Date: October 17th , 2018

Reference #1: Aycock, Westafer et al. Acute Kidney Injury After Computed Tomography: A Meta-analysis. Ann Emerg Med 2018 (CRD42017056195)

Reference #2: Weisbord SD, Gallagher M, Jneid H, et al; PRESERVE Trial Group. Outcomes after Angiography with Sodium Bicarbonate and Acetylcysteine. NEJM 2018 (ClinicalTrials.gov NCT01467466.)

Guest Skeptic: Dr. Lauren Westafer is a board certified emergency physician at Baystate Medical Center and instructor in the Department of Emergency Medicine at the University of Massachusetts Medical School. She is author of the blog, The Short Coat, and cofounder of the emergency medicine podcast, FOAMcast. Lauren is currently funded by an NHLBI K12 grant (1K12HL138049-01) studying the implementation of evidence-based diagnosis of pulmonary embolism in the emergency department.

Case: A 64-year-old woman with type-2 diabetes. She presents to the emergency department with chest pain and some shortness of breath. The acute coronary syndrome work-up is negative but she is Well’s high and needs a CTPA to rule-out a pulmonary embolism. Her GFR is 50 and you are wondering if the contrast needed for the CT will cause an acute kidney injury (AKI) and if so, can you do anything to mitigate causing an AKI?

Background: There has been a huge increase in the number of CT scans performed with more than 75 million CT scans performed in the US in 2013. Some scans require intravenous contrast (CTPA and CTCA) while in other cases it may improve image quality.

There has been a lot of ink spilled over contrast-induced nephropathy (CIN). It came out of case reports and non-controlled studies. Historically the CTs were done with high osmolar contrast material and these non-controlled studies showed a rise in AKI. However, we now use iso-osmolar or low osmolar contrast and we are not seeing kidneys die as a result.

Multiple observational studies have been published demonstrating that AKI in the modern era does not exist. Part of the difficulty with this topic is the inconsistent definition of contrast-induced nephropathy. A common definition is an increase in creatinine level by 25% or an absolute increase of 0.3 to 0.5 mg/dL within 3 days.

These are all disease-oriented outcomes (change in laboratory values) not patient-oriented outcomes like death or need for dialysis.

Clinical Question #1: Is CT contrast associated with acute kidney injury?

Reference: Aycock, Westafer et al. Acute Kidney Injury After Computed Tomography: A Meta-analysis. Ann Emerg Med 2018 (CRD42017056195)

  • Population: Adult humans
    • Exclusions: Pediatrics, non-human studies, studies of contrast enhanced procedures (ex: coronary angiography), interventional studies, case reports, review articles, clinical guidelines, other meta-analyses
  • Intervention: Contrast enhanced CT scans
  • Comparison: Noncontrast CT scan
  • Outcome:
    • Primary Outcome: Incidence of acute kidney injury
    • Secondary Outcomes: Mortality or need for renal replacement therapy

Authors’ Conclusions:“We found no significant differences in our principal study outcomes between patients receiving contrast-enhanced CT versus those receiving noncontrast CT. Given similar frequencies of acute kidney injury in patients receiving noncontrast CT, other patient- and illness-level factors, rather than the use of contrast material, likely contribute to the development of acute kidney injury.”

Quality Checklist for Systematic Review Prognostic Studies:

  1. The prognostic question is clinically relevant for ED patients? Yes
  2. The individual study patients were sufficiently homogeneous with respect to prognostic risk for the outcome? No
  3. The individual study assessment for the outcome used objective, reproducible, and unbiased criteria? Yes
  4. The individual study period of follow-up was sufficiently long and complete? Unsure
  5. The search for studies was detailed and exhaustive? Yes
  6. The methodological quality of primary studies were assessed for common forms of prognostic research bias? Yes
  7. The assessment of studies were reproducible? Yes

Key Results: A total of 28 studies were included with over 100,000 patients. All of the studies were observational with the majority being retrospective chart reviews. Of the 28 studies, 26 evaluated and defined AKI, 13 measured the need for renal replacement therapy and 9 quantified all-cause mortality.

Contrast-enhanced CT was not significantly associated with acute kidney injury

  • Primary Outcome: Incidence of acute kidney injury
    • OR 0.94; 95% CI 0.83 to 1.07
  • Secondary Outcomes:
    • Mortality OR 1.0; 95% CI 0.73 to 1.36
    • Need for renal replacement therapy OR 0.83; 95% CI 0.59 to 1.16


Listen to the SGEM podcast on iTunes to hear Lauren’s full response to my five nerdy questions.

