipment from various manufacturers. Therefore, the acquisition parameters, SID, frame and pulse rates, and added ltration used by the device differ. We have not performed detailed analyses, but this is one possible cause for the higher DRL in Japan. The present study included 43 facilities from an entire nation in which there might be a wide range of dose settings within each hospital.
Second, in addition to variation in dose settings, results might be in uenced by other factors including a wide variety of angiography units from many vendors, the size of the eld being irradiated, soft line elimination lters, bed height, and source image distance. In other words, while 75th percentile values were 2-3 times higher than in the articles referred to above, our result might fundamentally re ect a national reference level.
Third, Japan is in a unique situation regarding control of radiation exposure, as there are very few medical physicists in the entire country. Radiological technologists thus need to play the role of medical physicists in managing the radiation dose.
Comparing diagnostic and therapeutic interventions, Ubeda's results were similar. In this study there were no signi cant differences for K a,r between the diagnostic and therapeutic procedures.
However, the K a,r (189 mGy) of treatment procedures was 1.3 times the K a,r (142.1 mGy) of diagnostic procedures, and K a,r tended to be higher than the diagnosis.
In addition, it is necessary to consider diagnosis and treatment separately when formulating DRL because signi cant differences are found for FT and CI.
A further observation is that many facilities use the same uoroscopy rates (7.5 pulses/sec) and exposure rates (30 frames/sec) for both diagnostic and therapeutic cardiac catheterizations. Previous reports describe uoroscopy rates and exposure rates of 10-15 pulses/sec and 10-30 frames/sec [15], almost the same as in the present survey.
Among various methods used for exposure dose reduction, decreasing uoroscopy rates and exposure rates is thought to be one of the most effective techniques during cardiac catheterization [16,17]. It is necessary to use uoroscopy rates and exposure rates in a manner that is appropriate to the patient's speci c diagnosis and individual status.
In addition to delays in establishing DRLs for cardiac angiography, national guidelines regarding clinical examinations used for decision making in diagnosis of and drug therapy for pediatric patients with congenital heart disease and cardiovascular disorders have only recently been published [18]. Compared to other countries, acquisition of diagnostic guidelines and national DRLs may be somewhat late; we therefore hope that establishing DRLs for cardiac catheterization will quickly lead to national dose regulation.
Our study has some limitations. First, national DRLs are usually determined by the accumulation of data from many cases. For each age group in the diagnostic cardiac catheterization group, the number of cases was large 26 to 137 and we therefore consider the values to be reliable. However, the number of cases for therapeutic cardiac catheterization (excluding ages 1-5), for which DRL data are available, is small, and may therefore be less reliable.
We intend to examine more cases in the future. More detailed data may also be derived where missing values are present using multiple imputation of values. Second, we did not evaluate individual cases or diseases. Further evaluation will therefore be necessary to take into account differences between diagnostic and interventional radiology (IVR) procedures. Third, kV, mA, pulse duration and frame duration values were not collected, nor were details of any additional ltering. Also, pediatric exposure doses by age were evaluated. However, pediatric DRLs recommend the use of weight categories, rather than age bands [19]. DRLs presented as a function of patient weight rather than weight categories. Therefore, it is necessary to further analyze using weight categories. Fourth, uoroscopy rate and exposure rate are programmable techniques, and may therefore vary across the different protocols programmed into uoroscopic units. Therefore, uoroscopy and exposure rates were not collected.
However, the device manufacturer, inch size, number of uoroscopic pulses, and number of imaging frames were speci ed for each case. Fifth, there are numerous patient factors that in uence the derivation of DRLs in interventional procedures, unlike in general radiography. These patient factors may include variable patient anatomy (e.g., body habitus, vascular tree variations, blood vessel diameters), lesion variability, and factors related to operator and equipment. These factors may all play a role in DRLs. Finally, we did not evaluate the images acquired at cardiac catheterization for vascular enhancement and subjective image quality.

Conclusion
Our survey of diagnostic and therapeutic cardiac catheterization in Japanese pediatric patients showed that all agebased Japanese 75th percentiles (at ages less than one year, 1-5 years, 6-10 years, and 11-15 years) for the K a,r , P KA , FT, and CI were higher than in other surveys. To promote the optimization of diagnostic and therapeutic cardiac catheterization involving pediatric protocols, it is therefore necessary to establish DRLs for pediatric cardiac catheterization examinations in Japan.   Table 3 The minimum, 25th percentile, median, 75th percentile, and maximum values for K a,r , P KA , FT, and CI for therapeutic protocols in four pediatric age groups.