When Evidence Based Medicine Can't Happen

Oct 04 2011 Published by under Evidence Based Medicine

Today I reviewed a review article over at Stream of Thought that considers current treatment of IgA nephropathy (IgAN), one of the most common glomerular diseases in the world. Despite being common, most studies of its treatment are retrospective, anecdotal, or small. When the KDIGO group releases its report on glomerulonephritis later this year, most of what it reports for IgAN will not reach the level of "recommendation" because the evidence supporting it is so weak. Most treatments will be "suggestions" based on expert opinion.

The process that KDIGO uses to evaluate evidence is explained here. Two major groups participate: world leaders in the condition at hand (the working group or WG) and a separate professional evidence review team (ERT, practitioners of statistical black magic). The WG takes the output of the ERT and makes recommendations or suggestions in support of various treatments; public commentary occurs during the process as well.

I will not repeat the treatment algorithm for IgAN here at WhizBANG! (you can click over and read the other post). This article and post demonstrate the uncertainty that physicians deal with daily: how do you treat a patient when there is little or no evidence of what you should do?

For kidney diseases, a number of nonspecific treatments can generally be pursued. If the blood pressure is high, we make it normal, usually by inhibiting angiotensin II. If a study ever shows that this is bad for a kidney disorder, I won't believe it. If the patient has protein in the urine, lowering it is good. If cholesterol and other plasma lipids are increased, fix them. Encourage the patient to be active, maintain normal body weight, and avoid tobacco products.

There is value in expert opinion based on anecdotes and case reports as well. We have to have a starting point for valid clinical trials. Some prospective trials in IgAN may suggest "recommended" treatments in the next 4 to 5 years, in part because they expanded on smaller, less rigorous studies. However, telling a patient to wait 5 years for evidence before beginning treatment is impractical. Published guidelines help us choose a rational approach while we wait for evidence.

5 responses so far

  • Arlenna says:

    Pascale, thank you for this--this is the problem with a lot of rare (and even "non-rare") cancers. Because cancer is so heterogeneous and some rare ones have as few as a couple of hundred cases per year in the US, it's next to impossible to perform these kinds of studies. Or at best, it would take 25-50 years to get enough cases together to get good statistical analytics on the outcomes of any particular treatment or disease subtype.

    This is where I see a strong case for the personalized medicine concept (and not in just the "genetic profile" kind of a way, the "patient factors and individual response monitoring" kind of way like you describe above).

    • Pascale says:

      The pediatric oncologists all got together and started national collaborative study groups to overcome some of these issues. Even with well-established groups, there are still some tumors that are so rare we will never have EBM for them.

      We pediatric nephrologists have not been so organized until recently when some regional/national groups have come together.

  • D. C. Sessions says:

    Dr. Lane, this is the kind of issue that could be very interesting taken to sciencebasedmedicine.org

    Much as I love their focus, you could shake up their rather reactive approach by giving them something prospective to gnaw on. Any chance of you doing a guest appearance for Drs. Gorski and Novella?

  • D. C. Sessions says:

    I can drop Dr. Gorski a suggestion, but I'd probably just be in the way. He's the executive editor and one step off the main page.

    Your choice

Leave a Reply