Key points are not available for this paper at this time.
ABSTRACT Iodinated contrast media are central to modern medicine, but have historically been associated with contrast‐induced nephropathy (CIN), prompting widespread caution when imaging patients with kidney disease. This has led to delays, precautionary measures and avoidance of clinically indicated imaging. Contemporary evidence suggests that the risk of clinically meaningful kidney injury attributable to modern contrast agents has been overstated. We conducted a structured narrative review of the literature examining iodinated contrast media and kidney injury. Searches of PubMed/MEDLINE, Embase, Scopus and Google Scholar were performed in January 2026. Evidence was synthesised to evaluate the incidence, pathogenesis, clinical outcomes and prevention of contrast‐associated acute kidney injury. Early reports of CIN were largely based on uncontrolled observational studies that attributed post‐contrast creatinine rises to nephrotoxicity without appropriate comparator groups. Contemporary controlled studies demonstrate no independent association between intravenous contrast and clinically meaningful outcomes such as dialysis, chronic kidney disease progression or mortality. Experimental models show plausible mechanisms of injury but often rely on conditions not representative of routine practice. Biomarker studies highlight discordance between functional and structural kidney injury. Preventive strategies, including prophylactic hydration, have not demonstrated benefit in high‐quality trials. Overall, clinically significant kidney injury attributable to modern iodinated contrast media appears uncommon. Excessive caution may delay or prevent necessary diagnostic and therapeutic procedures. A balanced, evidence‐based approach prioritising timely imaging while addressing modifiable risk factors may improve patient outcomes and reduce barriers to care. The evidence synthesised in this review informs updated national guidance on contrast use in Aotearoa New Zealand.
Shaw et al. (Thu,) studied this question.