© A. Boulanouar et al., Published by EDP Sciences, 2025

1 Introduction

In recent years, the increased use of multidetector CT (MDCT) in medical imaging has raised concerns within the scientific community. CT scans account for 71.3% of the medical radiation dose (IRSN, 2014). MDCT exams may elevate the risk of developing cancer in children, who are more fragile and sensitive to radiation than adults (Pearce et al., 2012). In the United States, approximately 500 out of every 600,000 children may develop cancer as a result of CT scans (Brenner et al., 2001). Likewise, a large cohort study in Australia (Mathews et al., 2013) found that undergoing CT scans during childhood can lead to various types of cancer, including leukemia and brain tumors, and increases the overall cancer incidence by 24%.

To minimize the risk of radiation exposure, the International Commission on Radiological Protection (ICRP) established three fundamental principles of radiation protection in its publication N°60: justification, limitation, and optimization (ICRP, 1991). Additionally, in publication N°.73 (ICRP, 1996), ICRP proposed the development of diagnostic reference levels as a tool for optimization of radiation doses in medical imaging. This recommendation was subsequently adopted in France starting in 2004 (Roch et al., 2018).

Determining Diagnostic Reference Levels for children is more complex than for adults due to the wide variety of sizes among paediatric patients. However, many studies worldwide on paediatric CT imaging (Brisse et al., 2009; Roch et al., 2013; Shrimpton et al., 2000; Rehani et al., 2012) have established DRLs based on anatomical regions and age groups, in accordance with the recommendations of the ICRP (Vañó et al., 2017). Furthermore, several experts have defined DRL values for paediatric CT using weight bands (Célier et al., 2020; Watson et al., 2010; Strauss et al., 2017; Vassileva et al., 2015). Moreover, multiple European studies have analyzed children’s exposure as a function of CI in paediatric CT in Italy (Granata et al., 2015), Finland (Järvinen et al., 2015), Switzerland (Wagner et al., 2018), and France (Célier et al., 2020).

In Africa, studies by Joseph Zira (Joseph Zira et al., 2021), Samuel (Samuel et al., 2021), and Tchaou (Tchaou et al., 2016) aimed to establish DRL CI for paediatric head CT scans. In contrast, some Moroccan publications on diagnostic reference levels in paediatric CT focus specifically on the anatomical areas and age ranges of children (Benmessaoud et al., 2020; Semghouli et al., 2022; Benmessaoud et al., 2023; Khajmi et al., 2023). Nevertheless, Morocco had not previously conducted a survey of diagnostic reference levels based on clinical indications in paediatric CT.

This search aims to specify, for the first time, diagnostic reference levels per CI in paediatric CT at the Marrakech University Hospital Centre. We also seek to identify variations in these DRLs and compare them with international standards for each group of children studied.

2 Methodology

The study was conducted between April 2023 and July 2024 at the Med VI University Hospital of Marrakech’s. All CT examinations were performed using a 16-slice Siemens Somatom Emotion CT scanner, which has been in service since 2008 and is equipped with automatic tube current modulation (CARE Dose 4D). This equipment is regularly maintained to ensure it remains up to date. The Ethics and Health Research Committee of the hospital approved the study. Children were divided into three age groups: 1–5 years, 5–10 years, and 10–15 years, in accordance with European guidelines on paediatric diagnostic reference levels (Granata et al., 2019). The investigation included four anatomical regions: head, chest, abdomen-pelvis, and chest-abdomen-pelvis CTs. Fourteen clinical indications were selected, as detailed in Table 1. These indications were favored due to their prevalence in the department. For each examination, we collected various CT acquisition parameters, including the exam indication, age and sex of the child, tube current (mAs), tube voltage (kV), use of contrast medium, number of scans, scan mode and length, and pitch. Using our Picture Archiving and Communication System (PACS), we recorded the CTDI_vol (mGy) and DLP (mGy.cm). The children’s weight was not recorded at the time of the examination, so categorization by weight was not applicable.

2.1 Statistical analysis and testing

Descriptive statistics, including frequency, mean, quartiles, maximum, and minimum values, were computed using IBM SPSS Statistics software (v26). The DRLs for specific age groups of children and CI were assessed based on the median and third quartile of the CTDI_vol and DLP values (Vañó et al., 2017).

