Overcoming Bias

Elusive Placebos

Alas, I have treated the Placebo effect too uncritically in my health econ class.  I'll do better next year.  From Wikipedia:

The original 1955 article of Beecher "The Powerful Placebo" claimed a 35% placebo effect in 15 studies. The original article was in 1997 re-analysed and "no evidence was found of any placebo effect in any of the studies" used by Beecher. ... The claimed "effects" were produced by spontaneous improvement, fluctuation of symptoms, regression to the mean, additional treatment, conditional switching of placebo treatment, scaling bias, irrelevant response variables, answers of politeness, experimental subordination, conditioned answers, neurotic or psychotic misjudgment, psychosomatic phenomena, misquotation, etc. ...

Hróbjartsson and Gøtzsche published a study in 2001 and a follow-up study in 2004 questioning the nature of the placebo effect. ... They performed two meta-analyses involving 156 clinical trials in which an experimental drug or treatment protocol was compared to a placebo group and an untreated group, and ... found that in studies with a binary outcome, meaning patients were classified as improved or not improved, the placebo group had no statistically significant improvement over the no-treatment group. Similarly, there was no significant placebo effect in studies in which objective outcomes (such as blood pressure) were measured by an independent observer. The placebo effect could only be documented in studies in which the outcomes (improvement or failure to improve) were reported by the subjects themselves.

HT to Brandon Reinhart.

Faith in Docs

Today is my health econ final exam.  I also return their last paper, on faith healing.  After an entire semester hearing how we get little health value from a wide margin of medical spending, almost every student (21 undergrads & 9 grads) said that a big argument against legal faith healing is that it can discourage people from going to regular doctors.  Most also said it is hard to evaluate faith healer quality, and to know if they are just in it for the money.   

Sigh.  Regular docs are mostly in it for the money, and are also hard to evaluate.  If we on average get near zero health from our last units of medicine, we are better off replacing those units with anything cheaper, at least if it also gives near zero net health effect and similar non-health benefits.  Faith healing seems to fit this bill. 

Sure, we vary in how much medicine we get, and in how much we would substitute legal faith healing for medicine.  So yes a general trend toward more faith healing would no doubt produce a few people who sometimes get too little medicine.  But that harm should be far outweighed by a reduction in harmful overtreatment.  Alas, apparently even econ students after a semester of my indoctrination can't see this (only two mentioned it) - we all just love docs too much. 

I'm struck by how emotional was the opposition to faith healing and how timid were its supporters.  Most people believe prayer can make you well, but few believe religious specialists can use such powers to similarly help others.  Yet our faith in docs is so strong that when considering medical quantity variation, a few getting too little dominates our attention - we just can't see most getting too much.

Beware Transfusions

New Scientist:

Blood transfusion became a mainstay of medicine during the two world wars, where it was used as a last resort to save soldiers who had suffered massive blood loss. But now, far from being restricted to catastrophic bleeding, transfusions are routinely used as an optional treatment, most commonly for patients in intensive care or undergoing major surgery. ... The rationale behind such blood transfusions seems incontrovertible. Red cells deliver vital oxygen to tissues, and seriously ill patients who are also anaemic fare less well, so a transfusion should help. Those assumptions went untested for the better part of a century.

Things started to change in 1999 with a randomised controlled trial on 838 critical care patients in Canada that used haemoglobin levels to determine when a blood transfusion was given. Normal levels of haemoglobin ... range from 120 to 170 grams per litre. A normal haematocrit - the proportion of red cells in the blood - ranges from 36 to 50 per cent.  Doctors decide whether to give a transfusion based on a number of factors, including haemoglobin levels and haematocrit, and the patient's overall robustness. Many guidelines exist, and practice varies from one hospital or doctor to another, but it is common for patients to receive transfusions when their haemoglobin dips to between 70 and 100 g/l or their haematocrit to 21 to 30 per cent.

But the Canadian study found significantly fewer patients died in hospital, 22 versus 28 per cent, if they received transfusions only when their haemoglobin fell below 70 g/l rather than when it fell below 100 g/l.

Beware Supplements

The Copenhagen team reviewed more than 815 clinical trials into the benefits of vitamins A, E, and C, alongside beta-carotene and selenium - all commonly-used supplements.  They selected 68 whose methods were more likely to produce an accurate picture of vitamin benefits ... [and] eliminated a further 21 trials which had a slightly higher possibility of producing a skewed result, ...  While the risk of death was unchanged among selenium and vitamin C users, a statistically significant increase in risk emerged for the other three supplements.   Beta-carotene produced an approximate 7% increased risk, vitamin E a 4% increase and vitamin A, a 16% increase.

More here. 

Prevention Costs

More prevention will not cut health care costs.  Tuesday's Post:

Most of us naturally assume that preventing a disease is cheaper than waiting for the disease to appear and then treating it. That belief is especially dear to politicians, who often view prevention as an underused weapon in the battle against health-care costs. ...

In 1986, a health economist named Louise B. Russell published "Is Prevention Better Than Cure?," in which she concluded that prevention activities tend to cost more than they save. Since the book's appearance, her observation has been borne out by studies of hundreds of interventions -- everything from offering mammograms to all women and prescribing drugs to people with high cholesterol to requiring passenger-side air bags in cars and shortening the response time of ambulances. ...

For example, Australian researchers tried out a program in which general practitioners watched a video and read a booklet about how to help their patients lower their heart attack risk. The patients were then given a series of videos and a self-help booklet on the same topic.  How cost-effective is this instruction? When it is provided for women at low risk of heart disease, $9.8 million has to be spent for every year of life saved in the prevention of premature heart attack deaths. ...

