This tweet came through my time line yesterday:
You really should check out the tweet on twitter to see all of the discussion it generated (I sampled some of the best tweets in the Storify below).
I skimmed over the full text article and it is not very convincing. The study is a meta analysis of 13 non-randomized trials that compared iodinated contrast imaging to non-contrasted imaging. I suspect most of the imaging was CT scans. No intra-arterial contrast was allowed, this eliminated all of the cardiac cath data (though the idea of a non-contrasted cardiac cath is ludicrous).
Before I started reading the article I suspected he would be pulling lot’s of old studies form the early days of iodinated contrast, I mean who today is questioning the existence of contrast nephropathy. Wrong:
Red: contrast causes renal failure
Blue: contrast protects from renal failure
not determining if the contrast itself was safe. Looking at the references it becomes clear why I hadn’t been familiar with this supposed controversy, this stuff is not published in general medicine or nephrology journals. All doctors live inside specialty defined bubbles and, besides a single negative study published in Archives of Internal Medicine, none these articles pierced the bubble.
The thrust of the article is that renal dysfunction following IV contrast administration is no more common than renal dysfunction without contrast administration. If one looks at placebo (or placebo-like) controlled studies there was no signal that contrast increased the risk of acute renal failure:
Contrast: 6.4% (1004 of 15582)
No contrast: 6.5% (675 of 10368)
Contrast: 2.4% (178 of 7359)
No contrast: 6.7% (81 of 1216)
Contrast: 0.3% (24 of 7270)
No contrast: 1.2% (15 of 1214)
Reading this article I could not forget the faces of the patients I had taken care of who had died of contrast nephropathy. My disbelief was practically palpable. Trying to reconcile the data and my experience some of my thoughts were:
- If you had a patient at risk of ARF you would avoid contrast, so maybe we have a lower risk population getting contrast?
- Patients getting contrast may get prophylaxis with IV fluids and N-acetyl cysteine, both of which can lower creatinine. Does that lower the creatinine mask renal injury?
|From Tepel. NEJM 2000. Shows NAC not only prevents CN but lowers creatinine.|
These justifications are just another way of pointing out the problems that come from a lack of randomization and secondarily a lack of blinding. These methodology shortcomings are not solved by a meta-analysis.
UPDATE: My resident just pointed out that two of the authors have received money/grants from GE Healthcare, a manufacturer of CT scanners and iodinated contrast.
Here are some of the responses from Twitter
Note: My first grand rounds after finishing fellowship was on contrast nephropathy. I think it was the first time I used Keynote, way back in 2004. When I look at most of my presentations from that era I usually cringe but I think that old presentation holds up pretty well.: