Ron Winslow from the WSJ just wrote on 2/24/09 about a new JAMA study finding that "about half the medical recommendations for heart patients have limited scientific backing..."
"Heart disease is among the most studied illnesses in all of medicine, yet just 11% of more than 2,700 recommendations approved by cardiologists for treating heart patients are supported by high-quality scientific testing, according to new research..."
Experts Disagree
But such examples are the exception. For a variety of other conditions, treatment recommendations rely largely on non-randomized studies or expert opinion. For instance, it's unclear at what point patients who suffer excess bleeding during a heart procedure should get a blood transfusion, Dr. Tricoci says. And debate rages among cardiologists over two new tools for assessing a patient's long-term risk of a heart attack. When does a blood test for a marker called C-reactive protein aid in making such a prediction? How about a scan to check for calcium buildup in a patient's arteries? Mounting evidence supports each test, but more data are needed to determine how they might best be used, Dr. Tricoci says.
Harlan Krumholz, a cardiologist at Yale University School of Medicine, says doctors should disclose to patients the strength of the evidence behind the care they recommend. "Treatment decisions are often made very dogmatically even when the level of evidence isn't very strong," he says.
One reason for the lack of stronger evidence is that the large "megatrials" that have dominated cardiovascular research in the past decade were sponsored by drug and device companies. While those studies provide an important source of information, they are typically designed primarily to win approval for a treatment or to widen the market for a therapy already on the market, and not to guide treatment decisions, according to the JAMA study.
Limitations of Studies
Clinical studies also typically exclude patients with complicated illness, which can limit the ability of doctors to apply findings to many of their patients. For instance, many heart patients also suffer from kidney disease. But contrast agents typically given to patients undergoing a diagnostic X-ray called cardiac catheterization can be harmful to kidneys.
Dr. Tricoci says he and his colleagues also observed that guideline writers are picked for their expertise in the field -- but they are also often those who consult regularly with industry. Such possible conflict of interest raises the potential to introduce bias into the guidelines, undermining their credibility."
New PNAS Article: Statins Can Make You Dumb . . .
Along with the above commentary, yet another recently published journal article reports how statin therapy can affect and reduce brain neurotransmitter function. For those susceptible (eg, mitochondrial dysfunction, genetically present mitochrondrial variants, etc), those on excessively high dose statins or perhaps any dose can experience frequently reported statin side effects: 'brain fog', short-term memory loss, erectile dysfunction, muscle aches, tingling/burning in the extremities, peripheral neuropathy, etc.
Cholesterol-reducing Drugs May Lessen Brain Function, Says Researcher
If you dig through what is called "evidence" by the NHS and NICE in the UK you'll find most of it is based on metastudies.
ReplyDeleteNot only that but the studies chosen often specifically exclude the patients to which the advice is later applied, for example dietary studies that exclude diabetics are used to determine a "diabetic diet".
My "special subject" is Type 2 diabetes
http://www.library.nhs.uk/Diabetes/ViewResource.aspx?resID=328114
http://www.crd.york.ac.uk/crdweb/ShowRecord.asp?View=Full&ID=12009105859
http://guidance.nice.org.uk/CG66
When you dig around in back of some of these sites you'll find references to papers which don't actually get included in advice. Same for the ADA, WHO, USDA etc. etc.
Then you see who sponsors pretty much all the health "charities"
http://www.eatright.org/corporatesponsors/
http://www.diabetes.org/donate/sponsor/
hmmm, can't get the list to come up
http://www.diabetes.org.uk/Get_involved/Corporate
/Acknowledgements/
So first the "science" is selected by statisticians on the basis of not offending sponsors, from research paid for by drug companies, and after that it *really* goes downhill.
Little surprise then that when you find something that works your doctor will often (but by no means always) tell you to stop that at once.