Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH
QUICK SEARCH:   [advanced]


     

Published online before print May 5, 2008, 10.1148/radiol.2481071528

(Radiology 2008;248:106.)

A more recent version of this article appeared on July 1, 2008
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow All Versions of this Article:
2481071528v1
248/1/106    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Herts, B. R.
Right arrow Articles by Baker, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Herts, B. R.
Right arrow Articles by Baker, M. E.
© RSNA, 2008

Contrast Media

Identifying Outpatients with Renal Insufficiency before Contrast-enhanced CT by Using Estimated Glomerular Filtration Rates versus Serum Creatinine Levels1

Brian R. Herts, MD, Erika Schneider, PhD, Emilio D. Poggio, MD, Nancy A. Obuchowski, PhD, and Mark E. Baker, MD

1 From the Division of Radiology (B.R.H., M.E.B., E.S.), Department of Nephrology and Hypertension (E.D.P.), and Department of Quantitative Health Sciences (N.A.O.), Cleveland Clinic, 9500 Euclid Ave, Desk Hb6, Cleveland, OH 44195. Received August 29, 2007; revision requested October 29; revision received November 17; accepted January 8, 2008; final version accepted February 19. Supported in part by a research award from the Society of Computed Body Tomography. Address correspondence to B.R.H. (e-mail: hertsb@ccf.org).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Purpose: To determine whether using estimated glomerular filtration rate (eGFR) values rather than serum creatinine levels to identify patients with renal insufficiency facilitates any substantial change in the number of outpatients scheduled for computed tomography (CT) who are considered at increased risk for contrast medium–induced nephropathy.

Materials and Methods: The study was HIPAA compliant and institutional review board approved for medical chart review; the requirement for informed patient consent was waived. Patients (n = 5138; 2569 women, 2569 men, 753 African Americans, 4385 non–African Americans) examined during a 2-year period formed the final study group after exclusion of patients undergoing dialysis (n = 49), for whom no age data were recorded (n = 9), and younger than 18 years (n = 113). Patient age, sex, and race and the blood urea nitrogen, albumin, and serum creatinine levels most recently measured within 6 months before CT were obtained from the electronic medical records. The number of patients with creatinine levels higher than 1.4 mg/dL was directly compared with the number of patients with eGFR values (calculated with four- and six-variable Modification of Diet in Renal Disease [MDRD] equations) lower than 60 mL/min/1.73 m2 by using the two-tailed McNemar test. For 2689 patients, data to calculate the eGFR by using the four-variable equation were available, and for 2005 patients, data to calculate the eGFR by using the six-variable equation were available.

Results: Among the outpatients scheduled to undergo CT, the percentage of patients with an eGFR lower than 60 mL/min/1.73 m2 was significantly greater than the percentage of patients with a creatinine level higher than 1.4 mg/dL for both the four-variable (412 [15.3%] vs 166 [6.2%] of 2689 patients) and the six-variable (346 [17.3%] vs 117 [5.8%] of 2005 patients) MDRD equation groups (P < .001).

Conclusion: A significantly higher number of outpatients scheduled for contrast medium–enhanced CT met the National Kidney Foundation criteria for renal insufficiency when the MDRD equations were used to estimate the glomerular filtration rate compared with the number of outpatients who met the criteria on the basis of elevated creatinine levels.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
The reported prevalence of iodinated contrast medium–induced nephropathy (CIN) after administration of any amount of contrast material ranges from 1%–30%, and the relative risk of CIN increases with preexisting renal disease, diabetes (especially in patients with renal insufficiency), and congestive heart failure—among other conditions (1–4). Serious complications of CIN include permanently reduced renal function and renal failure requiring dialysis. However, most cases of CIN are self-limited, with renal function returning to normal within 2 weeks (5). CIN reportedly occurs after intravenous contrast medium injection for computed tomography (CT) in only 4% of patients with chronic kidney disease (CKD) (6). Therefore, CIN should occur less frequently after intravenous contrast medium injection for CT in patients with normal or mildly reduced renal function; however, to our knowledge, there are no studies that specifically address this clinical scenario.

Serum creatinine level generally has been used as a surrogate measure of renal function in place of creatinine clearance or inulin clearance, the reference standard, because serum creatinine level measurements are readily available, minimally invasive, and inexpensive. Before the Kidney Disease Outcomes Quality Initiative guidelines were published by the National Kidney Foundation, patients with mildly elevated creatinine levels (1.5–1.9 mg/dL) generally were considered to have mild renal insufficiency and to have some—but not a highly—increased risk for CIN. Patients with moderately elevated creatinine levels (≥2.0 mg/dL) were considered to have moderate renal insufficiency and a higher risk for CIN. The cutoff creatinine levels at which radiologists consider it safe to administer contrast media with respect to the risk of CIN vary widely (7).