Some disease-preventing activities ... save money, although they are relatively rare. Childhood vaccinations are the classic examples. ... Providing a single colonoscopy to men 60 to 64 years old also saves money. ...

Similar to the finding that prevention rarely saves money is the calculation that people in good health probably rack up higher lifetime medical costs than their less-healthy brethren.  The reason?  Healthy people tend to live longer.

Added:  I'm mentioned in today's NYT re prediction markets, alas again as the extremist.

Anxiety about what is true

Robin links to an article about apparent manipulation of the medical process by drug manufacturers, dealing particularly with the drug Fosamax which is supposed to improve bone strength. The article raises the possibility that many modern diseases don't exist at all but are the creation of the pharmaceutical industry to give them tools to sell more drugs.

My wife used to take Fosamax, so this was a topic of interest to me. To learn more, I did a search on scholar.google.com for "fosamax hip-fracture". It looks to me like most of the hits are pretty favorable to the drug, but how much does that really prove?

Brownlee on Selling Anxiety

Shannon Brownlee in the Post on drug companies inventing new diseases:

It turns out that much of what we -- and our doctors -- think we know about many health problems has been shaped by drugmakers and their marketers. ... Osteoporosis and osteopenia aren't really diseases. Before the 1990s, doctors decided that you had osteoporosis if you were elderly and you broke a bone. When the pharmaceutical company Merck came up with its anti-bone-loss durg Fosamax, it wanted a broader market than just elderly fracture patients. The solution? The company helped fund a panel of medical experts to create diagnostic criteria for osteoporosis so that a diagnosis could be made before the patient actually broke a bone.

The panel's first step was to define "normal" bone density as that of the average 30-year-old woman. Next, the experts chose as their cutoff for osteoporosis a statistical point that was slightly below the bone density of their normal 30-year-old -- a definition they admitted was "somewhat arbitrary." Finally, they came up with a completely new disease -- osteopenia -- for bone density that fell somewhere between that normal 30-year-old and their arbitrary definition of osteoporosis.

Voila -- 30 percent of post-menopausal women suddenly had a disease that needed to be treated early in order to prevent a problem -- hip fracture -- that wouldn't occur for many years, if ever. According to the new guidelines, millions more women now had osteopenia, which their doctors needed to watch like hawks so that their patients could be treated once they progressed to osteoporosis. Merck then took the added step of helping doctors buy DEXA scanners, X-ray machines needed to scan your bones to get that all-important diagnosis. ...

Fosamax ... can cause necrosis (death) of the jawbone. What's more, there aren't any valid scientific studies to show that treating osteoporosis early will prevent fractures down the road. The drug can also trigger serious heartburn.

Added 6Apr: More from Brownlee:

Showing That You Care

My ambitious theory paper, which attempts to explain diverse health behavior puzzles with just a few assumptions, has finally been published in Medical Hypotheses.  (Print copies were mailed today.)  The abstract: 

Human behavior regarding medicine seems strange; assumptions and models that seem workable in other areas seem less so in medicine. Perhaps, we need to rethink the basics. Toward this end, I have collected many puzzling stylized facts about behavior regarding medicine, and have sought a small number of simple assumptions which might together account for as many puzzles as possible.

The puzzles I consider include a willingness to provide more medical than other assistance to associates, a desire to be seen as so providing, support for nation, firm, or family provided medical care, placebo benefits of medicine, a small average health value of additional medical spending relative to other health influences, more interest in public than private signals of medical quality, medical spending as an individual necessity but national luxury, a strong stress-mediated health status correlation, and support for regulating health behaviors of the low status. These phenomena seem widespread across time and cultures.

Eternal Medicine

Most fans of modern medicine do not realize how similar in appearance was ancient medicine:

While the Greeks left a vast legacy of medical texts in a familiar language, we know of only 12 from the time of the pharaohs - written on papyrus in a vanished language that scholars are still grappling with. From their descriptions of diseases and treatments, the texts have left little doubt that the ancient Egyptians had considerable medical skills, but weighing up their pharmaceutical knowledge has proved trickier: although the papyri include some 2000 prescriptions, doubts surround the identity of many of the ingredients listed. ....

Antidepressant Publication Bias

Medical publication bias was real:

The makers of antidepressants like Prozac and Paxil never published the results of about a third of the drug trials that they conducted to win government approval, misleading doctors and consumers about the drugs' true effectiveness, a new analysis has found. ... The new analysis, reviewing data from 74 trials involving 12 drugs, is the most thorough to date. And it documents a large difference: while 94 percent of the positive studies found their way into print, just 14 percent of those with disappointing or uncertain results did. ... In the study, a team of researchers identified all antidepressant trials submitted to the Food and Drug Administration to win approval from 1987 to 2004.  The studies involved 12,564 adult patients testing drugs like Prozac from Eli Lilly, Zoloft from Pfizer and Effexor from Wyeth.

Fortunately, you needn't worry; the problem has been fixed:

Alan Goldhammer, deputy vice president for regulatory affairs at the Pharmaceutical Research and Manufacturers of America, said the new study neglected to mention that industry and government had already taken steps to make clinical trial information more transparent. "This is all based on data from before 2004, and since then we've put to rest the myth that companies have anything to hide," he said.

Can you guess what he will say if the same results are found in data through 2007, but not published until 2011?