Original Contribution

Impact of Contrast Dose Reduction on Incidence of Acute Kidney Injury (AKI) Among Patients Undergoing PCI: A Modeling Study

Hitinder S. Gurm, MD1;  Milan Seth, MS1;  Roxana Mehran, MD2;  Louis Cannon, MD3;  Cindy L. Grines, MD4;  Thomas LaLonde, MD5;  Carlos Briguori, MD6; for the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2)

Hitinder S. Gurm, MD1;  Milan Seth, MS1;  Roxana Mehran, MD2;  Louis Cannon, MD3;  Cindy L. Grines, MD4;  Thomas LaLonde, MD5;  Carlos Briguori, MD6; for the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2)

Abstract: Objective. Acute kidney injury is a common complication associated with angiography and percutaneous coronary intervention (PCI). Increasing doses of contrast are associated with an increase in the likelihood of AKI. The objective of our study was to estimate projected reduction in the burden of AKI in association with varying degrees of contrast media dose reduction among patients undergoing PCI. Methods. We assessed the relationship between contrast volume to creatinine clearance among consecutive patients undergoing PCI in the state of Michigan between January 2010 and September 2013. Computational modeling was used to estimate the anticipated reduction in risk of AKI across varying degrees of reduction in contrast volume. Results. The risk of AKI was significantly and substantially increased in patients in whom the contrast dose exceeded 2.99 times the creatinine clearance. The benefit of contrast dose reduction was most evident in those at greater predicted risk of AKI. An across the board 30% reduction in contrast dose would be expected to prevent one-eighth of AKI cases, although clinical benefits could also be anticipated with smaller dose reductions. Conclusion. Our study provides estimates of reduction in AKI that could be achieved with contrast dose reduction in clinical practice. These data should help guide planning of clinical trials and the application of contrast-saving strategies to routine clinical practice.

J INVASIVE CARDIOL 2016;28(4):142-146. Epub 2016 January 15.

Key words: contrast media, acute kidney injury, percutaneous coronary intervention


Contrast-induced nephropathy (CIN) is a common complication associated with invasive cardiac procedures and is strongly associated with morbidity, mortality, and significantly increased health-care expenses.1-4 CIN is the most common cause of acute kidney injury (AKI) among hospitalized patients. While AKI can be multifactorial in patients undergoing PCI, contrast media directly cause or contribute significantly to the risk of developing AKI among patients undergoing PCI.5

The toxicity of contrast media appears to be directly related to the amount of contrast media that is delivered to the renal tubule and there appears to be a non-linear relationship between contrast dose and risk of AKI in patients undergoing PCI.6,7 Furthermore, measures that reduce tubular exposure to contrast media, such as aggressive hydration or reducing contrast dose, are the only measures that have consistently demonstrated efficacy in reducing the development of AKI.3,8

Based on the non-linear relationship of contrast dose and the risk of AKI, a number of approaches have been advocated to significantly reduce the total contrast dose among patients undergoing PCI.9,10 These approaches have varied in complexity and the associated risk to the patient and it is unclear how these efforts will be incorporated into routine clinical practice. Furthermore, it is unclear as to the degree of contrast reduction that would be needed for a device or strategy to be clinically meaningful. We accordingly developed a computational model to assess the impact of varying degrees of contrast reduction on the occurrence of AKI.


Study population. We included data from patients undergoing PCI at 47 hospitals participating in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2). The details of the BMC2 registry and its data collection and auditing process have been described previously.11-13 Briefly, procedural data on all patients undergoing PCI at participating hospitals are collected using standardized data collection forms. Baseline data include clinical, demographic, procedural, and angiographic characteristics, as well as medications used before, during, and after the procedure, and in-hospital outcomes. All data elements were prospectively defined, and the protocol was approved by local institutional review boards at each hospital. In addition to a random audit of 2% of all cases, medical records of all patients undergoing multiple procedures or coronary artery bypass grafting and of patients who died in the hospital are reviewed routinely to ensure data accuracy. 

The study population for this analysis included all consecutive patients who underwent PCI between January 1, 2010 and September 30, 2013. We excluded patients who were on dialysis at the time of the procedure, those missing preprocedural or postprocedural creatinine values, and those in whom creatinine clearance or baseline AKI risk could not be estimated due to missing information. We also excluded patients whose contract volume exceeded 10x the creatinine clearance since these patients are at extremely high risk of AKI and such contrast dosing usually reflects unusual clinical circumstances.

