Using contrast in CT scans has been thought to cause acute kidney injury, renal replacement therapy and even death. But is this just a case of confounding bias?
Using contrast in CT scans has been thought to cause acute kidney injury, renal replacement therapy and even death. But is this just a case of confounding bias?
Should emergency physicians be a little concerned, very concerned or not concerned at all about Contrast-Induced Acute Kidney Injury (CI-AKI)?
From previous studies, here’s what we think we know today: The use of contrast-enhanced CT scans is a completely iatrogenic cause of renal failure, recognized as one of the most common causes (between 11%-12%) of hospital-acquired Acute Kidney Injury (AKI). Acute Renal Replacement Therapy (RRT) is then required for between 3%-15% of CI-AKI patients [1,2,3,4,5,6,7]. In addition, hospitalized patients complicated by CI-AKI, compared to patients with similar pre-morbid characteristics, experience longer hospital stays, and suffer significantly higher in-hospital and long-term mortality rates [5,6,7,8,9,10,11]. Those patients who survive CI-AKI then are predisposed to continued loss of kidney function [12,13,14,15].
So, based on this data, the best answer to the leading question would probably be “very concerned.” But let’s take a closer look at the evidence behind the causal chain linking contrast-enhanced CT Scans with AKI, progression to Chronic Kidney Disease (CKD), the need for RRT and the potential possibility of death to determine whether that answer really is correct.
Contrast-Induced Acute Kidney Injury is a syndrome defined by an absolute (0.5 mg/dl) or relative increase (25%) in Serum Creatinine (SCr) 48-72 hours after intravenous or intra-arterial iodinated contrast administration. As is the case in most toxic exposures, CI-AKI has been defined solely by observational studies showing an association between contrast-exposure and AKI. Observational studies for causal relationships are especially prone to confounding bias where hidden variables explain the causal link between predictor and outcomes better than the observed temporal association. Without a prospective randomized controlled trial of sufficient sample size with appropriate safeguards against biases, a definitive causal relationship between contrast-exposure and AKI cannot be automatically accepted.
Here’s an example of confounding bias: Blaming the canary for the observed deaths of coal miners. Check this out — before poisonous gases could be objectively detected the death of coal miners, a canary was used as an early warning signal of carbon monoxide poisoning. If you did not already know that it was the poisoned air that killed the canary before the coal miners, a non-informed observer may mistake the temporal relationship between the bird’s death and sick coal miners as cause-and-effect. One may then falsely blame the canary and not the invisible poisonous gas (confounding variable) as the cause of the coal miners’ illness. The paper would be dramatically and, of course erroneously, titled “Canary Illness Leads to Death of Coal Miners.”
Lipsitch et al stated that noncausal associations between outcomes and exposures are the result of either mismeasurement (recall bias), confounding or selection bias. To prevent confounding, Lipsitch et al suggested a counterfactual test by designing a negative control experiment where the observation is repeated under conditions not expected to produce the outcome of interest. If the outcome is encountered without the exposure, then a confounding bias may exist [16].
Rao & McDonald Studies: Counterfactual Methodology
A counterfactual methodology for testing the causal link between contrast and AKI would compare the incidence of AKI between those patients’ exposed and unexposed to intravenous contrast. Such a counterfactual methodology of studying CI-AKI was reported in systematic reviews of CI-AKI by Rao et al in 2006 and McDonald et al in 2013 [17,18]. These two systematic reviews found no statistically significant difference in the incidence of AKI in contrast exposed compared to unexposed patients. The majority of the studies reviewed in these two systematic reviews mixed inpatients with ED patients with and without CKD. The McDonald study has been criticized, however, for not including earlier studies from the iodinated-contrast era when AKI was more common, and for conflicts of interest (several authors have ties to manufacturers of CT scanners, or contrast agents).
Sinert Study: ED Patients with Normal Baseline Renal Function
Sinert et al studied ED patients with normal baseline renal function (SCr < 1.5 mg/dl) comparing 773 contrast-exposed to 2,956 unexposed [19]. Using the conventional definition of CIAKI, unexposed compared to contrast-exposed patients had a significantly (P=0.003) higher incidence of AKI 8.96% vs. 5.69%, respectively.
From this data it would appear that intravenous contrast was not a risk for AKI, but actually protected against hospital-acquired AKI. The patients in the contrast-exposed and unexposed groups that developed AKI by the CI-AKI criteria had similar outcomes with respect for RRT requirements and mortality rates.
