乌鸦传媒

PHILADELPHIA -- The myth that certain patients need to avoid iodinated contrast media (ICM) for fear of acute kidney injury (AKI) persists despite evidence that there is no additional danger for them, a researcher said here at the annual meeting of the American College of Emergency Physicians.

"[Our study] supports previous research which has essentially not shown an association, and questions the fact that we continue to restrict this modality that really would help diagnose patients in almost a discriminatory way against people that have reduced renal function," said Shawna Bellew, MD, MPH, of the University of South Carolina College of Medicine in Greenville.

The first association between the use of intravenous contrast media and AKI was reported in the 1950s, and this association "really came from case reports and case series," Bellew said. "There were no controlled studies until the 1980s, and they didn't show an association. And these studies really wouldn't be replicated until the end of the 2000s."

Research in this area is made more difficult "because there's this long-held belief in the association between contrast and renal injury, and it creates a selection bias when you're doing retrospective work," said Bellew. "You also really can't randomize giving someone contrast, for somewhat obvious reasons."

Various research methodologies have been used to try to address these issues, including propensity score matching of large datasets "and then more recently, this idea of 'fuzzy regression,'" but nothing has shown a consistent association between the use of contrast and renal injury, she said, adding that "despite this, many facilities, including my own, continue to restrict the use of contrast in patients felt to be at risk for renal injury, particularly patients with baseline low renal function."

But then, in April 2022, "the plant that makes contrast for General Electric shut down secondary to COVID, and this caused many facilities, including my own, to have to vastly decrease the use of contrast to avoid running out of it completely," Bellew said. "And so I wondered, 'here's this huge natural experiment where we're going to change the way we do everything -- what's going to be the effect on the incidence of the development of AKI?'"

To test the hypothesis that there would be no association between the rationing of contrast media and a decreased incidence of AKI, the investigators looked at data on adults who had a CT of the abdomen performed at one of six emergency departments and who had at least two creatinine measurements, at least 24 hours apart. Data abstracted from the electronic medical record included age, sex, race, comorbid conditions (hypertension, diabetes, chronic kidney disease, congestive heart failure), acuity score, ICM administration, and creatinine levels within 7 days of presentation.

The maximum change in creatinine was determined by subtracting the maximal creatinine obtained with the initial creatinine. The primary outcome was the development of AKI, which was defined as an increase in creatinine of 0.3 mg/dL or a percentage increase in creatinine ≥50%. The researchers compared the incidence of AKI between groups and calculated an adjusted odds ratio. They also used subgroup analysis to compare the incidence of AKI by whether contrast was administered, and also stratified it by the initial estimated glomerular filtration rate (eGFR).

Patients were divided into two groups of about 1,100 each -- those who were treated before contrast rationing started and those treated after it had begun -- "and there was no significant difference in the patients' demographic information, comorbidities, or initial baseline kidney function," Bellew said. In the "pre-rationing" group, 87.7% of patients received contrast compared with 42.7% in the rationing group, but "interestingly, there was no change in the development of AKI; it was about 11% in both groups," she said.

The study does have some major limitations, Bellew said. "It's retrospective, the sample size was a lot smaller than I was hoping, and there's still selection bias [because] our facility still restricts [it], so patients that got contrast in this study were probably sicker if their eGFR was less than 30 because someone was willing to take the risk and argue with the radiologist to get it done anyway. If anything, that should bias the data towards showing more of an association, and we still didn't find one."

Bellew said she'd like to do the study again with a bigger sample size, and in the future, she is hoping that institutional policies on restricting contrast will be changed. Once that's done, hopefully the researchers can perform a more rigorous prospective study, she said.

Comment