1) Quality of Studies: This SRMA only included observational studies, some retrospective and some prospective. I’m not saying garbage in, garbage out but it does limit the strength of conclusions that can be drawn.

Absolutely. We would have loved to have seen some RCTs but it’s kinda hard to randomize someone who needs a scan for aortic dissection to a non-contrast scan – I think that might raise some flags with the IRB. On the other hand, I think it’s important to take a gander at the evidence that was used to say PC-AKI was a thing to begin with – and that was case series and non-controlled studies…i.e. really subpar “evidence”.

2) Publication Bias: We know that publication bias can exist with positive studies being more likely to be published. Did you check for publication bias and if so, what did you find?

We did a funnel plot of publications, with an equal distribution of studies demonstrated visually. The Harbord-Egger test of bias was calculated to be –0.18 (P=0.70), indicating a low likelihood of publication bias.

3) Selection Bias: Because none of the studies were randomized, that can introduce selection bias. How could selection bias have been introduced into this SRMA and what impact do you think it would have on the results?

The type of CT scan being ordered would have been influenced by the baseline renal function. Also, could have resulted from the requirement for follow-up creatinine-level measurement, including a sicker cohort.

4) Measurement Bias: There were differing definitions of AKI in the included studies and the timing of renal function measurements. Less than one in five studies reported renal function greater than 72 hours. How do you think that could impact the results?

We thought the differing definitions may affect the outcome because using the 25% rise in creatinine (Cr) definition, someone could go from a Cr of 0.7mg/dL (61.9 μmol/L) to 0.875 (84.4 μmol/L) and would meet the definition although these are both within normal limits – it’s a super conservative definition. So we did a subgroup analysis – and no major differences except using 25% increase in creatine alone as a definition was associated with less AKI in the contrast arm…probably because it was just more sensitive in both arms. With the timing of follow-up this could overreport meaningful AKI if the Cr just goes back to normal. Again, we did a subgroup analysis and didn’t find a difference in these groups either.

5) Heterogeneity: The heterogeneity for the primary outcome of AKI as measured by the I2 metric was fairly high (65.1%) indicating moderate heterogeneity.  Some would suggest these studies should not be combined due to the differences between the studies.

We saw the heterogeneity and it made sense given that a bunch of the studies weren’t matched and were relatively small. Because of this we did, yes another subgroup analysis because we hypothesized that the higher quality or matched studies would be more homogenous and woila….when we examined those the I2 was 0% for matched studies and there was still no difference in odds of AKI. Also, the heterogeneity was low for the secondary outcomes…the patient oriented ones of mortality and need for dialysis.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree with the authors’ conclusions.

SGEM Bottom Line #1: CT contrast is not associated with acute kidney injury?

Clinical Question #2: Do intravenous sodium bicarbonate or sodium chloride with oral acetylcysteine or placebo prevent acute kidney injury and major adverse outcomes in high-risk patients undergoing angiography?

Reference: Weisbord SD, Gallagher M, Jneid H, et al; PRESERVE Trial Group. Outcomes after Angiography with Sodium Bicarbonate and Acetylcysteine. NEJM 2018 (ClinicalTrials.gov NCT01467466.)

  • Population: Patients with compromised renal function (GFR 15-45 or 45-60 if diabetic)
    • Exclusions: Patients who were undergoing emergency angiography and those with unstable baseline levels of blood creatinine. See list in Supplementary Appendix.
  • Intervention: IV Sodium Bicarbonate or IV normal saline
  • Comparison: Oral acetylcysteine (NAC) or Placebo
  • Outcomes:
    • Primary Outcome: Composite of death, need for dialysis or >50% increase in creatinine 90-104 days post angiography (persistent impaired renal function)
    • Secondary Outcomes: Contrast-associated AKI (≥ 25% or 44 μmol/L) from baseline at 3–5 days after angiography; confirmed persistent kidney impairment; death; dialysis; hospitalization with acute coronary syndrome, heart failure or stroke by 90 days; or admission to hospital within 90 days.

Authors’ Conclusions: “Among patients at high risk for renal complications who were undergoing angiography, there was no benefit of intravenous sodium bicarbonate over intravenous sodium chloride or of oral acetylcysteine over placebo for the prevention of death, need for dialysis, or persistent decline in kidney function at 90 days or for the prevention of contrast-associated acute kidney injury.”

