If ever there was a month that demonstrated how little medicine knows what it's doing, it has to be this one. In the same month, we find out that estrogen doesn't protect against heart disease, angioplasty isn't as safe or effective as bypass surger

The study about estrogen, conducted by the University of California at San Diego's Department of Family and Preventive Medicine, examined 50 year old women and also those who were post-menopausal, and discovered that women with heart disease didn't have lower levels of estrogen than those with healthy hearts. And neither high blood pressure nor raised cholesterol didn't pose any risk either. This is the second such study to come to these conclusions.

Just sit back ponder for a minute what this means. It means that post-menopausal women aren't "missing" estrogen and don't need to have it replaced. It kicks away one of the main platforms of hormone replacement therapy to protect older women from heart attacks. It means that millions of women have been exposed to the cancer risks of HRT for no other reason than to prevent hot flushes.

As for angioplasty, doctors just assumed that a procedure without surgery "had" to be safer than a big full scale operation, so nobody thought it needed to be tested before being universally adopted as the favourite alternative. It was only after decades of use, when all the bad results of angioplasty finally prompted a giant multinational trial, that we discover that, actually, it's killing more people that major heart surgery, which, admittedly, is going some.

As for the tuberculosis vaccine, it's been around for so long that everyone just assumed it eradicated the disease. It's only now that tuberculosis is back again that we bother to find out why, and discover that the BCG jab only works against leprosy (and I'd like that confirmed in a randomized trial before putting my money on it). It means that countless millions of children have been exposed to all the risks of this vaccine for nothing.

Lately, articles have abounded about "evidence based medicine" which simply means looking up what has been proven in medical research before using it on patients (BMJ, April 29, 1995). The Cochrane Collaboration named after epidemiologist Archie Cochrane, who spent most of his life pointing out the weakness of the evidence supporting much of conventional medical practice has been set up to create and maintain a register of all randomized, controlled trials in biomedical research.

But at the moment, this approach which might seem the obvious way forward to you and me is only the subject of gentlemenly debate or thoughtful review in the medical literature. Doctors are treating it like an arcane notion, that might be considered at some point in the distant future.

Whenever I present the argument that medicine ought to be tested before being used on people, doctors get very highhanded, arguing, "And where would we be if we waited for penicillin to meet the test of time?" (Penicillin is medicine's usual trump card.)

I tell you where we'd be. We'd have a better set of statistics. Every other day, medicine kills two jumbo jets' worth of people in the US alone.

But I suppose waiting for good doctrine is a little like waiting for Godot.

!ALynne McTaggart