The National Kidney Foundation has reported that the glomerular filtration rate (GFR), either measured or estimated, is a better determinant of kidney function than the serum creatinine level (8). Furthermore, nephrologists now recommend using estimated or measured GFR rather than creatinine level alone to identify individuals at risk for CIN (9). The estimated GFR can be calculated by using formulas based on the serum creatinine level in combination with the subject's weight, age, sex, race, blood urea nitrogen (BUN) level, and/or albumin level (10,11). It should be noted, however, that the equations used to estimate the GFR are empirically derived and are not considered accurate in patients who have normal renal function (12).

It is estimated that 4% of outpatients who present for contrast medium–enhanced CT have an elevated creatinine level (13,14). To our knowledge, however, the relationship, among patients referred for CT, between number of patients with serum creatinine level–based reduced renal function and number of patients with estimated GFR–based reduced renal function is unknown. The use of estimated GFR rather than creatinine level to identify patients with CKD could result in a larger number of patients who would need to be considered for renal protective strategies or who might be considered at an unacceptable risk of CIN. The purpose of our study was to determine whether using estimated GFR values rather than serum creatinine levels to identify patients with renal insufficiency facilitates any substantial change in the number of outpatients scheduled for CT who are considered at increased risk for CIN.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Selection Criteria and Patient Population
Our retrospective study was Health Insurance Portability and Accountability Act compliant and institutional review board approved for medical chart review; the requirement for informed patient consent was waived. All contrast-enhanced CT scans at our medical center during a 2-year period (January 2004 to December 2005) were identified from our radiology information system (IDX, version 9.0; IDX, GE Healthcare, Milwaukee, Wis). From this group of more than 140 000 examinations, 6000 examinations were randomly selected by using the data of 30 patients selected per day on consecutive dates starting June 1, 2004, and outpatient examination codes. (A total of 6000 examinations were chosen on the basis of a consultation with our information technology division to limit the data set size.) Only the first contrast-enhanced CT scan obtained in each patient (n = 5309) was included in the study. Forty-nine patients who were undergoing dialysis (determined from direct review of the medical records of patients with creatinine values higher than 2.5 mg/dL), nine patients for whom no age data were recorded, and 113 patients younger than 18 years were excluded. Therefore, the final study group consisted of 5138 individuals who underwent contrast-enhanced CT scanning (Figure).


Figure 1
View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Flow diagram shows patient inclusion and exclusion criteria and numbers of patients with four- and six-variable Modification of Diet in Renal Disease (MDRD) equation data available for calculation of estimated GFR (eGFR).

 
Our institutional practice of using serum creatinine levels to screen for CKD before any CT examination closely follows the American College of Radiology guidelines (3): Patients younger than 70 years with neither a history of renal insufficiency nor a medical history that suggests renal insufficiency are not required to have their serum creatinine level measured prior to intravenous contrast medium administration. In all other patients, a serum creatinine sample must be taken within 6 months before the CT examination or more recently if the medical history dictates it (eg, in a patient undergoing chemotherapy).

Electronic medical records (EpicCare; EPIC Systems, Verona, Wis) were used to gather the following data for each patient: date of birth, sex, race—specifically African American or non–African American, all serum creatinine levels measured within the 180 days prior to the date of CT, all BUN measurements, and all albumin measurements.

The 5138 patients whose data were included had a median age at the time of CT of 60 years (mean age, 58.5 years; age range, 18–97 years). There were 2569 male and 2569 female patients; 753 patients were African Americans, and 4385 were non–African Americans. The creatinine levels of 2689 patients (1326 women, including 260 [19.6%] African Americans; 1363 men, including 175 [12.8%] African Americans) were recorded within 180 days before the CT examination. In 2005 of these 2689 patients (1003 women, including 184 [18.3%] African Americans; 1002 men, including 127 [12.7%] African Americans), albumin and BUN levels were also recorded within 6 months before CT (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Patients with Available Four- and Six-Variable MDRD Equation Data

 
The median and mean times between creatinine measurement and CT examination were 30.0 and 43.7 days, respectively. For the patients in whom creatinine, BUN, and albumin levels were measured, the median and mean times between BUN measurement and CT examination were 30.0 and 47.5 days, respectively, and the median and mean times between albumin measurement and CT examination were 15.0 and 37.2 days, respectively.