Quality Checklist for Randomized Clinical Trials:

  1. The study population included or focused on those in the emergency department. Yes
  2. The patients were adequately randomized. Yes
  3. The randomization process was concealed. Yes
  4. The patients were analyzed in the groups to which they were randomized. Yes
  5. The study patients were recruited consecutively (i.e. no selection bias). Yes
  6. The patients in both groups were similar with respect to prognostic factors. Yes
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. Yes
  8. All groups were treated equally except for the intervention. Yes
  9. Follow-up was complete (i.e. at least 80% for both groups). Unsure
  10. All patient-important outcomes were considered. Yes
  11. The treatment effect was large enough and precise enough to be clinically significant. Yes

Results: There were 4,993 patients included in this trial randomized into 2-by-2 factorial design (2,511 sodium bicarb [NaHCO3], 2,482 sodium chloride [NaCl], 2,495 NAC and 2,498 placebo). The average age was 70 years, the vast majority (94%) were men, the median GFR was 50, 80% had diabetes and 90% of the procedures were coronary.

No difference in primary composite outcome or AKI between sodium bicarbonate and normal saline or between NAC and placebo.

  • Primary Outcome: Composite of death, the need for dialysis, or a persistent increase of at least 50% from baseline in the serum creatinine level at 90 days.
    • No difference – around 4.5% for all four groups

  • Secondary Outcomes: No difference in any of the secondary outcomes
    • Contrast-associated acute kidney injury (no difference)
    • Death by 90 days (no difference)
    • Need for dialysis by 90 days (no difference)
    • Persistent kidney impairment by 90 days (no difference)
    • Hospitalization with ACS, heart failure, or stroke by 90 days (no difference)
    • All-cause hospitalization by 90 days (no difference)


1) Selection Bias/External Validity: The population in this study was military veterans with stage 3 or 4 chronic kidney disease of whom 94% were male, 80% had diabetes and most (90%) were getting a contrast enhanced CT scan of their coronaries. This is a fairly select group and may limit the external validity to women, those without diabetes and those getting other forms of contrast enhances CTs.

2) Differences in Intervention: The timing of initiation, duration, and rate of fluid administration varied between different sites.

3) Composite Outcome: Having a composite outcome always makes the target bigger. They did not find a difference between the four groups. A concern always is that not all outcomes included in the composite are considered equal in importance. Persistent impairment in kidney function is not clinically equal to death or even the need for hemodialysis. However, they did report each component of the composite outcome and there were no differences between any of those outcomes either.

4) Stopped Early: The trial was stopped early at the final interim analysis. They included only 5,177 (67.4%)  of the patients minus those (4.1 to 9.2%) with missing creatinine levels. This is probably the biggest threat to the validity of this trial. The SGEM has talked about the problems of stopping trials early on previous episodes (SGEM#133SGEM#137, and SGEM#183).

5) Confounders: The primary endpoint was assessed at 90 days; therefore, the effect of the intervention may be confounded by other treatments in between the CT scan and the composite endpoint at three months.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree with the authors’ conclusions.

SGEM Bottom Line #2: The risk of AKI from CT contrast is not as great as it was thought to be, and it might not even exist. The risk of missed or delayed diagnosis likely outweighs any from the exposure in a patient who requires a contrast CT study.

Dr. Lauren Westafer

Case Resolution: The patient has the CTPA study done without receiving bicarb or NAC and no pulmonary embolism was identified. There was a small suspected pneumonia but requires clinical correlation.

Clinical Application: If a contrast enhanced CT study is needed for patient management you should get the scan even in high-risk patients.

What Do I Tell My Patient? There is no sign of heart attack or blood clot with the CT scan. However, we did pick up a possible early pneumonia on the CT scan. I have written a prescription for antibiotics. See your PCP next week and come back to the ED if you develop a rash, your shortness of breath gets worse or you are worried.

Keener Kontest: Last weeks’ winner was Paul Ehlers a PGY1 from UCSF-ZSFG EM Residency. Paul knew ketamine is an NMDA receptor antagonist.

Listen to the SGEM podcast on iTunes to hear this the new keener question. If you know the answer send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Other FOAMed Resources:

  • The Bottom Line: Outcomes after Angiography with Sodium Bicarbonate and Acetylcysteine
  • REBEL EM– Is CIN really not a thing?
  • EM Cases– Post Contrast Acute Kidney Injury
  • First10EMDoes contrast cause kidney injury? The evidence
  • PulmCrit– Do CT scans cause contrast nephropathy?
  • EM Lit of Note– Punching Holes in CIN
  • emDocs– Contrast-Induced Nephropathy: Confounding Causation

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.