One of the core features of science (and therefore science-based medicine) is to precisely identify and control for variables, so that we know what, exactly, is exerting an effect. The classic example of this principle at work is the Hawthorne effect. The term refers to a series of studies performed between 1924 and 1932 at the Hawthorne Works. The studies examined whether or not workers would be more productive in different lighting conditions. So they increased the light levels, observed the workers, and found that their productivity increased. Then they lowered the light levels, observed the workers, and found that their productivity increased. No matter what they did, the workers improved their productivity relative to baseline. Eventually it was figured out that observing the workers caused them to work harder, no matter what was done to the lighting.
This “observer effect” – an artifact of the process of observation – is now part of standard study design (at least well-designed studies). In medical studies it is one of the many placebo effects that need to be controlled for, in order to properly isolate the variable of interest.
There are many non-specific effects – effects that result from the act of treating or evaluating patients rather than a physiological response to a specific treatment. In addition to observer effects, for example, there is also the “chearleader” effect from encouraging patients to perform better. There are training effects from retesting. And there are long-recognized non-specific therapeutic effects just from getting compassionate attention from a practitioner. It is a standard part of medical scientific reasoning that before we ascribe a specific effect to a particular intervention, that all non-specific effects are controlled for and eliminated.
One of the recurrent themes of science-based medicine is that any medical intervention that can plausibly cause physiological benefit can also plausibly cause physiological harm. There is no such thing as “it can’t hurt.” Sometimes the risk may be minuscule – but we should never assume that it is zero. Being “natural” or “holistic” or being blessed with some other alleged marketable virtue does not affect the risk vs benefit calculation of an intervention.
Vitamins are an excellent example. There is widespread sentiment that vitamins are harmless, and that supplementing with vitamins is therefore a no risk-possible benefit scenario. It is certainly reasonable to conclude from the evidence that vitamins (at usual supplemental levels) are low risk, compared to many other types of medical interventions. High doses, or megadoses, of vitamins, however, risk toxicity and this risk increases with the dose.
But even at sub-toxic doses vitamins should not be assumed to be risk free. This is especially true when we take a public health perspective – what is the net effect of large scale supplementation on the population? A new meta-analysis looking at the net effects of Vitamin E supplementation on stroke risk reinforces this caution.
Most shots in the dark miss. Scientists learn this early in their career – most of the guesses we make as to how things work will turn out to be wrong. In fact, a proper understanding of science requires thorough knowledge of all the ways in which humans deceive themselves into believing things that are not true. In fact, most shots in well-lit conditions (informed by prior knowledge) miss. Ignoring prior knowledge results in chances that are all but hopeless.
Therefore the title of the 1985 book DPT: A Shot in the Dark by Harris Coulter and Barbara Loe Fisher, is perhaps unintentionally ironic. The book sparked the first modern popular concern about the risk of neurological damage from vaccines, in this case the pertussis vaccine that is part of the DTP vaccine.Fisher, of the National Vaccine Information Center (NVIC) still promotes the book and its content, even though the science has progressed in the last 25 years.
At the time the whole cell pertussis vaccine was part of the diptheria, tetanus, pertussis vaccine (DTwP). This combination has been largely replaced with the DTaP vaccine, which contains an acellular pertussis component. This change was partly due to safety issues, rare cases of neurological disease (seizures and encephalopathy) following DTwP being given. DTaP has a lower incidence of fever, seizures, and other side effects.
Over the last couple of days I have been engaged at NeuroLogica in a discussion with a fellow blogger, Marya Zilberberg who blogs at Healthcare, etc. Since the topic of discussion is science-based medicine I thought it appropriate to reproduce my two posts here, which contain links to her posts.
A Post-Modernist Response to Science-Based Medicine
I receive frequent commentary on my public writing, which is great. The feature that most distinguishes blogs is that they are conversations. So I am glad to see that science-based medicine (a term I coined) is getting targeted for criticism in other blogs. One blogger, Marya Zilberberg at Healthcare, etc., has written a series of posts responding to what she thinks is our position at Science-based medicine. What she has done, however, is make many of the logical fallacies typically committed in defense of unscientific medical modalities and framed them as one giant straw man.
