« January 2009 | Main | March 2009 »
February 28, 2009
Flu vaccine
There has been plenty written for and against the flu vaccine. I have been seeing a fair number of flu cases and more people who think they have it but have a negative test. So I find the study comparing the intranasal live vaccine with the inactive interesting on account of the cases of pneumonia and influenza prevented by the vaccine. As I read it, over a three year period there were about 8 cases per 1000 person-years in the vaccinated group, and between 10 and 19 cases in the unvaccinated group. So the NNT varies from 90 to 500. I am not sure how to average that. But in a bad year for flu/pneumonia, 90 people are vaccinated to prevent one case. Not one death but one case. Those that I have seen this year have not been terribly ill, not much more ill than many who don't have the flu. Think of the cost, the time, the worry.
And how does one know if one has the flu? The Positive predictive value of the test is less than a coin toss. That is, even if one tests positive, one cannot be sure one has the flu. This is a result of the poor specificity (15% false positives) and the low prevalence. The Negative predictive value is much better. So why do the test? You tell me.
Posted by Robert Maddox at 5:09 PM | Comments (0)
Help for Obama reforming health care
JAMA also reports (in the March 4 issue - this is amazing that I received it before it was published, that I have read it, and that I have access to a computer to write this) on the Obama transition team community meetings to discuss health care. (Either I was not invited, didn't know about it, was busy doing important things with the family, didn't think Obama's team cared what I thought, or didn't think it would make a difference.) The White House Office of Health Reform is using volunteer labor to compile "synthesize comments into a manageable document" to deliver to Obama tomorrow. Interestingly, there is "no consensus on its (health care system) biggest problems and possible solutions to fix it."
Apparently, the transition team gave a list of acceptable answers to each of seven questions. I suspect that the answer to the biggest problem question did not allow the answers "hubris" or "idolatrous expectations."
One professor highlighted in the article is quoted to summarize the article, "There's this fundamental metric that's missing from our discourse -- what do individual patients value in a health care setting? We need to shift perspectives; we assume to know what value is, what the patient wants, but we make that assumption without any data."
I might agree with Dr. Corvera, if she were to go on to say that such a metric is noncollectable, and that therefore government, or other third-parties have no right or responsibility to make such decisions. But if she presumes that such a metric can be obtained and systematized for a single or third party payor, she will top the hubris charts.
Posted by Robert Maddox at 4:45 PM | Comments (0)
death by acne treatment or not?
In a interesting abstract from Archives of Dermatology published in JAMA (current issue not yet on-line as of this writing , which is curious in this day and age), the VATTC study from 2005 is reported. This caught my eye for several reasons.
The study was stopped in 2005 due to the excessive number of deaths in the tretinoin-treated group. It was not published in the AD until January 2009. Why was it delayed and what was the point of publishing it now?
Secondly, it is not clear to my reading what they were treating and what they were expecting. It appears that they were treating a certain type of skin cancer which has responded to systemic (as opposed to topical) medicines like this one. There is evidence of increased deaths among smokers treated systemically. Yet they claim that death was not a contemplated end point, when they were evaluating the "relation of topical tretinoin... with all-cause mortality... ."
I am even more curious about the view of science exhibited herein. There was a clear and significant association between use of this cream and death, yet the authors are unwilling to infer a causal association. The endpoint they were studying was a type of cancer. If there had been that association, would they have published? If the trial had shown an association of desired benefit, would they have inferred causality? By what right do we decide when an association is causal?
Disclaimer: This does not constitute advice to anyone using this acne cream.
Posted by Robert Maddox at 4:17 PM | Comments (0)
February 8, 2009
An Amazing Life
My mentor and friend, Dr. Hilton Terrell, died suddenly last week. His obituary is reproduced here. Below are comments I was privileged to make at his funeral.
Dr. Hilton Pack Terrell
Florence, SC
Continue reading "An Amazing Life"
Posted by Robert Maddox at 3:44 PM | Comments (1)