Study endpoints. Our study had two endpoints for this analysis: AKI and new need for dialysis. Acute kidney injury was defined as an elevation in serum creatinine of ≥0.5 mg/dL. We (and others) have previously demonstrated this definition to be preferable to more sensitive definitions of AKI.14-16 

Statistical analysis. Creatinine clearance (CrCl) was estimated using the Cockgroft Gault equation.17 Patient preprocedural risk of AKI was estimated using the BMC2 random forest prediction model, which estimates baseline risk from 21 preprocedural clinical, laboratory, and demographic variables. The methodology used to develop the models has been described previously,18 and model estimates can be evaluated for sample patients using the online calculator at https://bmc2.org/calculators/multi or with a downloadable free app at http://scaipciriskapp.org/pci_welcome.

Contrast volume (CV) was indexed to CrCl, and the added effect of CV on risk of in-hospital mortality was assessed using logistic regression models, adjusting for estimated preprocedural AKI risk along with preprocedural covariates age, gender, PCI indication, coronary artery disease presentation, cardiogenic shock, cardiac arrest, heart failure, diabetes, and CrCl. CV (indexed to CrCl) was included in this model using a linear spline, with knots determined by deciles of indexed CV in the cohort. Spline coding was utilized to address potentially non-linear relationships between contrast dosage and outcome risk. Likelihood ratio tests were used to assess the association between CV and in-hospital outcomes, and whether spline regression significantly improved the model fit compared with a model assuming the effect of contrast was linear across its range.

The regression models were then used to estimate the expected reduction in outcome incidence resulting from an across-the-board percent reduction in CV applied to all patients or for a subset of patients defined by predicted baseline AKI risk >1% or >3%. The impact of contrast reduction ranging from 5%-50% by 5% increments was evaluated. Using the coefficient estimates and variance-covariance matrix from the fitted models, the linear predictor value for each patient was estimated along with its standard error assuming a given across-the-board reduction in contrast dose or for a given reduction value applied only to those patients with baseline risk estimates above a cut-off (1% or 2%). The total expected number of AKI cases or nephropathy requiring dialysis (NRD) cases was then estimated by summing the inverse logit of the patient-level linear predictor values, and 95% confidence limits for total cases were obtained by summing over the inverse logit of the upper and lower 95% confidence limits for the patient-level linear predictor values). These cohort-level estimates of outcome incidence rates after a hypothetical intervention were used to estimate the number needed to treat (NNT) for each intervention.


A total of 113,587 patients underwent PCI between January 1, 2010 and September 30, 2013. A total of 2830 patients (2.5%) were on dialysis at the time of the procedure, and 14,144 patients (12.5%) were missing preprocedural or post procedural creatinine. In 557 patients (0.5%), creatinine clearance or baseline outcome risk could not be estimated due to other missing demographic or clinical data elements and 431 patients (0.38%) had CV levels >10x creatinine clearance. After exclusions, the analysis cohort consisted of 95,625 patients (84.2%). In this cohort, 2976 cases (3.1%) resulted in AKI, and 324 cases (0.34%) required dialysis. Deciles of CV/CrCl with the unadjusted AKI and NRD rates are provided in Table 1. 

After adjusting for baseline characteristics and predicted risk, CV/CrCl was significantly associated with both AKI (LRT statistic, 193.5 on 11 df; P<.001) and NRD (LRT statistic 52.1 on 11 df; P<.001). Furthermore, for the AKI outcome, the spline regression model significantly improved the model fit compared with a simpler model adjusting for all other included covariates but assuming a simple linear relationship between CV/CrCl and outcome risk (LRT statistic, 27.49 on 10 df; P=.01); however, this was not the case for NRD (LRT statistic, 10.37 on 10 df; P=.41).  

Model-estimated odds ratios of AKI relative to the median value of 2.05 and 95% confidence intervals at quantiles of CV/CrCl from the 5th to 95th percentile are plotted in Figure 1. Estimated AKI odds ratios increase with CV/CrCl throughout the range of values, with confidence intervals around estimates for CV/CrCl values greater than the 3rd quartile of 2.99 entirely above 1.0, demonstrating a statistically significant increase in risk of AKI. The low incidence of NRD rendered such analysis unreliable, and hence these were not further explored.