We employed the non-parametric Kruskal-Wallis test to assess variations in Diagnostic Reference Levels between different indications within the same anatomical region for specific age groups. The choice of this test was guided by the non-normal distribution of the CTDI_vol and DLP values, making the Kruskal-Wallis test appropriate for comparing the medians of these variables among several independent groups. All analyses were conducted with a significance level set below 0.01. Subsequently, we compared the DRLs by indication for each age group, including the 75th percentiles of CTDI_vol and DLP, against those reported by other researchers in Europe and Africa (Järvinen et al., 2015; Granata et al., 2015; Wagner et al., 2018; Célier et al., 2020; Joseph Zira et al., 2021). The dose distributions for the 25^th, 50^th, and 75^th percentiles of CTDI_vol (mGy) and DLP (mGy.cm) as a function of age and CI for the head CT scans studied are illustrated in Figures 1 and 2.

Fig. 1 Distribution of CTDIvol (mGy) for paediatric head CT by age group and CI. Data are presented as the minimal, the 25th, 50th, 75th percentiles, and maximal values.

Fig. 2 Distribution of DLP (mGy.cm) for paediatric head CT by age group and CI. Data are presented as the minimal, 25th, 50th, 75th percentiles, and maximal values.

3 Results and discussions

Radiation dose data from 1730 paediatric CT exams involving children were gathered. Among the participants, 60% (n = 1038) were male and 40% (n = 692) were female. The age distribution was as follows: 34.3% of the children were between one and five years old, 32% were aged five to ten years, and 31.4% were between ten and fifteen years old. Head CT exams were the most common, accounting for 42.2% of the total, with trauma and hemorrhage being the primary indications at 12%. Other types of scans included chest CT scans (19.5%), CAP CT scans (19.8%), and abdomen-pelvis CT scans (18.5%). The main clinical indications for paediatric CT scans of the chest, CAP, and abdomen-pelvis included infectious, inflammatory, malformation, and tumor conditions, each representing an average of 6.4% of the total indications.

Overall, there is a noticeable difference in CT acquisition parameters depending on the indications reviewed, as shown in Tables 2 and 3. For all the CI examined, the tube voltage was set at 110 kV for children under 5 years and ranged from 110 to 130 kV for children over 5 years. We used intensity modulation at 83% for head CT scans and at 100% for chest, CAP, and abdomen-pelvis CT scans in order to adjust the mAs load based on the child’s body size. We prefer to perform 94% of head CT scans and all other tests in helical mode for CT acquisition.

Moreover, within each age group, there are variations in practices regarding the number of series, scan lengths based on indications, and the timing of contrast product injections for most protocols. The mean value of series per scan differs depending on the reason for the exam and specific circumstances. It is approximately 1 series for trauma, hydrocephalus, and convulsive crisis indications; between 1 and 2 series for bronchiectasis/bronchitis and abdominal pain; between 2 and 3 series for brain tumors, tumor control, cellulitis, and pneumonia; and between 3 and 4 series for other indications. The slice thickness is consistently set at 5 mm for all indications; however, the pitch increases throughout the series both with and without contrast injection to minimize examination time and ensure proper vessel opacification at the appropriate times. Tables 2–5 provides detailed information on the median values and DRL for CT exams based on age and clinical indication.

The third-quartile values of the CTDI_vol (mGy) for the five CI of head studies were as follows: 22 mGy to 26 mGy, 20 mGy to 29 mGy, 22 mGy to 38 mGy, 22 mGy to 38 mGy, and 21 mGy to 38 mGy, respectively. These measurements correspond to three groups of children who were examined. A similar trend was observed in chest CT studies. The CTDI_vol values for these chest scans varied from 3.7 mGy to 5.1 mGy, 3.8 mGy to 4.8 mGy, and 3.8 mGy to 5.8 mGy. In the case of abdomen-pelvis CI studies, the CTDI_vol values ranged from 5.1 mGy to 7.3 mGy, 4.6 mGy to 7.7 mGy, and 4.9 mGy to 8.1 mGy, respectively. For CAP CI examinations, CTDI_vol values varied from 5.4 mGy to 7.8 mGy, 4.7 mGy to 7.1 mGy, and 4.9 mGy to 7.6 mGy, respectively. It is important to note that children aged 10 to 15 years exhibited higher CTDI_vol values due to the increased voltage and intensity used for this age group. Significant differences in the distributions of CTDI_vol were observed among the five CI of head CT scans across all age groups of children (p < 0.01). However, for chest scans, CAP examinations, and abdomen-pelvis studies, no significant variations in CTDI_vol based on CI were found (p > 0.01).