The estimated GFR was calculated by using the MDRD equations. These equations were used because they are the best formulas available for estimating the GFR in outpatients with CKD (11,12) and because this is the method for estimating kidney function advocated by the National Kidney Foundation. As such, the findings of our study can be better generalized to other institutions. The four- and six-variable MDRD equations are cited in the Appendix.

Populations Examined by Using Serum Creatinine Level versus Estimated GFR
In those patients who had the data needed to calculate the estimated GFR by using the four-variable MDRD equation, the numbers and percentages of patients with serum creatinine levels of less than or equal to 1.4 mg/dL, 1.5–1.9 mg/dL, 2.0–2.4 mg/dL, and higher than 2.4 mg/dL and of patients with estimated GFR values of greater than or equal to 90 mL/min/1.73 m2, 60–89 mL/min/1.73 m2 (stage 2 or mild CKD), 30–59 mL/min/1.73 m2 (stage 3 or moderate CKD), and less than 30 mL/min/1.73 m2 (stage 4 or 5, or severe CKD and renal failure) were determined.

Similarly, for those patients who had the data needed to use the six-variable MDRD equation, the numbers and percentages of patients with serum creatinine levels of less than or equal to 1.4 mg/dL, 1.5–1.9 mg/dL, 2.0–2.4 mg/dL, and higher than 2.4 mg/dL and of patients with estimated GFR values of greater than or equal to 90 mL/min/1.73 m2, 60–89 mL/min/1.73 m2, 30–59 mL/min/1.73 m2, and less than 30 mL/min/1.73 m2 were determined (8).

Subsequently, among both the patients with available four-variable equation data and those with available six-variable equation data, the number of patients with reduced renal function based on a creatinine level higher than 1.4 mg/dL was compared with the number of patients with reduced renal function based on an estimated GFR lower than 60 mL/min/1.73 m2. We chose these definitions to identify patients at higher risk for CIN due to CKD on the basis of a greater than 50% reduction in renal function; this cutoff is also used by the Canadian Association of Radiologists (15). This comparative analysis was performed for the entire study population and then for the race- and sex-based subpopulations: African Americans, non–African Americans, women, and men.

Patients with Creatinine Level–based Normal Renal Function and Abnormal Estimated GFR
We also determined, among the patients with available four- and six-variable MDRD equation data, the number of patients whose creatinine level–based renal insufficiency status differed from the estimated GFR–based renal insufficiency status. Specifically, we determined the number of patients with a creatinine level of less than or equal to 1.4 mg/dL (normal) who had estimated GFR values of less 60 mL/min/1.73 m2 and less than 30 mL/min/1.73 m2 (abnormal), and the number of patients with creatinine levels of 1.5–1.9 mg/dL (abnormal) who had estimated GFR values of less 60 mL/min/1.73 m2 and less than 30 mL/min/1.73 m2.

Data Abstraction
We also manually abstracted data from the electronic medical records to verify the electronic data derivation in a subset of the patients. A sample of 48 patients was randomly selected for abstraction. Abstraction was performed by a licensed registered nurse with more than 10 years of critical care experience and more than 1 year of abstraction experience. Abstraction consisted of a manual review of the electronic medical records that spanned the time range of 3 months before to 3 months after every contrast-enhanced CT examination. The following variables were abstracted and compared: medical record number; patient age, sex, and race; CT examination date; serum creatinine level and measurement date; and BUN level and measurement date. Data abstraction was used to assess the accuracy of the electronic data because it was not feasible to check all data for accuracy in a data set of this size. The abstraction process consisted of manually checking data in a smaller subset of patients and estimating the accuracy for the entire data set.

Statistical Analysis
The two-tailed McNemar test was used to compare the numbers of patients considered to have reduced renal function on the basis of the two criteria (creatinine level > 1.4 mg/dL and estimated GFR < 60 mL/min/1.73 m2). P < .05 was considered to indicate a significant difference.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Patients with Creatinine Level– versus Estimated GFR–based Reduced Renal Function
Of the 2689 patients with available four-variable MDRD equation data (creatinine level, age, sex, and race), 2523 (93.8%) had a creatinine level of less than or equal to 1.4 mg/dL; 142 (5.3%), a creatinine level of 1.5–1.9 mg/dL; 23 (0.9%), a creatinine level of 2.0–2.4 mg/dL; and one (0.04%), a creatinine level higher than 2.4 mg/dL.

When we used the four-variable MDRD equation to calculate the estimated GFR, 1251 (46.5%) of the 2689 patients had an estimated GFR of greater than or equal to 90 mL/min/1.73 m2; 1026 (38.2%) patients, an estimated GFR of 60–89 mL/min/1.73 m2; 406 (15.1%) patients, an estimated GFR of 30–59 mL/min/1.73 m2; and six (0.2%) patients, an estimated GFR lower than 30 mL/min/1.73 m2.

