Purpose: To suggest a more precise tool when assessing the risk of contrast-medium-induced nephropathy (CIN), i.e. the ratio between contrast medium (CM) dose expressed in grams of iodine (g-I) and estimated glomerular filtration rate in ml/min (eGFR; based on equations using serum-creatinine (s-Cr), weight, height, age, and/or sex), here named I-dose/GFR ratio.
Material and Methods: A Medline search of published CIN investigations reporting mean eGFR and mean dose of low-osmolality CM (LOCM) identified 10 randomized controlled prophylactic and 2 cohort coronary investigations, and 3 randomized and 1 cohort computed tomographic (CT) investigation. From the randomized trials, data were collected only from the placebo or control arms, unless there was no significant difference between the control and test groups. The mean I-dose/GFR ratio of each study was correlated with the mean frequency of CIN-1 (s-Cr rise⩾44.2 µmol/l or ⩾20–25%) and CIN-2 (oliguria or requiring dialysis). A maximum dose according to an I-dose/GFR ratio = 1 in patients with s-Cr ranging from 100 to 300 µmol/l was compared with that of Cigarroa's formula and with a “European consensus” threshold published by the European Society of Urogenital Radiology, both using s-Cr alone to predict renal function. McCullough's formula was used to assess the risk of CIN requiring dialysis at an I-dose/GFR ratio = 1 with LOCM.
Results: The coronary investigations revealed a linear correlation with a correlation coefficient between the I-dose/GFR ratio and the frequency of CIN-1 and CIN-2 of 0.91 (P<0.001) and 0.84 (P = 0.001), respectively. At a mean I-dose/GFR ratio = 1, the regression line indicated a 10% risk of CIN-1 and a 1% risk of CIN-2. At a mean I-dose/GFR ratio = 3, the risk of CIN-1 and CIN-2 increased to about 50% and 15%, respectively. Pooled weighted data from the CT investigations revealed a 12% risk of CIN-1 at a mean I-dose/GFR ratio = 1.1 and no cases of CIN-2. The maximum CM dose according to an I-dose/GFR ratio = 1 was about 30–50% of that of both Cigarroa's formula and the “European consensus” in elderly low-weight individuals, while it was similar for middle-aged individuals weighing about 90 kg. McCullough's formula suggests that there will be an exponentially increasing risk of CIN requiring dialysis, but at an I-dose/GFR ratio = 1 and using LOCM it will not exceed 1% until GFR decreases below 30 ml/min in diabetics and below 20 ml/min in non-diabetics.
Conclusion: Using the I-dose/GFR ratio may be a more expedient way of improving risk assessment of CIN than today's common practice of estimating CM dose from volume alone and renal function from s-Cr alone. Prospective studies based on individual patient data are encouraged to define the risk of CIN at various I-dose/GFR ratios and correlated to type of CM, examination, risk factors, etc.
