Establishment of Local Diagnostic Reference Levels in Conventional Radiography: A Pilot Study in Dakar, Senegal

Diagnostic Reference Levels (DRLs) are indicators that allow assessing the quality of equipment and procedures from the point of view of the doses delivered to patients and subsequently initiate corrective actions if necessary. The purpose of this study is to encourage health professionals to investigate patient radiation doses and to determine whether those doses comply with the principles of radiation protection in medical fields so as to improve practices by reducing patient exposure without reducing clinical effectiveness. To perform this work, we have investigated patient doses for different radiological examinations from six (6) medical centers in Dakar, including the following nine routine types: chest (PA), abdomen (AP), pelvis (AP), cervical spine (AP), lumbar spine (AP, Lat), hip (AP), thoracic spine (AP, Lat). Three types of data were collected, i.e., X-ray tube machine data, patient data and output measurements. The data were analyzed statistically and the median, minimum, maximum, and third quartile values were calculated and displayed throughout boxplots graphs for all exams and medical centers. The two sigma range (95% confidence interval) was also checked. Comparison of third quartiles of Entrance Surface Dose (ESD) and Dose Area Product (DAP) by type of examination with recommended international DRLs was performed. protection practices. The results show the need to develop protocols for dose measurement as well as to carry out quality assurance programs and dose optimization in Senegal.


Introduction
Diagnostic X-rays represent the major portion of radiation exposure from artificial origin to the population. According to the World Health Organization (WHO), more than 3.6 billion diagnostic radiology exams are performed every year around the world. X-ray examinations are an established tool of medical diagnosis and patients can undoubtedly obtain enormous benefits from these examinations. However, the ionizing nature of the X-rays means that their use is not entirely without risk. The stochastic effect of low doses from ionizing radiation follows a linear model without threshold and at a long term. For this reason, all exposures to diagnostic X-rays need to be justified and optimized in terms of benefit and risk [1]. These effects are not well known, that is why it is important to monitor patient dose.
The International Commission on Radiological Protection (ICRP) suggested general principles of radiation protection: justification, optimization and dose limit. In medical exposure, dose limits are not at all relevant since ionizing radiation, used at the appropriate level of dose for a particular medical purpose, is an essential tool that will cause more good than harm. Therefore medical radiation has no dose limits, and generally uses diagnostic reference level (DRL) as a reference value for optimization of practice [1]. The goal of the optimization process is to provide an acceptable image quality by keeping the corresponding radiation dose As Low As Reasonably Achievable (ALARA) [2]. In 1997, the ICRP introduced DRLs with the goal to reduce the unnecessary radiation exposure by setting given thresholds [3]. DRLs have been defined in European Commission's legislation (EC, 1997) as dose levels in medical radiodiagnostic practices in the case of radiopharmaceuticals levels of activity for typical examination for group of standard patients or standard phantoms for broadly defined type of equipment. The International Atomic Energy Agency (IAEA) through the International Action Plan on Radiation Protection of Patients and the ICRP have for some time carried out important efforts to ensure that in the medical applications of the ionising radiations, the optimisation of radiological protection of patients is fundamental, to such a point that they include it directly as a requirement for these practices (in the International Basic Safety Standards for Protection against Ionising Radiation and for the Safety of Radiation Sources (BSS)-GSR Part 3, 2014).  [9]. The dosimetric parameter should bear a nearly linear relationship with radiation risks associated with examinations. To achieve these objectives the following dosimetric parameters have been widely adopted for monitoring in conventional radiography: 1) Entrance surface dose (ESD)-conventional radiography (could be obtained with TLD or by calculations); 2) Dose-area product (DAP)-conventional radiography (obtainable with DAP meter or by calculations).
These dosimetric parameters were introduced to verify that the dose descriptors used during imaging process are below the defined European values established after many trials. In Senegal, both local and national DRLs (LDRLs and NDRLs, respectively) are not available.
Registration of the existing X-ray units using conventional radiography, CT, fluoroscopy etc. is currently being undertaken throughout Senegal. In addition, a National Radiation Dose Database (NRDD) is required to access dose information and facilitate processing for future optimization programs. To establish patient DRLs for various radiography tests and raise public awareness about patient dose, it is useful to identify the medical centers associated with higher radiation doses. As a result, the adoption of measures such as equipment quality control can lead to a reduction in patient doses while improving image quality [10].
The lack of radiation protection culture and patient's radiation protection practices, which are poorly documented in Senegal, make a detailed study very timely. The one presented here is a first that aims at examining the situation in

