Determination of serum creatinine prior to iodinated contrast media: is it necessary in all patients?

The risk of contrast-associated nephrotoxicity (CAN) is increased in the presence of preexisting renal disease. Although routine determination of serum creatinine (Cr) prior to imaging studies is the traditional method of assessing renal function, it is a costly and time-consuming practice. The purpose of this study was to investigate whether a patient survey could identify patients with a high likelihood of having normal Cr values and who, therefore, did not require serum testing. A survey was administered to 673 consecutive adult patients who were scheduled for contrast-enhanced computed tomography. Survey questions were designed to elicit a history of renal disorders as well as additional risk factors for CAN. Each patient had a Cr level determined within 48 hours prior to the injection of iodinated contrast media. Cr levels were assessed in the patients who gave negative responses to all survey questions. The degree to which positive responses to each survey question predicted elevated Cr levels was determined using the odds ratio (OR). Among the 673 respondents, 577 (85%) had normal Cr values (< or =1.3 mg/dL for women and < or =1.4 mg/dL for men). Completely negative responses to the questionnaire occurred in 191 (28%) of 673 of respondents, and 176 (92%) of these 191 had normal Cr values. A Cr cutoff value of 1.7 mg/dL is used in our department, i.e., patients with Cr values >1.7 mg/dL usually do not receive iodinated contrast media. Using this Cr cutoff value, 189 (99%) of 191 patients with negative responses had Cr values less than or equal to the cutoff value. The survey questions most strongly associated with elevated Cr values pertained to preexisting renal disease (OR 13.6), proteinuria (OR 8.7), prior kidney surgery (OR 8.1), hypertension (OR 5.4), gout (OR 4.6), and diabetes (OR 3.2). If the survey had been limited to these six questions, completely negative responses would have occurred in 450 (67%) of 673, 424 (94%) of these 450 would have normal Cr values, and 446 (99%) of 450 would have had Cr values at or below the 1.7 mg/dL cutoff for iodinated contrast. A completely negative response to a simple (six question) patient survey prior to iodinated contrast administration can identify a significant fraction of patients with normal Cr levels. Use of this survey could reduce by 67% the number of patients undergoing routine Cr determinations prior to imaging studies. This could reduce costs, decrease delays, and increase patient satisfaction associated with imaging studies.