Tables 2–5 present our estimated 75th percentiles for DRL CI for DLP measured in mGy.cm. For head CT indications, the DRL values range from 1321 mGy.cm to 1708 mGy.cm, 849 mGy.cm to 1306 mGy.cm, 533 mGy.cm to 875 mGy.cm, 486 mGy.cm to 808 mGy.cm, and 498 mGy.cm to 991 mGy.cm, respectively. In the case of chest CT indications, the DRL values vary from 119 mGy.cm to 238 mGy.cm, 193 mGy.cm to 351 mGy.cm, and 221 mGy.cm to 656 mGy.cm, respectively. For abdomen-pelvis exams, the DRL CI ranges from 316 mGy.cm to 623 mGy.cm, 586 mGy.cm to 1167 mGy.cm, and 788 mGy.cm to 1654 mGy.cm, respectively. The CAP CI tests indicated that the DRL results varied from 966 mGy.cm to 1738 mGy.cm, from 552 mGy.cm to 1417 mGy.cm, and from 471 mGy.cm to 1106 mGy.cm, respectively. Notably, we observed substantial variations (p < 0.01) in the distribution of DRL among the five CI categories for head CT. This analysis revealed that the DRL CI values for congenital malformations, adrenal neuroblastoma, and tumor extension were significantly higher (p < 0.01) compared to those for pneumonia, bronchiectasis/bronchitis, ileus, and abdominal pain across all groups of children examined.

4 Discussions

The head CT was the most frequently examined anatomical region in the mother and child hospital’s CT unit, and head trauma was its main indication. This result is in line with previous studies by Joseph Zira (Joseph Zira et al., 2021) in Nigeria, Järvinen (Järvinen et al., 2015) in Finland, Granata (Granata et al., 2015) in Italy, Tchaou (Tchaou et al., 2016) in Togo, and Samuel (Samuel et al., 2021) in Cameroon. The 10-15 years-old age group of children received the maximum dose for all clinical indications established, as highlighted by Tchaou (Tchaou et al., 2016). The findings from Joseph Zira (Joseph Zira et al., 2021) in Nigeria, Wagner (Wagner et al., 2018) in Switzerland, and Järvinen (Järvinen et al., 2015) in Finland align with the results of the Kruskal-Wallis test for head CT indications. These studies showed a significant difference between CTDI_vol and DLP doses for CI in the head CT. Different professionals use distinct protocols and CT technical parameters, which causes the doses to be spread out unevenly. This is mostly seen for the "cellulite and tumor/mass" indications. ln contrast, for the chest, abdominal, and CAP CTs, the findings of the statistical tests show a non­significant difference in CTDIvol (p > 0.0 1) and a statistically significant difference in DLP between clinical CT indications (p < 0.01) for all groups of children. Whereas, Järvinen (Järvinen et al., 2015) observed no significant difference in CTDI_vol and DLP doses between CI in any of the anatomical regions examined.

The differences observed in the diagnostic reference levels, in terms of DLP, between the CI analysed are mainly due to variations in the duration and number of series prescribed for each protocol. They are also due to the repetition of failed series because of the absence of appropriate restraint devices for agitated children and the programming of adult protocols for children. For all the indications examined, the increase in CTDIvol and DLP with age can be attributed to the growth of children and the adjustment of parameters according to age and height.

4.1 Comparison of DRL CI with international standards

The 3th quartiles statistics for the CI in this exercise were compared with similar data from European surveys (Järvinen et al., 2015; Granata et al., 2015; Wagner et al., 2018; Celier et al., 2020) and the African study (Joseph Zira et al., 2021), as presented in Tables 6 and 7.

