Association of Contrast and Acute Kidney Injury in the Critically Ill: A Propensity-Matched Study

Abstract

Background: Despite evidence that low osmolar radiocontrast media is not associated with acute kidney injury, it is important to evaluate this association in critically ill patients with normal kidney function.

Methods: This retrospective observational study included 7,333 adults with an ICU stay at a six-hospital health system in south Florida. Patients who received contrast were compared with unexposed control subjects prior to and following propensity score (PS) matching derived from baseline characteristics, admission diagnoses, comorbidities, and severity of illness. Acute kidney injury (AKI), defined as initial onset (stage I) or increased severity, was determined from serum creatinine levels according to Kidney Disease: Improving Global Outcomes guidelines.

Results: Based on 2,306 PS-matched pairs obtained from 2,557 patients who received IV contrast and 4,776 unexposed control subjects, the increase in AKI attributable to contrast was 1.3% (19.3% vs 18.0%; P = .273), and no association was found between contrast and the pattern of onset and recovery. Hospital mortality increased by 14.3% subsequent to AKI (18.0 vs 3.6; P < .001), but the risk ratio in relation to patients with stable AKI did not vary when stratified according to contrast. Multivariable regression identified sepsis, metabolic disorders, diabetes, history of renal disease, and severity of illness as factors that were more strongly associated with AKI.

Conclusions: In critically ill adults with normal kidney function, low osmolar radiocontrast media did not substantively increase AKI. Rather than limiting the use of contrast in ICU patients, efforts to prevent AKI should focus on the susceptibility of patients with sepsis, diabetes complications, high Acute Physiology and Chronic Health Evaluation scores, and history of renal disease.

Keywords: ICU; acute kidney injury; community hospital; contrast; critical care; propensity score matching.