Figure 2 provides a plot of model-estimated expected percent reductions in AKI incidence for hypothetical contrast-sparing interventions over a range of effectiveness (percent reduction in CV) applied to all patients and to subgroups of patients defined by baseline predicted risk values (>1%, >2%). The expected number of patients who would need to be treated with the intervention to prevent 1 case of AKI (NNT) is plotted in Figure 3.


The key finding of our study is that an across-the-board reduction in CV would be associated with a reduction in AKI and dialysis, with the benefit being proportional to the magnitude of the reduction. Furthermore, a risk-based approach, whereby the contrast reduction strategy is applied selectively to those who are at high risk for AKI would allow optimal utilization of such strategies. 

Contrast media are one of the leading causes of AKI among hospitalized patients, and the only strategies that have been demonstrated to reduce the occurrence of AKI are the choice of contrast media and hydration.19-21 The association of high contrast dose with the risk of AKI has been previously demonstrated, and there appears to be a non-linear relation between contrast dose and the risk of AKI. Professional societies and contrast media manufacturers advocate a principle of minimizing contrast media use.22,23 More importantly, most cardiologists understand the need to balance optimal angiographic visualization against excessive contrast administration.24 

Methods of limiting contrast media use include foregoing left ventriculography or aortography, use of biplane angiography, and in select cases, use of aggressive contrast minimization techniques and use of fluoroscopy and IVUS to guide interventional procedures.25,26 In addition, there have been efforts to use device-based strategies that either bind contrast media, remove contrast media from the coronary sinus, or limit excessive reflux of contrast media from guiding catheters in the aorta.27-34 These device-based strategies will imply varying degrees of risk and introduce varying degrees of complexity to the procedure. Furthermore, the amount of reduction in CV that can be achieved with these devices will vary. 

Our study would suggest that there is a nearly linear relationship between the magnitude of contrast reduction and the expected reduction in AKI. Secondly, the AKI reduction, while clinically important, would be of small magnitude and devices demonstrating small reduction in contrast use would need to be studied in extremely large studies before a statistically meaningful reduction in AKI could be demonstrated. Our study would suggest that a reduction of 30% in CV could reliably be expected to translate into a 12.8% reduction in AKI, while a 20% reduction would be associated with an 8.8% reduction in AKI, although smaller degrees of CV reduction would also be anticipated to demonstrate clinically relevant benefits.

The use of these devices would likely be most impactful in those at the highest risk for AKI. Use of the device that reduces contrast by 30% among the patients with a predicted AKI risk of >1% would require treatment of <37% of patients and would achieve 77.1% of the benefit of treating the entire population. Similarly, the application of this strategy to those with a risk >2% would treat just about one-quarter of the population and achieve 69.2% of the benefit of treating the entire population. Thus, the cost effectiveness of these strategies would vary depending on the population to which they are applied, and while some strategies such as biplane angiography or modification of procedural approach (such as avoiding ventriculography) would almost always be cost effective for any patient, the cost effectiveness of any novel therapies would vary based on the population in which it is used.

Study limitations. The results of our study should be interpreted with certain caveats. This is a simulation study and the results of this exercise assume that only CV would be impacted and other factors would remain unchanged. That is unlikely in a clinical trial or in clinical practice. However, our data should provide a reasonable estimate of the projected benefits of contrast reduction and could guide planning of clinical trials. Secondly, we used a highly specific definition of AKI that indicates more serious renal injury and is thus more strongly associated with risk of death or need for dialysis. Most clinical trials use a more sensitive definition of AKI, but the clinical benefit of using such definitions remains to be established. Finally, our work is derived from patients undergoing procedures across the entire state and the hydration and other prophylactic measures were not standardized. This, however, increases the generalizability of our findings to routine clinical practice.


Our study provides a projection of the anticipated benefit of strategies to reduce the dose of contrast media administered to patients undergoing PCI. These estimates should help guide ongoing and future clinical trials, and might be helpful in targeting clinical application of these devices in clinical practice. 


1.    Rihal CS, Textor SC, Grill DE, et al. Incidence and prognostic importance of acute renal failure after percutaneous coronary intervention. Circulation. 2002;105:2259-2264.

2.    Best PJ, Lennon R, Ting HH et al. The impact of renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary interventions. J Am Coll Cardiol. 2002;39:1113-1119.