Of the 2005 patients with available six-variable MDRD equation data (age, sex, race, creatinine level, BUN level, and albumin level), 1888 (94.2%) had a creatinine level of less than or equal to 1.4 mg/dL; 101 (5.0%), a creatinine level of 1.5–1.9 mg/dL; 15 (0.8%), a creatinine level of 2.0–2.4 mg/dL; and one (0.05%), a creatinine level higher than 2.4 mg/dL.

When we used the six-variable MDRD equation to calculate the estimated GFR, 898 (44.8%) of the 2005 patients had an estimated GFR of greater than or equal to 90 mL/min/1.73 m2; 761 (38.0%), an estimated GFR of 60–89 mL/min/1.73 m2; 333 (16.6%), an estimated GFR of 30–59 mL/min/1.73 m2; and 13 (0.6%), an estimated GFR lower than 30 mL/min/1.73 m2.

Thus, with use of the four-variable MDRD equation, 6.2% (166 of 2689) of patients had an elevated creatinine level (>1.4 mg/dL) compared with 15.3% (412 of 2689) of patients with an estimated GFR lower than 60 mL/min/1.73 m2, and with use of the six-variable MDRD equation, 5.8% (117 of 2005) of patients had an elevated creatinine level compared with 17.3% (346 of 2005) of patients with an estimated GFR lower than 60 mL/min/1.73 m2. The percentage of patients with an estimated GFR lower than 60 mL/min/1.73 m2 was significantly larger than the percentage of patients with a creatinine level higher than 1.4 mg/dL at both four-variable and six-variable MDRD equation analyses (P < .001).

Race- and Sex-based Differences in Renal Sufficiency Status
Race- and sex-based differences between the percentage of patients with renal insufficiency based on a creatinine level higher than 1.4 mg/dL and the percentage of patients with renal insufficiency based on an estimated GFR lower than 60 mL/min/1.73 m2 were also significant (Table 2). Among the African American patients, there were significantly fewer patients with a creatinine level higher than 1.4 mg/dL than patients with an estimated GFR lower than 60 mL/min/1.73 m2 at both four-variable (25 [5.8%] vs 38 [8.7%] of 435 patients) and six-variable (16 [5.1%] vs 36 [11.6%]) of 311 patients) MDRD equation analyses (P < .001).


View this table:
[in this window]
[in a new window]

  		Table 2. Percentages of Patients with Creatinine Level– versus Estimated GFR–based Reduced Renal Function at Four-Variable MDRD Equation Analysis

 
Among the non–African American patients, there were significantly fewer patients with a creatinine level higher than 1.4 mg/dL than patients with an estimated GFR lower than 60 mL/min/1.73 m2 at both four-variable (142 [6.3%] vs 374 [16.6%] of 2254 patients) and six-variable (101 [6.0%] vs 294 [17.4%] of 1694 patients) MDRD equation analyses (P < .001).

Among the female patients, there were significantly fewer patients with a creatinine level higher than 1.4 mg/dL than patients with an estimated GFR lower than 60 mL/min/1.73 m2 at both four-variable (46 [3.5%] vs 206 [15.5%] of 1326 patients) and six-variable (34 [3.4%] vs 168 [16.8%] of 1003 patients) MDRD equation analyses (P < .001).

Among the male patients, there were significantly fewer patients with a creatinine level higher than 1.4 mg/dL than patients with an estimated GFR lower than 60 mL/min/1.73 m2 at both four-variable (119 [8.7%] vs 206 [15.1%] of 1363 patients) and six-variable (83 [8.3%] vs 164 [16.4%] of 1002 patients) MDRD equation analyses (P < .001).

Patients with Creatinine Level–based Normal Renal Function but Estimated GFR–based Reduced Renal Function, and Vice Versa
Of the patients with available MDRD equation data, 2524 had normal creatinine values (≤1.4 mg/dL), and 250 (9.9%) of these patients had an estimated GFR lower than 60 mL/min/1.73 m2. Four (0.2%) of the 2269 patients with an estimated GFR greater than or equal to 60 mL/min/1.73 m2 had a creatinine level higher than 1.4 mg/dL (Table 3). Of the patients with available six-variable MDRD equation data, 1888 had normal creatinine values, and 217 (11.5%) of these patients had an estimated GFR lower than 60 mL/min/1.73 m2. Two (0.1%) of the 1888 patients, who had an estimated GFR higher than 60 mL/min/1.73 m2, had a creatinine level higher than 1.4 mg/dL.


