Utility, Appropriateness of Request and Report of Computed Tomography Scan for the Diagnosis of Paranasal Sinuses Pathologies in a Sub-Saharan Africa Urban Setting

Background: Computed tomography (CT) Scan is the imaging of choice for the diagnosis of sino nasal (SN) pathologies in sub-Saharan African setting. However, it is often requested by non experienced practitioners. We aimed to describe the utility and the appropriateness of CT Scan request forms and results in the diagnosis of SN pathologies. Method: We conducted a retrospective cross-sectional study in four health facilities in Yaoundé, Cameroun, from October 2018 to September 2019. All patients to whom SN CT Scan was performed were called for an appointment. After informed consent, data on the appropriateness and conformity of the request form were collected and compared to the guidelines of the French society of radiology. CT Scan results were also collected and the quality of the report was evaluated. Results: A total of 206 SN CT Scan request forms and reports were included, with a mean patient age of 37.7 ± 16.1 years. The request form was conformed to guidelines in 8.3% of cases and was found appropriate in 62.6% of cases. No CT Scan report had all the items required, and 159 (77.2%)


Introduction
Sino nasal pathologies are a group of diseases affecting aerial cavity within the bones of the face or the base of the skull. They can be grouped into three main categories: 1) post-traumatic pathologies; 2) inflammatory diseases including acute and chronic rhinosinusitis and Sino nasal polyposis pattern; 3) tumoral pathologies mostly benign and malignant primitive tumours [1]. These pathologies are especially important and potentially serious diseases because they extend to the central nervous system either through the venous system or by contiguity. They also represent a great public health concern due to their high socioeconomical impact and alteration of quality of life. Indeed, chronic rhinosinusitis patients in United Kingdom were found to have a high level of healthcare service use, out-of-pocket expenditure and productivity loss while it results in an average of 20.6 workdays missed during one year in Canada [2] [3]. In the USA, the disease affects more than 16% of the population [4]. A similar prevalence was found in Cameroon between 2011 and 2015, with 11.7% of people affected with chronic rhinosinusitis [5].
Imaging for the diagnosis of sino nasal pathologies is based on standard X-ray, computed tomography (CT) scan, and magnetic resonance imaging (MRI). The usage of X-ray is limited to exploration of intrasinusal liquid, as it underestimates the extent of mucosal and bone lesions. Furthermore, X-ray permits a poor analysis of sphenoidal and ethmoidal sinuses. On the other hand, MRI gives a detailed analysis of paranasal sinuses and their link to encephala and meninges [6].
However, the high cost and the few number centers where MRI can be done in our setting (including urban setting) reduce the accessibility of this exam. CT Scan, therefore, remains the analysis of choice in sub-Saharan Africa setting for the diagnosis of paranasal sinuses pathologies. Due to the lack of specialists, it can be requested by every practitioner, including general practitioners and nurses which may not be trained for the diagnosis of sino nasal diseases. Radiologists are not always experienced in the report of sino nasal CT Scan results. We aimed to study the practice and utility of CT Scan for the diagnosis of paranasal sinuses pathologies in a sub-Saharan urban setting. The results will help to design better strategies for capacity building in the management of sino nasal pathology in order to reduce their burden in our setting.

Study Design and Population
This was a retrospective cross-sectional study design conducted in radiology and Open Journal of Radiology

Data Collection
Patients with paranasal sinuses CT Scan performed during the study period were identified through the units' registries and their phone numbers (or the number of their relatives) were collected. Participants or their relatives were therefore contacted by phone for an appointment in the unit where the exam was done, the patient or relative was asked to come with the CT Scan request form and report. Selection criteria were checked on arrival and a written consent was ob- study without systematic application of a reference standard; Level of evidence 4: Case-control study; study using a non-independent or low-quality reference standard; Level of evidence 5: Deductive reasoning. Open Journal of Radiology The results of the CT Scan (radiological diagnosis, location of lesions and anatomical variants) were collected from the report and the quality of reporting was assessed using the 18 criteria described by the FSR. An exam was considered useful when it was done according to the protocol recommended for the condition, the semiology of lesions was presented on the report and there was an answer to the clinical question asked on the request form (if there was such a question on the request form) [9].

Statistical Analysis and Sample Size Calculation
The minimum sample size was estimated at 93 request forms and reports, using a precision of 10%, a type one error of 5% and an estimated proportion of appropriate imaging of 59.5% found by Moifo et al. [10]. Data collected were entered in a data entry application designed in Census and surveys processing system (CsPro) version 7.0. Qualitative data were described as number and percentage while continuous data were described as mean ± standard deviation.

Ethical Consideration
Prior to patient inclusion, authorization was obtained from the directors of the corresponding hospitals and an ethical clearance was obtained from the institu-

Description of the Study Population
A total of 206 CT Scan request forms and reports were included in the study.
Females were more frequent (53.9%) and the mean age of patients was 37.7 ± 16.1 years, ranging from 1 to 86 years ( Table 1). Most of the exams were requested by specialists (53.9%), mostly otorhinolaryngologists (41.7%), and the practitioner requesting for the analysis was not specified in 27 (13.1%) cases. Table 1 gives more details on the study population.

Conformity of the CT Scan Request form
Conformity criteria are described in Table 2. All the 08 items recommended by the NAH guidelines were found on 17 (8.3%) request forms ( Table 2). The patient's name and age, and the anatomical region to be analyzed were present on Y. C. M. Eng et al. The quality of the CT Scan request forms was described according to the French NAH guidelines for the conformity of imaging exams requests. The appropriateness of the request was described according to Guide for proper use of medical imaging examinations of the French Society of Radiology. CT: Computed Tomography; NAH: National Authority for Health. *The item couldn't be assessed due to absent or insufficient information for the assessment.
almost all the request forms; while the purpose of the exams and the name of the requesting practitioner were reported on less than 25% of request forms ( Table 2).
Clinical information provided on forms was mainly clinical diagnosis, symptoms and syndromes, respectively in 35%, 18.4% and 17% of request forms. However, no clinical information was present on 19.9% of request forms (Table 2).

Appropriateness of the Request
The CT Scan requested for paranasal sinuses pathologies diagnosis was found pertinent for 129 (62.6%) patients; the request was not appropriate or the appropriateness couldn't be evaluated for 75 (36.4%) cases (Table 2). Among those which were appropriate, 110 (53.4%) were Grade B recommendation (Table 2).

Quality of CT Scan Report and Utility
All the CT Scan reports contained patient identification, radiologist identity, date of report writing, anatomical region examined results of the exam and the validator signature. No report had the date on which the exam was done, and 11.2% reported the date of report validation ( Table 3). Characteristics of radiocontrast

Radiological Profile of Study Population
The CT Scan was normal for 52 (25.2%) patients. The most common diagnosis in patients with abnormal results were chronic sinusitis and sino nasal polyp/polyposis (Table 4). These pathologies were mostly found in the maxillary and the anterior ethmoidal sinuses, respectively for 153 (74.3%) and 104 (50.5%) cases (Table 4). At least one anatomical variant was observed in 49 (23.8%) results, Concha bullosa (12.1%) and septal deviation with bone spur (8.3%) been the more common. See Table 4 for more details on radiological profile. Clinoid apophyse pneumatisation 2 1 Dehiscence of the planum bone 1 0.5 No anatomical variant

Discussion
CT Scan like other imaging methods are intended to provide accurate and complete identification of relevant findings, a coherent opinion of the radiologist concerning the underling cause of abnormalities and if appropriate a guidance on other investigation to be done. However, an appropriate request form and a good report are required for these purposes. In some sub-Saharan Africa settings, CT Scan is a key element for diagnosis of paranasal sinuses pathologies.
The request forms are often done by non specialist and the results report by non experienced radiologists. We found in this study that less than 10% of request forms had all the required items, about two third of the requests were appropriate and the CT Scan was found useful in 77.2% of cases.
Among the 206 request forms included in this study, the name and age of the patient was written on almost all the forms; a result found also by Roussel et al. [12]. Indeed, the identity of the patient is the first element requested on the form, and define the subject to whom the exam should be done.  [14]. Similarities can be related to the prescribing practitioners which included general practitioners in those studies and in our population. The proportion of inappropriate CT Scan request is still high, suggesting the need of capacity building for those who prescribe this exam, especially in our low resource setting where most of the patients have to pay out pocket. These inappropriate requests constitute a source of wasting resources, and may explain the 52 (25%) normal CT Scan results found in our population.
No report was contained all the required item for a good imaging result report, yet all these reports were validated by a senior radiologist. The dosimetry and the equipment used for the analysis were among the less reported items, present respectively on 11.2% and 18% of the reports. Although the radiation dosimetry is required to inform the level of exposure to ionisation and thus the risk from the CT Scan especially for children which may have a lifetime risk of fatal cancer as high as 1 in 1000 cases. The radiation dosimetry is also intend to assess the degree of optimisation of radiologic procedure in the unit [15]. Radi-Open Journal of Radiology ologist should be aware that these easy to obtain information are really important both for the patients' protection and the radiology unit improvement. Other key elements which were absent on the reports are the characteristics of radiocontrast agents used for the exam. Indeed, they were reported for only 5% of the patients. Knowing that adverse events after administration of radiocontrast media can occur up to one week after injection-especially in young adult and women which were more frequent in this population-lacking information on the type, the concentration and the volume of radiocontrast media will really hamper management of these patients in case of adverse events [16].
The overall utility of the CT Scan was good in 77.2% of cases, which is lower than the 87.5% of useful CT Scan found by Chin et al. concerning non traumatic acute abdominal pain [17]. This difference may be related to the easiest identification of lesions on abdominal CT Scan compare to paranasal sinuses analysis.
Furthermore, the utility of the exam may have been further altered by the poor quality of request forms and finding reports. For an efficient usage of resources, all these components of the CT Scan results chain should be improved in our context.
This study presents some limitations including the fact that requesting practitioner was not contacted for more information in order to better assess the appropriateness of the request, nor the patient's file used for this purpose. This approach would have produced more reliable results concerning the real appropriateness of the requested exams. Another limitation can be the exclusion of non validated results, as they may also represent poor reporting practices. We choose to exclude this population because non validated results may have been produced out of the normal path of results in the facility.

Conclusion
CT Scan requested for diagnosis of paranasal sinuses pathologies were useful for less than 80% of patients. The request forms were delivered by non specialists in about half of cases, and were not appropriate for 63.6% of patients. No report of findings contained all the required items, with some key elements like dosimetry and characteristics of radiocontrast agent used absent in about 90% of reports.
These results suggest the need for capacity building of all the practitioners who may request CT Scan and radiologist in charge of interpretating and validating results in our setting. In order to efficiently use the resources which are sparse, and to improve the security of patients in case of adverse events.