A CT scan with contrast can injure the kidneys. That much has been known for decades.
What has changed is how often it actually happens — and how many patients are now harmed by the fear of it rather than by the contrast itself.
The evidence moved
Early studies compared patients who received contrast to patients who did not, and found more kidney injury in the contrast group. But sicker patients get more scans. Those studies could not separate the dye from the illness that prompted the scan.
Studies that matched patients properly found the effect far smaller than previously believed, and in patients with normal or near-normal kidney function, difficult to detect at all.
The risk is real. It is concentrated in a specific population, and it was substantially overestimated for everyone else.
Who is actually at risk
eGFR below 30 is where risk becomes clinically meaningful. Below 45 warrants attention, particularly with other factors.
Diabetic nephropathy. Volume depletion. Heart failure. Advanced age. Multiple myeloma. Concurrent nephrotoxic drugs. Large contrast volumes, and repeated exposures close together.
Intra-arterial contrast — a cardiac catheterization — carries more risk than intravenous contrast for a CT scan. The two are frequently discussed as though they were the same thing.
The harm of withholding
This is the part worth stating plainly.
A patient with a suspected pulmonary embolism does not get a CT angiogram because their creatinine is 1.6. They get a lower-quality study, or no study. The embolism is missed.
A patient with a possible aortic dissection. A patient with an occult malignancy. Withholding a necessary contrast study because of an exaggerated risk causes more harm today than contrast nephropathy does.
The question is never whether contrast carries risk. It is whether the information from the scan is worth that risk. Usually it is.
What actually prevents it
Intravenous isotonic saline before and after the study, in genuinely high-risk patients. This is the only intervention with consistent support.
Using the smallest effective contrast volume.
Stopping nephrotoxic drugs — NSAIDs above all — around the study.
Spacing repeated studies when both are needed.
What does not work
N-acetylcysteine. Studied extensively, and it does not prevent contrast nephropathy. It persists in order sets anyway.
Sodium bicarbonate offers no advantage over saline.
Prophylactic dialysis after contrast does not help, and in patients not already on dialysis it introduces the risks of an unnecessary catheter.
Metformin does not cause kidney injury. It is held because a patient who develops kidney injury for any reason and continues metformin risks lactic acidosis. The distinction matters, and patients are routinely told the opposite.
When it does happen
Creatinine rises within 48 to 72 hours and typically peaks around day three to five. Most patients recover within one to two weeks.
The injury is usually mild and self-limited. Dialysis is rarely required. But an episode of contrast nephropathy raises the long-term risk of chronic kidney disease, which is why anyone who has one should have kidney function rechecked afterward.
What we do
Assess the actual risk rather than the reflexive one. Advise the ordering physician on whether prophylaxis is warranted. Give saline when it is. Review the medication list.
And say clearly, when it is true, that the scan should proceed.