View this table:
[in this window]
[in a new window]

  		Table 3. Patients with Creatinine Level– or Estimated GFR–based Normal Renal Function at Four-Variable Equation Analysis but Reduced Renal Function with Other Criteria

 
Nearly all patients with creatinine levels of 1.5–1.9 mg/dL had estimated GFR values lower than 60 mL/min/1.73 m2 at both four-variable (138 [97.2%] of 142 patients) and six-variable (all 101 [100%] patients) MDRD equation analyses. No patient with a creatinine level of 1.5–1.9 mg/dL had an estimated GFR lower than 30 mL/min/1.73 m2 at four-variable MDRD equation analysis. Six patients (5.9%) had an estimated GFR lower than 30 mL/min/1.73 m2 at six-variable MDRD equation analysis.

Data Abstraction
There were no discrepancies regarding patient age, sex, or race between the abstracted and retrospectively measured data. In one discrepant case, the creatinine level category (>1.4 vs ≤1.4 mg/dL) of one (2.1%) of the 48 patients changed: Creatinine data for days 1 and 7 before CT were missing from the analyzed data set; the creatinine level was 1.5 mg/dL on both days. The analyzed data set rather included a creatinine value of 1.4 mg/dL for 24 days prior to CT. There were two (4.2%) discrepancies in which the patient's estimated GFR category (<60 vs ≥60 mL/min/1.73 m2) changed: The analyzed data set included slightly different creatinine values (1.4 mg/dL rather than abstracted value of 1.2 mg/dL and 1.6 mg/dL rather than abstracted value of 1.5 mg/dL) and thus yielded slightly different estimated GFR values (54 vs 64 mL/min/1.73 m2 and 59 vs 63 mL/min/1.73 m2, respectively).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
CIN is a major cause of morbidity and mortality (16). However, what constitutes an acceptable risk of CIN has not been defined. There are numerous articles, editorials, and commentaries regarding the incidence and pathogenesis of CIN and the available prevention strategies. Investigators in the majority of clinical studies have evaluated differences in the incidence of CIN between high- and low-osmolar contrast media, between low-osmolar and iso-osmolar contrast material, and between different types of low-osmolar contrast agents (1,2,6). However, most of these published trials involved intraarterial contrast material injections (17). Only a minority of the published clinical trials have involved the examination of patients who received intravenous contrast material injections, and an even smaller number of studies have included control groups of patients who did not receive contrast material. With such limitations, the issue of which patients actually develop nephropathy in association with intravenous contrast material administration, as opposed to other morbidities, is even less clear (17,18). Only rarely have researchers critically evaluated the practice guidelines, which are decidedly inconsistent among institutions (7).

The risk factors for CIN have been addressed and the prevention strategies have been compared in many publications (19–22). Diabetes and preexisting CKD, particularly when these coexist, represent the biggest risk factors for developing CIN (1,2). Therefore, it is critically important to identify patients with CKD before administering contrast medium.

Serum creatinine measurement is not standardized among different laboratories, so these values are often poor measures of CKD (23–25). In one study, 15% of patients with a normal creatinine level had a creatinine clearance of less than 50 mL/min when the Cockcroft-Gault (CG) equation was used (26). Furthermore, the relationships between serum creatinine level and patient age, sex, and race are poorly understood. Nonnephrologists do not identify CKD in patients in large part because of their inability to interpret serum creatinine levels in the context of the patient's age, sex, and race (27). The National Kidney Foundation now recommends using estimated GFR values to screen for and stage CKD. Patients with stage 1 CKD have a GFR of greater than or equal to 90 mL/min/1.73 m2 without reduced renal function. Patients with stage 2 CKD have a GFR of 60–89 mL/min/1.73 m2, which corresponds to mildly reduced renal function. Patients with stage 3 CKD have a GFR of 30–59 mL/min/1.73 m2, which corresponds to moderately reduced renal function. Patients with stage 4 CKD have a GFR of 15–29 mL/min/1.73 m2, which corresponds to severely reduced renal function. Patients with stage 5 CKD have a GFR lower than 15 mL/min/1.73 m2, which corresponds to renal failure. Patients with stage 3 CKD are considered to have a greater than 50% reduction in renal function (8).

If serum creatinine level is no longer recommended as the single best marker of renal function, it is imperative that radiologists understand how using the GFR—the estimated GFR especially—will affect the identification of patients who are considered to have CKD and thus may require renal protection strategies before contrast medium administration. However, to our knowledge, there are no data yet to suggest that we will be able to reduce the incidence of CIN by using estimated GFR values instead of serum creatinine levels.