3.    Howe M, Gurm HS. A practical approach to preventing renal complications in the catheterization laboratory. Interv Cardiol Clin. 2014;3:429-439.

4.    McCullough PA. Contrast-induced acute kidney injury. J Am Coll Cardiol. 2008;51:1419-1428.

5.    Geenen RW, Kingma HJ, van der Molen AJ. Pathophysiology of contrast-induced acute kidney injury. Interv Cardiol Clin. 2014;3:363-367.

6.    Gurm HS, Dixon SR, Smith DE, et al. Renal function-based contrast dosing to define safe limits of radiographic contrast media in patients undergoing percutaneous coronary interventions. J Am Coll Cardiol. 2011;58:907-914.

7.    Lauver DA, Carey EG, Bergin IL, Lucchesi BR, Gurm HS. Sildenafil citrate for prophylaxis of nephropathy in an animal model of contrast-induced acute kidney injury. PloS One. 2014;9:e113598.

8.    Rojkovskiy I, Solomon R. Intravenous and oral hydration: approaches, principles, and differing regimens. Interv Cardiol Clin. 2014;3:393-404.

9.    Stub D, Duffy SJ, Kaye DM. Device-based therapy in the prevention of contrast-induced nephropathy. Interv Cardiol Clin. 2014;3:421-428.

10.    Nayak KR, Mehta HS, Price MJ, et al. A novel technique for ultra-low contrast administration during angiography or intervention. Catheter Cardiovasc Interv. 2010;75:1076-1083.

11.    Moscucci M, Share D, Kline-Rogers E, et al. The Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) collaborative quality improvement initiative in percutaneous coronary interventions. J Interv Cardiol. 2002;15:381-386.

12.    Share DA, Campbell DA, Birkmeyer N, et al. How a regional collaborative of hospitals and physicians in Michigan cut costs and improved the quality of care. Health Aff (Millwood). 2011;30:636-645.

13.    Moscucci M, Rogers EK, Montoye C, et al. Association of a continuous quality improvement initiative with practice and outcome variations of contemporary percutaneous coronary interventions. Circulation. 2006;113:814-822.

14.    Slocum NK, Grossman PM, Moscucci M, et al. The changing definition of contrast-induced nephropathy and its clinical implications: insights from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2). Am Heart J. 2012;163:829-834.

15.    Harjai KJ, Raizada A, Shenoy C, et al. A comparison of contemporary definitions of contrast nephropathy in patients undergoing percutaneous coronary intervention and a proposal for a novel nephropathy grading system. Am J Cardiol. 2008;101:812-819.

16.    McCullough PA. Contrast-induced nephropathy: definitions, epidemiology, and implications. Interv Cardiol Clin. 2014;3:357-362.

17.    Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31-41.

18.    Gurm HS, Seth M, Kooiman J, Share D. A novel tool for reliable and accurate prediction of renal complications in patients undergoing percutaneous coronary intervention. J Am Coll Cardiol. 2013;61:2242-2248.

19.    Reed M, Meier P, Tamhane UU, Welch KB, Moscucci M, Gurm HS. The relative renal safety of iodixanol compared with low-osmolar contrast media: a meta-analysis of randomized controlled trials. JACC Cardiovasc Interv. 2009;2:645-654.

20.    Meier P, Ko DT, Tamura A, Tamhane U, Gurm HS. Sodium bicarbonate-based hydration prevents contrast-induced nephropathy: a meta-analysis. BMC Med. 2009;7:23.

21.    Solomon R, Werner C, Mann D, D’Elia J, Silva P. Effects of saline, mannitol, and furosemide to prevent acute decreases in renal function induced by radiocontrast agents. N Engl J Med. 1994;331:1416-1420.

22.    Nallamothu BK, Tommaso CL, Anderson HV, et al. ACC/AHA/SCAI/AMA-convened PCPI/NCQA 2013 performance measures for adults undergoing percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures, the Society for Cardiovascular Angiography and Interventions, the American Medical Association-Convened Physician Consortium for Performance Improvement, and the National Committee for Quality Assurance. J Am Coll Cardiol. 2014;63:722-745.

