Risk of contrast induced nephropathy in the critically ill: a prospective, case matched study

Abstract

Introduction: Computerized tomography is frequently employed in the critically ill, often using intravenous radiocontrast material. Many of these patients have clinical features that are considered risk factors for contrast induced nephropathy, but are simultaneously at risk for renal injury from other factors related to their acute illnesses. The attributable risk for renal dysfunction from radiocontrast exposure has not been well quantified in this population.

Methods: A prospective matched cohort study was conducted of patients scanned with or without radiocontrast enhancement while receiving intensive care in a Veterans Affairs Medical Center. Patients were matched for pre-scan measured creatinine clearance, diabetes, mechanical ventilation, and vasopressor use. Measured clearance was followed for three days after scanning. Evolution of nephropathy, as determined by change in measured clearance, was compared within matched pairs.

Results: Fifty-three pairs of patients satisfied matching criteria. Unmatched characteristics were similar among the pairs, including serum creatinine variability during the week preceding scanning (67 ± 85% among contrast recipients, 63 ± 62% among others) and clinical risk factors for renal failure. In 29 pairs, pre-scan measured clearances were less than 60 mL/minute/1.73 m2. Following scanning, measured clearance declined by at least 33% in 14 contrast and 19 non-contrast patients (95% confidence interval for contrast associated difference in nephropathy rates -27% to 9%), while a 50% reduction in clearance persisted three days after scanning in three contrast and nine non-contrast patients (95% confidence interval for difference in rates -25% to 2%).

Conclusions: Among established intensive care unit patients declines in glomerular filtration following contrast-enhanced scanning are common, but these changes are far more likely to be attributable to factors other than the contrast exposure itself. The upper bound for the incidence of contrast induced renal injury lasting even three days was 2% in the population studied.

Figure 1

Screening, enrollment, and matching schema.

Figure 2

Variability and optimization of serum creatinine before scanning. Filled circles in the left panel represent the ratio of highest to lowest serum creatinine during the week preceding scanning in individual matched patient pairs, and in the right panel the ratio of the last serum creatinine measured before scanning to the lowest measured during the week preceding scanning. Filled triangles denote mean values for the pairs. Serum creatinine declined toward its recent minimum with pre-scanning management similarly in both groups.

Figure 3

Minimum mCr_Cl within the three days following scanning relative to pre-scan values. Filled circles represent individual matched patient pairs, and the filled diamond mean values for contrast and non-contrast patients. Points above the diagonal line imply greater loss of renal function in the patient who received contrast, and below the line greater loss in the patient who did not receive contrast.

Figure 4

mCr_Cl three days after scanning, relative to pre-scan values. Filled circles represent individual matched patient pairs, and the filled diamond mean values for contrast and non-contrast patients. Points above the diagonal line imply greater loss of renal function in the patient who received contrast, and below the line greater loss in the patient who did not receive contrast.