Estimated GFR is a far more practical tool than measured creatinine for identifying patients with CKD in a screening setting, such as identifying patients at risk for CIN before CT scanning. Estimated GFR is not an ideal parameter, but it is considered a better screening tool than creatinine level alone (28). Both the CG equation and the MDRD equation are commonly used to estimate renal function with use of creatinine clearance and GFR values, respectively (10,29). In estimating renal function, one attempts to adjust creatinine levels for muscle mass and patient age, which is an independent risk for CKD that is factored into the CG and MDRD equations (30). In addition to patient age, patient weight is factored into the CG equation as a surrogate marker of muscle mass, and for all females, a constant factor of 0.85 is used to adjust creatinine levels for the lower mass. With the MDRD equation, a combination of the sex and the race of the patient is used to account for differences in muscle mass by applying data derived from a cohort of subjects known to have CKD. It is important to note that the patient's weight is not needed to apply the MDRD equation because this variable is modeled into the mathematical equation; hence, the result is reported in milliliters per minute per 1.73 m2. It is noteworthy that both the CG equation and the MDRD equation are empirically derived and thus have limitations; however, the MDRD equation is considered more accurate than both the CG equation and the measured creatinine clearance (9).

Although the MDRD equation is more accurate than the CG equation for estimating GFR in patients with CKD, neither formula is particularly accurate for estimations in healthy patients (31). The MDRD is more precise and accurate than the CG equation for estimating low GFR values, which are found in populations at risk for CIN (32). This is why we used the MDRD formula in our study. However, this equation is also not accurate in acutely ill hospitalized patients and therefore may not be a good screening tool for GFR in this population (33). Thus, we did not include inpatients in our study. Furthermore, the CG equation was developed by using an inpatient population and therefore was not appropriate for use in our study. Other parameters used to estimate the GFR include level of serum cystatin C, a low-molecular-weight cysteine proteinase, but this is not in widespread use (34).

Although there are a few published studies on the incidence of CIN in patients who receive intravenous injections of contrast medium (17), none of these address the number of patients with risk factors for CIN who present for CT scanning or how using the estimated GFR rather than the creatinine level could alter treatment management (17,35,36). Our study results show that a significantly larger number of patients meet the criteria for moderately reduced renal function (stage 3 or greater CKD) when the MDRD equation is used to estimate the GFR compared with the number of patients who meet the criteria on the basis of mildly or moderately elevated creatinine levels. This was true regardless of the race or sex of the patient. We found that approximately 2.5 times as many patients met the GFR-based criterion for reduced renal function—an estimated GFR lower than 60 mL/min/1.73 m2—than had an elevated serum creatinine level (>1.4 mg/dL). Similar to Duncan et al (26), we identified nearly 10% of patients with normal serum creatinine levels who had moderately reduced renal function by using estimated GFR values calculated with the MDRD equations. Restated, nearly 10% of the patients who initially were considered not to have an increased risk for CIN actually had stage 3 CKD. This is a potentially serious problem and suggests that using estimated GFR values may be a better method of identifying those at risk for CIN.

There were limitations to our study. First and foremost, data retrieval from the electronic medical records was limited to discrete data points around the time of the CT examination. Thus, we were not able to practically exclude patients with normal creatinine values who were undergoing hemodialysis or peritoneal dialysis; however, in actual practice, this constitutes a small percentage of our outpatients. Also, we have not yet studied the changes in renal function after CT examination to determine whether we can better identify those who will develop CIN with estimated GFR values or with creatinine levels. The number of patients who go on to develop CIN and the question of how to define CIN by using GFR values are the subjects of an ongoing investigation at our institution. Furthermore, data abstraction revealed that electronically derived data are not 100% accurate.

We also did not review the medical records of the patients who did not have available creatinine values to determine if any adverse effects on renal function occurred after intravenous contrast medium administration. However, this would have been impractical in a patient population this large and was not the primary purpose of our study. Finally, we did not include all patients who presented for CT scanning, regardless of whether they received contrast medium, because we had no way of reliably determining how many patients underwent CT without contrast material because of a specific indication (eg, sinus CT or head CT for intracranial bleeding) or a history of contrast medium allergy and thus would have been excluded or how many patients did not receive contrast medium because of elevated creatinine levels and thus would have been included. Many patients in our study did not have a serum creatinine sample taken before undergoing CT, but this is standard clinical practice and should not be considered a major limitation (3,8,13,37).