The findings of these studies suggest a significant risk of a confounding bias in the current definition of CI-AKI, based on timed changes of SCr. Since by definition CIAKI only occurs in patients with a pre-existing reason for contrast, the possibility that other etiologies beside contrast exposure are in part, or wholly, responsible for AKI. Confounding the risk of contrast causing AKI includes not only the primary disease that prompted the CT, but also pre-existing or intercurrent processes such as progression of CKD, hypovolemia, hypotension, sepsis or other nephrotoxic exposures (e.g., NSAID’s, antibiotics, toxins, etc.).
In other words, what’s injuring these patient’s kidneys may be their disease, or may still be iatrogenic, but not necessarily due to contrast exposure.
PROSPERO Study: Systemic Review of CI_AKI after CT Scans Planned
The PROSPERO study is a planned systematic review of CI-AKI after CT Scans will utilize a similar counterfactual methodology, including more recent studies by Davenport et al and McDonald et al that used propensity-matching analysis to reduce confounding bias of observational studies [20,21,22]. Both the Davenport et al and McDonald et al studies of patients with normal kidney function (SCr < 1.5 mg/dl) also failed to find a higher rate of CI-AKI in contrast-exposed compared to unexposed patients receiving CT-Scans.
The hope is that the PROSPERO study will finally dispel the causal link between contrast exposure with AKI, CKD, need for RRT or death. In a coal mine there is a direct causal link between the poisonous gases and the death of the canary, which alerts the coal miners to danger. In the ED, contrast exposure does not appear to have a direct causal link either to the increases in SCr, the need for RRT, or death.
Not Safe in All Patients: Future Studies Needed
The implications of questioning the definition of CI-AKI to define a specific disease are that it is safe to administer intravenous contrast to patients with normal renal function without the fear of AKI. Informed consent procedures discussing the risks of AKI after contrast may no longer be justified. This is not to say, however, that contrast is safe in patients with an elevated SCr or an estimated GFR less than 60 ml/min/1.73m2.
Future studies of CI-AKI in patients with baseline renal insufficiency should include suitable negative controls of unexposed patients with the addition of propensity matching for the conventional risks for AKI [23]. These studies must have appropriate power to detect clinically significant outcomes such as rates of RRT and mortality.
Richard Sinert, DO is professor of emergency medicine and vice-chair in charge of research, in the Department of Emergency Medicine, SUNY-Downstate Medical Center, Kings County Hospital Center
Editor’s Note: Another Perspective on the Risks and Benefits of Contrast
Research that suggests contrast isn’t so dangerous has intrinsic appeal to emergency physicians. After all, the pressure is on us to make the diagnosis – and we’re the ones negotiating with radiology to prioritize and protocol the CT (and in many cases, consent the patients for contrast). If AKI does happen, it’s comfortably removed from our environment – something that can be monitored or managed by someone else.
Professor Sinert asks, as we tend to associate contrast with increases in creatinine and renal injury, have we been falsely blaming the canary (contrast exposure), instead of some other factor, for the death of the coal miners (AKI)?
At the risk of extending the analogy too far: blaming the canary may be understandable, if we recall it used to be a large bird of prey. Furthermore, some of the folks who suggest we stop blaming the canary have a vested interest in the bird feeder business, and own pet stores.
The effect of contrast on nephropathy seemed more obvious in the era of high osmolar formulations; looking at observational studies, case-control studies and randomized trials from the 1970s and 1980s, contrast was more easily associated with nephrotoxicity than recently. Take a look at Older et al (in AJR 1976) or Krumlovsky (JAMA 1978) or Taliercio et al (Annals of Internal Med 1986) or Weinberg (Kidney Int 1994). Baseline characteristics and definitions of renal injury differ, but there’s a reason that “contrast-induced nephropathy” is an uncontroversial term.
Yet in the recent McDonald meta-analysis of 13 non-randomized trials, only 2 studies predate the year 2000 and no study was designed to assess the safety of IV contrast; they sought only to find a safer dose. McDonald’s control group exhibited selection bias (higher baseline creatinine in the non-contrast group). No studies of angiography were included, though evidence of a causal relationship is more clear with IA (and it’s hard to assert that IV contrast could be safe while IA is not – see the recent randomized, prospective CARE2 trial for example). Furthermore, a few of the authors had a relationship with GE Healthcare, which makes CT scanners and contrast material.
Placing too much stock in a meta-analysis of non-randomized and retrospective trials is risky; there are likely more confounders to retrospective studies, than potential confounders to the contrast/AKI relationship. Hopefully the planned PROSPERO systematic review will be able to include some of the older trials, with older contrast agents, that may not have been as well-conducted as modern trials but nonetheless defined “contrast-induced nephropathy.” In the meantime, though, it’s premature to stop discussing the risks and benefits of contrast to patients, especially to those with renal insufficiency.