Materials and Methods
This study began after we received permission from each of the managers of the six ( The patients that underwent AP/PA and LAT projection were considered as separate cases to estimate the dose for each projection. The data were collected by the physicist operating in the rooms included in the survey. The exposure parameters displayed on the console of the X ray unit during examination were recorded by the physicist. The examinations were performed in seven radiographic rooms in total, all of them equipped with systems using AGFA Computed Radiography (CR) with Automatic Exposure Control (AEC). The X-ray equipment was checked through quality control program to ensure the consistency of the equipment performance, the reliability and reproducibility of the exposure parameters. The quality control tests included the tube voltage accuracy and reproducibility, the current time product (mAs) linearity, the Half Value Layer (HVL) and the X-ray tube output measurement. The latter was measured at a distance of 100 cm from the X-ray source to the chamber (detector), for a tube voltage between 45 and 130 kVp in incremental steps of 5 kVp. All measurements were performed using Xi model R/F, serial number 185512 produced by Unifors RaySafe, Sweden. This is a multi-parameter X-ray detector which can measure kVp, dose, dose rate, HVL and time. To ensure the accuracy and precision of the quantities derived from calculation, we performed dosimetric measurements based on the recorded parameters used during examination.
Details on 2217 X-ray examinations were collected during a period of one year, and at least 30 patients were observed for each examination type. All the nine exams were not performed in each room. For each examination, personal data and technical parameters were collected according to a questionnaire designed for the patient's dosimetry protocol as follows:  [13]. The ESD represents the absorbed dose at the patient entrance and is given by Equation (1). For each patient, it was calculated using real examination data.
( ) In this equation, BSF is the backscatter factor, a is the patient thickness, b is the distance between the top of the table and the film and mAs is the tube current time product [14]. The X-ray tube output of each system was obtained during quality control with this power function: In this equation, A is the fitting factor and n is the power, both derived from the plotting of the tube output measurement against the tube voltage [15].
The dose area product (DAP), which is the product of the Incidence Air Kerma (IAK) (i.e. ESD in the absence of backscattering) by the irradiated area, is not only a quick and simple measurement but also a valuable radiation dose descriptor. Its advantage is that the biological effects of radiation are dependent on the radiation dose and the irradiated area of the body. The DAP is also applicable for quality assurance and functional analysis of X-ray machines [16] [17] [18]. It was calculated based on this formula [19]: The local DRLs were estimated at the 75th percentile of ESD and DAP for each exam and room to further assist in the optimization process by providing a local comparator linked to the differences between technology and variations in

Results
To obtain a typical dose estimate in a typical patient, the measurements were performed on a representative sample of adult patients. 2217 patients were included in this study. Table 1 shows the specific data of the X-ray units investi-         Table 3. The 75th percentiles of ESD and DAP values for each exam were compared with recommended international DRLs for some European countries in Table 4 and Table 5. Sweden [23] Italy [15] DRLs [24] Chest (

Discussions
This study evaluates the dose received by adult patients who underwent X-rays examinations in seven (7) radiographic rooms in Senegal. Great variations in patient doses were found in this survey as shown by the boxplots (Figure 3 and without degrading the quality of the image when using additional Cu filters after simulating 20 cm acrylic phantom [28]. Increasing the FFD also can lead to a  proportional to the collimation and could degrade the subject contrast without carrying information to the detector [29]. Reducing the field of view (FOV) to the area of interest could reduce the scattered radiation and the dose received by other organs as well as improve image quality. The use of anti-scatter grid, post processing algorithms (window level and width), specific training for technologists and radiologists, and implementation of quality assurance program can lead to dose reduction and improvement of the image quality [30]. The local dose audit reported in this study reveals the practice in few X-ray units. Although it is not representative of what happens in every hospital, it is an indication that dose optimization is possible in Senegal. The differences can be due to the effect that the diagnostic reference levels available for comparison are of European origin which have not been determined with the same equipment, training of machine users and patient morphology but also used a large sample.
The DRLs play critical role in the optimization of radiation dose. The great variations in dose found between medical centers indicate the necessity to optimize the practices. Our results can be applied to meet some requirement for the establishment of NDRLs which in turn can help to prevent unnecessary radiation dose. Low doses are critical because of their stochastics effect at long term.
That is why it is necessary for each institution to develop protocols for dose measurement that could contribute to both the establishment of LDRLs as well as in the evaluation of the local radiographic practice.

Conclusion
The survey has been conducted to investigate patient doses for nine routine types of X-ray examinations in radiology. The exposure of 2217 patients was analyzed; entrance surface dose and dose area product were evaluated. This study indicates the need to standardize the medical X-ray examination technique. It is very important to perform real patient dose measurements in hospitals. In that sense, the survey results are the link between patient dosimetry, as the first step in optimization of radiation protection and quality assurance program in diagnostic radiology. It is of great importance to extend the survey to a larger number of hospitals in order to establish diagnostic reference levels at the national level. Reference dose levels for diagnostic radiology examination provide the benchmark for comparing X-rays exposures from different facilities, to reduce patient doses and maintain good image quality with respect to basic principles of radiation protection of patient (justification and optimization). It is expected that this survey will encourage further efforts in organizing radiation protection program to effectively monitor patient dose and optimize the radiology practice.

Declarations
The research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sector.