In summary, our study findings show that a significantly larger number of outpatients, approximately 2.5 times, would be considered to have moderately reduced renal function if estimated GFR values rather than creatinine levels were used to screen for renal insufficiency before CT scanning. This was true in our study, regardless of the patient's race or sex. Further study is imperative to determine the incidence and consequences of CIN in outpatient CT populations, whether estimated GFR is a better predictor of CIN than creatinine level, and what level of reduced renal function indicates an unacceptably elevated risk of CIN.


   APPENDIX
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
The four-variable MDRD equation used to estimate the GFR is as follows: GFR = 186 · Cr–1.154 · A–0.203 · 0.742 (if female) · 1.212 (if African American), where Cr is serum creatinine level and A is patient age.

The six-variable MDRD equation used to estimate the GFR is as follows: GFR = 170 · Cr–0.999 · A–0.176 · 0.762 (if female) · 1.18 (if African American) · BUN–0.170 · ALB0.318, where ALB is albumin level.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   ACKNOWLEDGMENTS
 
The authors thank Roseann Spitznagel, BS, for data compilation and James Walter, BA, for assistance with data extraction and chart and figure preparation.


   FOOTNOTES
 

Abbreviations: BUN = blood urea nitrogen • CG = Cockcroft-Gault • CIN = contrast medium–induced nephropathy • CKD = chronic kidney disease • GFR = glomerular filtration rate • MDRD = Modification of Diet in Renal Disease

Author contributions: Guarantors of integrity of entire study, B.R.H., E.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, B.R.H., E.S., M.E.B.; clinical studies, B.R.H., E.S.; statistical analysis, N.A.O.; and manuscript editing, all authors