Nicholas Genes, MD, PhD is a senior editor at EPM and a clinical informaticist at Mount Sinai Medical Center
REFERENCES
1. Nash K, Hafeez A, Hou S. Hospital-acquired renal insufficiency. Am J Kidney Dis 2002;39:930-6.
2. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-acquired renal insufficiency: a prospective study. Am J Med 1983;74:243-8.
3. Shusterman N, Strom BL, Murray TG, Morrison G, West SL, Maislin G. Risk factors and outcome of hospital-acquired acute renal failure. Clinical epidemiologic study. Am J Med 1987;83:65-71.
4. Nash K, Hafeez A, Hou S. Hospital-acquired renal insufficiency. Am J Kidney Dis 2002;39:930-6.
5. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-acquired renal insufficiency: a prospective study. Am J Med 1983;74:243-8.
6. Nikolsky E, Mehran R, Turcot D, et al. Impact of chronic kidney disease on prognosis of patients with diabetes mellitus treated with percutaneous coronary intervention. Am J Cardiol 2004;94:300-5.
7. Marenzi G, Lauri G, Assanelli E, et al. Contrast-induced nephropathy in patients undergoing primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 2004;44:1780-5.
8. Wickenbrock I, Perings C, Maagh P, et al. Contrast medium induced nephropathy in patients undergoing percutaneous coronary intervention for acute coronary syndrome: differences in STEMI and NSTEMI. Clin Res Cardiol 2009;98:765-72.
9. Shema L, Ore L, Geron R, Kristal B. Contrast-induced nephropathy among Israeli hospitalized patients: incidence, risk factors, length of stay and mortality. Isr Med Assoc J 2009;11:460-4.
10. Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality. A cohort analysis. Jama 1996;275:1489-94.
11. Senoo T, Motohiro M, Kamihata H, et al. Contrast-induced nephropathy in patients undergoing emergency percutaneous coronary intervention for acute coronary syndrome. Am J Cardiol 2010;105:624-8.
12. Medalion B, Cohen H, Assali A, et al. The effect of cardiac angiography timing, contrast media dose, and preoperative renal function on acute renal failure after coronary artery bypass grafting. J Thorac Cardiovasc Surg 2010;139:1539-44.
13. Rihal CS, Textor SC, Grill DE, et al. Incidence and prognostic importance of acute renal failure after percutaneous coronary intervention. Circulation 2002;105:2259-64.
14. Amdur RL, Chawla LS, Amodeo S, Kimmel PL, Palant CE. Outcomes following diagnosis of acute renal failure in U.S. veterans: focus on acute tubular necrosis. Kidney Int 2009;76:1089-97.
15. Goldenberg I, Chonchol M, Guetta V. Reversible acute kidney injury following contrast exposure and the risk of long-term mortality. Am J Nephrol 2009;29:136-44.
16. Lipsitch M, Tchetgen Tchetgen E, Cohen T. Negative controls: a tool for detecting confounding and bias in observational studies. Epidemiology 2010;21:383-8.
17. Rao QA, Newhouse JH. Risk of nephropathy after intravenous administration of contrast material: a critical literature analysis. Radiology 2006;239:392-7.
18. McDonald JS, McDonald RJ, Comin J, et al. Frequency of acute kidney injury following intravenous contrast medium administration: a systematic review and meta-analysis. Radiology 2013;267:119-28.
19. Sinert R, Brandler E, Subramanian RA, Miller AC. Does the current definition of contrast-induced acute kidney injury reflect a true clinical entity? Academic emergency medicine : official journal of the Society for Academic Emergency Medicine 2012;19:1261-7.
20. Kayibanda JF, Hiremath S, Knoll GA, et al. Does intravenous contrast-enhanced computed tomography cause acute kidney injury? Protocol of a systematic review of the evidence. Systematic reviews 2014;3:94.
21. Davenport MS, Khalatbari S, Dillman JR, Cohan RH, Caoili EM, Ellis JH. Contrast material-induced nephrotoxicity and intravenous low-osmolality iodinated contrast material. Radiology 2013;267:94-105.
22. McDonald RJ, McDonald JS, Bida JP, et al. Intravenous contrast material-induced nephropathy: causal or coincident phenomenon? Radiology 2013;267:106-18.
23. Luo Z, Gardiner JC, Bradley CJ. Applying propensity score methods in medical research: pitfalls and prospects. Medical care research and review : MCRR 2010;67:528-54.