23.    Bashore TM, Balter S, Barac A, et al. 2012 American College of Cardiology Foundation/Society for Cardiovascular Angiography and Interventions expert consensus document on cardiac catheterization laboratory standards update: a report of the American College of Cardiology Foundation Task Force on Expert Consensus documents developed in collaboration with the Society of Thoracic Surgeons and Society for Vascular Medicine. J Am Coll Cardiol. 2012;59:2221-2305.

24.    Mavromatis K. The imperative of reducing contrast dose in percutaneous coronary intervention. JACC Cardiovasc Interv. 2014;7:1294-1296. Epub 2014 Nov 17.

25.    Ogata N, Matsukage T, Toda E, et al. Intravascular ultrasound-guided percutaneous coronary interventions with minimum contrast volume for prevention of the radiocontrast-induced nephropathy: report of two cases. Cardiovasc Interv Ther. 2011;26:83-88.

26.    Mariani J Jr, Guedes C, Soares P, et al. Intravascular ultrasound guidance to minimize the use of iodine contrast in percutaneous coronary intervention: the MOZART (Minimizing cOntrast utiliZation With IVUS Guidance in coRonary angioplasTy) randomized controlled trial. JACC Cardiovasc Interv. 2014;7:1287-1293.

27.    Movahed MR, Wong J, Molloi S. Removal of iodine contrast from coronary sinus in swine during coronary angiography. J Am Coll Cardiol. 2006;47:465-467.

28.    Meyer M, Dauerman HL, Bell SP, Lewinter MM, Lustgarten DL. Coronary venous capture of contrast during angiography. J Interv Cardiol. 2006;19:401-404.

29.    Chang H, Hassan AH, Kim YL, et al. A novel technique for endovascular detection and removal of radiographic contrast during angiography. J Invasive Cardiol. 2009;21:314-318.

30.    Michishita I, Fujii Z. A novel contrast removal system from the coronary sinus using an adsorbing column during coronary angiography in a porcine model. J Am Coll Cardiol. 2006;47:1866-1870.

31.    Danenberg HD, Lotan C, Varshitski B, Rosenheck S, Weiss AT. Removal of contrast medium from the coronary sinus during coronary angiography: feasibility of a simple and available technique for the prevention of nephropathy. Cardiovasc Revasc Med. 2008;9:9-13.

32.    Duffy SJ, Ruygrok P, Juergens CP, et al. Removal of contrast media from the coronary sinus attenuates renal injury after coronary angiography and intervention. J Am Coll Cardiol. 2010;56:525-526.

33.    Kaye DM, Stub D, Mak V, Doan T, Duffy SJ. Reducing iodinated contrast volume by manipulating injection pressure during coronary angiography. Catheter Cardiovasc Interv. 2014;83:741-745.

34.    Lamps G, Yadav J, Mizaikoff B. Radiographic contrast media capture material. Google Patents, 2008.


From the 1Department of Internal Medicine, Division of Cardiovascular Medicine, University of Michigan Medical Center, Ann Arbor, Michigan; 2Mt. Sinai Hospital, New York, New York; 3McLaren-Northern Michigan Regional Hospital, Cheboygan, Michigan; 4Harper-Hutzel Hospital, Detroit, Michigan; 5St. John-Providence Medical Center, Detroit, Michigan; and 6Clinica Mediterranea, Naples, Italy.    

Funding: Support for BMC2 is provided by Blue Cross and Blue Shield of Michigan and Blue Care Network as part of the BCBSM Value Partnerships program.

Disclaimer: Although Blue Cross Blue Shield of Michigan and BMC2 work collaboratively, the opinions, beliefs, and viewpoints expressed by the authors do not necessarily reflect the opinions, beliefs, and viewpoints of BCBSM or any of its employees.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Gurm reports research funding from Blue Cross Blue Shield of Michigan and the National Institutes of Health; consultant fees from Osprey Medical. Dr Mehran reports research grant support from Eli Lilly/DSI, AstraZeneca, the Medicines Company, BMS, and OrbusNeich; consulting fees from Janssen Pharmaceuticals, Medscape, Osprey Medical, and Watermark Research Partners. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted September 25, 2015, provisional acceptance given September 29, 2015, final version accepted October 12, 2015.

Address for correspondence: Hitinder S. Gurm, MD, University of Michigan Cardiovascular Center, 2A394, 1500 E. Medical Center Drive, Ann Arbor, MI 48109-5853. Email: hgurm@med.umich.edu