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. Parfrey PS, Griffiths SM, Barrett BJ, et al. Contrast material-induced renal failure in patients with diabetes mellitus, renal insufficiency or both. N Engl J Med 1989;320:143–149.[Abstract]
  2. Lautin EM, Freeman NJ, Schoenfeld AH, et al. Radiocontrast-associated renal dysfunction: incidence and risk factors. AJR Am J Roentgenol 1991;157:49–58.[Abstract/Free Full Text]
  3. American College of Radiology manual on contrast media, version 5.0. American College of Radiology Web site. http://www.acr.org/SecondaryMainMenuCategories/quality_safety/contrast_manual.aspx.
  4. Gleeson TG, Bulugahahpitiya S. Contrast-induced nephropathy. AJR Am J Roentgenol 2004;183:1673–1689.[Free Full Text]
  5. Persson PB. Contrast-induced nephropathy. Eur Radiol 2005;15(suppl 4):D65–D69.[CrossRef][Medline]
  6. Barrett BJ, Katzberg RW, Thomsen HS, et al. Contrast-induced nephropathy in patients with chronic kidney disease undergoing computed tomography: a double-blind comparison of iodixanol and iopamidol. Invest Radiol 2006;41:815–821.[CrossRef][Medline]
  7. Elicker BM, Cypel YS, Weinreb JC. IV contrast administration for CT: a survey of practices for the screening and prevention of contrast nephropathy. AJR Am J Roentgenol 2006;186:1651–1658.[Abstract/Free Full Text]
  8. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Ann Intern Med 2003;139:137–147.[Abstract/Free Full Text]
  9. Lin J, Bonventre JV. Prevention of radiocontrast nephropathy. Curr Opin Nephrol Hypertens 2005;14:105–110.[Medline]
  10. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41.[Medline]
  11. Bostom AG, Kronenberg F, Ritz E. Predictive performance of renal function equations for patients with chronic kidney disease and normal serum creatinine levels. J Am Soc Nephrol 2002;13:2140–2144.[Abstract/Free Full Text]
  12. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function: measured and estimated glomerular filtration rate. N Engl J Med 2006;354:2473–2483.[Free Full Text]
  13. Tippins RB, Torres WE, Baumgartner BR, Baumgarten DA. Are screening serum creatinine levels necessary prior to outpatient CT examinations? Radiology 2000;216:481–484.[Abstract/Free Full Text]
  14. Choyke PL, Cady J, DePollar SL, Austin H. Determination of serum creatinine prior to iodinated contrast media: is it necessary in all patients? Tech Urol 1998;4:65–69.[Medline]
  15. Benko A, Fraser-Hill M, Magner P, et al. Canadian Association of Radiologists: guidelines for the prevention of contrast induced nephropathy. Canadian Association of Radiologists Web site. http://www.car.ca/Files/Nephropathy.pdf.
  16. McCullough P. Outcomes of contrast-induced nephropathy: experience in patients undergoing cardiovascular intervention. Catheter Cardiovasc Interv 2006;67:335–343.[CrossRef][Medline]
  17. Rao QA, Newhouse JH. Risk of nephropathy after intravenous administration of contrast material: a critical literature analysis. Radiology 2006;239:392–397.[Abstract/Free Full Text]
  18. Katzberg RW, Barrett BJ. Risk of iodinated contrast material–induced nephropathy with intravenous administration. Radiology 2007;243:622–628.[Free Full Text]
  19. Barrett BJ, Carlisle EJ. Meta-analysis of the relative nephrotoxicity of high- and low-osmolality iodinated contrast media. Radiology 1993;188:171–178.[Abstract/Free Full Text]
  20. Louis BM, Hoch BS, Hernandez C, et al. Protection from the nephrotoxicity of contrast dye. Renal Failure 1996;18:639–646.[Medline]
  21. Garcia-Ruiz C, Martinez-Vea A, Sempere T, et al. Low risk of contrast nephropathy in high-risk patients undergoing spiral computed tomography angiography with the contrast medium iopromide and prophylactic oral hydration. Clin Nephrol 2004;61:170–176.[Medline]
  22. Waybill MM, Waybill PN. Contrast media-induced nephrotoxicity: identification of patients at risk and algorithms for prevention. J Vasc Interv Radiol 2001;12:3–9.[Medline]
  23. Ashley JB, Milleard SF. Contrast agent-induced nephropathy: a simple way to identify patients with preexisting renal insufficiency. AJR Am J Roentgenol 2003;181:451–454.[Free Full Text]
  24. Swedko PJ, Clark HD, Paramsothy K, Akbari A. Serum creatinine is an inadequate screening test for renal failure in elderly patients. Arch Intern Med 2003;163:356–360.[Abstract/Free Full Text]
  25. Myers GL, Miller WG, Coresh J, et al. Recommendations for improving serum creatinine measurement: a report from the laboratory working group of the national kidney disease education program. Clin Chem 2006;52:5–18.[Abstract/Free Full Text]
  26. Duncan L, Heathcote J, Djurdjev O, Levin A. Screening for renal disease using serum creatinine: who are we missing? Nephrol Dial Transplant 2001;16:1042–1046.[Abstract/Free Full Text]
  27. Levey AS, Stevens LA, Hostetter T. Automatic reporting of estimated glomerular filtration rate: just what the doctor ordered. Clin Chem 2006;52:2188–2193.[Free Full Text]
  28. Poggio ED, Hall PM. Estimation of glomerular filtration rate by creatinine-based formulas: any room for improvement? [editorial]. NephSAP 2006;5:131–140.
  29. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461–470.[Abstract/Free Full Text]
  30. Detrenis S, Mechi M, Bertolini L, Savazzi G. Contrast medium administration in the elderly patient: is advancing age an independent risk factor for contrast nephropathy after procedures? J Vasc Interv Radiol 2007;18:177–185.[CrossRef][Medline]
  31. Poggio ED, Wang X, Greene T, et al. Performance of the modification of diet in renal disease and Cockcroft-Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol 2005;16:459–466.[Abstract/Free Full Text]
  32. Cirillo M, Anastasio P, De Santo NG. Relationship of gender, age, and body mass index to errors in predicted kidney function. Nephrol Dial Transplant 2005;20:1791–1798.[Abstract/Free Full Text]
  33. Poggio ED, Nef PC, Wang X, et al. Performance of the Cockcroft-Gault and modification of diet in renal disease equations in estimating GFR in ill hospitalized patients. Am J Kidney Dis 2005;46:242–252.[CrossRef][Medline]
  34. Dharnidharka VR, Kwon C, Stevens G. Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 2002;40:221–226.[CrossRef][Medline]
  35. Mitchell AM, Kline JA. Contrast nephropathy following computed tomography angiography of the chest for pulmonary embolism in the emergency department. J Thromb Haemost 2007;5:50–54.[CrossRef][Medline]
  36. Josephson SA, Dillon WP, Smith WS. Incidence of contrast nephropathy from cerebral CT angiography and CT perfusion imaging. Neurology 2005;64:1805–1806.[Abstract/Free Full Text]
  37. Katzberg RW, Haller C. Contrast-induced nephrotoxicity: clinical landscape. Kidney Int Suppl 2006;100:S3–S7.[Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
J. H. Ellis and R. H. Cohan
Reducing the Risk of Contrast-Induced Nephropathy: A Perspective on the Controversies
Am. J. Roentgenol., June 1, 2009; 192(6): 1544 - 1549.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow All Versions of this Article:
2481071528v1
248/1/106    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Herts, B. R.
Right arrow Articles by Baker, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Herts, B. R.
Right arrow Articles by Baker, M. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE