The diagnostic value of contrast administration following normal non-contrast brain computed tomography in Mankweng Hospital

Abstract

BACKGROUND: The use of intravenous contrast during computed tomography (CT) scans enhances the detection and characterization of pathologies. However, its high cost, along with the potential adverse effects of the contrast agents and radiation, presents significant challenges for the resource-limited public health sector in South Africa. It is the responsibility of radiologists to justify the use of intravenous contrast and the exposure of patients to ionizing radiation. AIM: This study aims to determine whether the administration of contrast significantly alters the diagnosis of normal non-contrast CT brain scans. METHODS: A descriptive retrospective cross-sectional study was conducted at Mankweng Hospital, analysing 181 brain CT images acquired between January 1, 2021, and January 31, 2022. All patients underwent consecutive non-contrast CT (NCCT) and contrast-enhanced CT (CECT) scans, with CECT performed immediately following NCCT. A simple random sampling approach was employed. Senior registrars reported the scans, a consultant radiologist verified the reports, followed by a discussion of a subset of equivocal scans in a consensus meeting. Results were recorded and statistically analysed using SPSS. The kappa coefficient was utilized to measure the agreement between the NCCT and CECT scans, with all statistical analyses conducted at a 5% level of significance. RESULTS: Among the 181 sampled scans, 88 (48.6%) were initially reported as normal on NCCT. Of these, 84 (95.5%) remained normal after contrast administration, while 4 (4.5%) exhibited significant diagnostic changes after receiving the contrast. The kappa coefficient of 0.945 indicated almost perfect agreement between the NCCT and CECT scans (p < 0.005). The administration of contrast provided additional diagnostic value primarily in immunocompromised patients or those with fever, seizures, or decreased levels of consciousness. CONCLUSION: This study demonstrates that administering contrast following a normal NCCT brain scan may not be necessary. Omitting this step could reduce healthcare costs, scan time, and radiation exposure. Radiologists should justify the use of contrast and repeat scans based on clinical information. We recommend adopting a “scan, review, and justify” approach to optimize resource utilization and minimize harm. Future studies should investigate the cost-effectiveness of this approach.