For head CT scans indicated for trauma, the DRLs of CTDI_vol and DPL are generally lower than those established in Nigeria (Joseph Zira et al., 2021), Switzerland (Wagner et al., 2018), and Italy (Granata et al., 2015) across all paediatric age groups. Additionally, the DRLs of CTDI_vol for other head indications are also lower than those reported in Nigeria (Joseph Zira et al., 2021), except for the age group of 10-15 years concerning cellulitis and hydrocephalus indications. The values observed for the tumour, cellulitis, and hydrocephalus indications are generally lower than those reported in Switzerland (Wagner et al., 2018), with the exception of the 10–15 age group for the cellulitis indication. Furthermore, the DRLs CI of DLP for head were significantly higher than the standards set in Switzerland (Wagner et al., 2018) but lower than the DRLs recorded in Nigeria (Joseph Zira et al., 2021). Except for the 10–15 age group, the DRLs for CTDI_vol related to lung and bronchial infection indications in chest studies exceed those reported in France (Célier et al., 2020) and Italy (Granata et al., 2015). Furthermore, the DRLs for DLP for these indications are significantly higher than the comparable values reported across all age groups in France (Célier et al., 2020) and Italy (Granata et al., 2015). For the 5–10 and 10–15 age groups, the DRLs in this study for CTDI_vol and DLP regarding malformation indications are also greater than the equivalent levels reported in France (Célier et al., 2020).

The DRLs for CTDI_vol in neuroblastoma cases related to abdomen-pelvis investigations are lower than the statistics reported in Italy (Granata et al., 2015) but higher than those observed in France (Célier et al., 2020). However, the DRLs for this indication in terms of DLP remain higher than the standards set for both France and Italy (Granata et al., 2015; Célier et al., 2020). For certain indications, such as CAP CI, abdominal pain, and ileus, a comparison could not be made due to, to the best of our knowledge, the lack of published results.

4.2 Limitations of the study

Due to the lack of standardization in clinical indications and paediatric groups, comparing our results with the current worldwide diagnostic reference levels for computer tomography was challenging. Some groups of children have shown varying DRL CI results in the paediatric CT dosimetry literature. This inconsistency is particularly evident for the two abdominal-pelvic CT indications related to abdominal pain and ileus, as well as the three CT indications for the chest. Additionally, the research conducted in France by Célier (Célier et al., 2020) regarding indications for chest and abdomen-pelvis CT did not include children in the 10 to 15-year age range. One of the study’s intrinsic limitations is that we were unable to account for the weight of the children involved.

5 Conclusion

This monocentric survey established new diagnostic reference levels for paediatric computed tomography examinations based on clinical indications, marking the first instance of such a study in Morocco at the Mohammed VI University Hospital Centre in Marrakech.

The results revealed a significant variation (p < 0.01) in CTDI_vol and DLP among all clinical indications for head CT. In contrast, for chest, abdomen-pelvis, and chest-abdomen-pelvis CT indications, the variations in CTDI_vol were not significant (p > 0.01), while the variations in DLP were significant (p < 0.01). These findings underscore the importance of tailoring CT examination protocols to specific clinical indications.

Acknowledgments

We would like to express our gratitude to the radiologists and radiology technicians of the mother and child hospital of the Mohammed VI University Hospital Centre in Marrakech (Morocco) for their help in data collection.

Funding

This research did not receive any specific funding.

Conflicts of interest

There is no conflict of interest for this paper.

Author contribution statement

A. Boulanouar: Statistics preparation, Methodology, A. Tounsi: Supervision Conceptualization, Visualization, H. Khajmi: Investigation, Writing- Reviewing and Editing. H. Jalal: Investigation, Verification.

Ethics approval

This study was approved by the Ethics Committee of the Mohammed IV UHC in Marrakech in 2024.

Informed consent

This article does not contain any studies involving human subjects.

References

All Tables

All Figures

	Fig. 1 Distribution of CTDIvol (mGy) for paediatric head CT by age group and CI. Data are presented as the minimal, the 25th, 50th, 75th percentiles, and maximal values.
In the text

	Fig. 2 Distribution of DLP (mGy.cm) for paediatric head CT by age group and CI. Data are presented as the minimal, 25th, 50th, 75th percentiles, and maximal values.
In the text