Contrast injection protocols used in oncological CT and opportunities for practice optimisation

INTRODUCTION: Contrast-enhanced CT is central to oncological imaging, yet no official guidelines exist for contrast injection protocols. As a result, practices vary widely between institutions affecting image quality and patient safety. Understanding this variability is essential to decrease this heterogeneity. This study aimed to map current oncological body CT contrast protocols in Western Switzerland. METHODS: An online survey was specifically designed and disseminated by using LimeSurvey platform to 74 medical imaging institutions to collect data about: a) socio-demographic characteristics of contrast users, b) technical acquisition protocols, c) injection protocols, d) optimisation practices, e) clinical information. The survey was composed of 22 closed and 7 open-ended questions. RESULTS: Of the 74 institutions, 42% (31) completed the survey. Regarding the identified protocols, 74% used single-energy CT (23/31), while 26% used dual-energy CT (8/31). Regarding image reconstruction, 68% employed iterative reconstruction algorithms (21/31), whereas 32% used deep-learning-based reconstruction (10/31). Contrast volume was primarily calculated based on total-body-weight (84%; 27/31), with an average injection rate of 3 mL/s. The most used iodine concentrations were 350 mgI/mL (55%; 17/31) and 400 mgI/mL (29%; 9/31). Radiographers (90%; 28/31), radiologists (58%; 18/31) and application engineers (55%; 17/31) were the specialists most involved in protocol optimisation. Dose adjustments were reported in cases of renal impairment (70%; 22/31) and patient morphology (61%; 19/31). CONCLUSION: These results highlight a marked heterogeneity in oncologic CT practices across Western Switzerland, with significant optimization potential using technologies like spectral imaging and deep learning reconstructions. IMPLICATIONS FOR PRACTICE: There is a need to update injection protocols in line with current technological advances, such as spectral imaging and new reconstruction algorithms. These updates should be increasingly based on evidence-based practice.