She is partly responding to this article of mine on SBM (What’s the harm) in which I make the point that medicine is a risk vs benefit game. Ethical responsible medical practice involves interventions where there is at least the probability of doing more benefit than harm with proper informed consent, so the patient knows what those chances are. Using scientifically dubious treatments, where there is little or no chance of benefit, especially when they are overhyped, is therefore unethical. And further, the “harm” side of the equation needs to include all forms of harm, not just direct physical harm.
Tonight (Friday Night) we will be moving SBM to a new faster host. This will improve the performance of SBM, which has been sluggish recently, and give us the ability to increase our resources as needed as SBM continues to grow.
Comments posted between Friday night and approximately Sunday morning may be lost in the gap as the location of the new servers propagates through the internet. The site will be up throghout this process, but comments may be lost during this period. We are making the move over the weekend because that is when traffic is lowest. SBM should be fully functional by Monday morning, and in any case I will update this post when it appears that the move is complete.
Thanks for your patience.
The SBM move is now complete. If you are seeing this addendum then you are pointing to the new host. Performance seems much better already, but we will be closely monitoring it to keep performance optimal.
In my group practice, the Yale Medical Group, drug-company sponsored lunches and similar events have been banned. This is part of a trend, at least within academic medicine, to create some distance between physicians and pharmaceutical companies, or at least their marketing divisions. The justifications for this are several, and are all reasonable. One reason is the appearance of being too cozy, which compromises the role of academic physicians as independent experts.
But the primary reason is the belief that “detailing” by pharmaceutical sales representatives has a negative effect on the prescribing habits of physicians. There is reason to believe this may be the case because of cases of bad behavior on the part of pharmaceutical marketing divisions – ghost writing white papers, for example. The concern, backed by evidence, is that pharmaceutical companies introduce spin and bias into the information they provide to physicians, whether though CME, detailing, literature, or sponsored lectures. Even when the information itself is not massaged, it is cherry picked, so in the end physicians are not getting a thorough and unbiased assessment of the facts.
The FDA does heavily regulate the marketing of information about pharmaceuticals, but marketers are very clever about exploiting loopholes and seem to be one step ahead of the regulators.
On the other hand there are those who argue that physicians can handle access to information and they are equipped to take it with a grain of salt and put it into context. Certainly most physicians I speak to believe this about themselves. Further, information provided by pharmaceutical companies may actually improve prescribing habits if it makes physicians aware of new products on the market and new information about the drugs they prescribe. The information itself is FDA approved (or at least should be), even if it is selective and wrapped in spin.
Any promoter of science-based medicine often faces the question – what’s the harm? What is the harm if people try treatment modalities that are not based upon good science, that are anecdotal, or provide only a placebo benefit? There are generally two premises to this question. The first is that most “alternative” placebo interventions are directly harmless. The second is that direct harm is the only type worth considering. Both of these premises are wrong.
The pages of SBM are filled with accounts of direct harm from unscientific treatments: argyria from colloidal silver, death from chelation therapy, infection or other complications from acupuncture, burns from ear candling, stroke from chiropractic neck manipulation – the list goes on. You can read anecdotal accounts of such harm on the website, whatstheharm.net. Of course, as we often point out, harm and risk is only one end of the equation – one must also consider benefit. It is the risk/benefit ratio of an intervention that is important. But generally we are talking about interventions that lack any evidence for benefit, and therefore any risk of harm is arguably unacceptable.
It is always flu season somewhere in the world. As Australia’s flu season comes to a close, we are getting ready for ours in the Northern Hemisphere. This is a good time to start thinking about getting the flu vaccine, and as always there is a lot of flu vaccine news to sort through.
Mark Crislip has already reviewed the evidence for the efficacy of the seasonal flu vaccine. Like most questions in medicine, the evidence is deceptively complex, and Mark does an excellent job of sorting through it, so I won’t repeat it here.
This year the H1N1 pandemic flu virus will be incorporated into the seasonal flu vaccine, so there will not be two separate vaccines as there was last year. H1N1 remains the dominant strain of seasonal flu, and as predicted the pandemic has simply been incorporated into the seasonal flu pattern.
Here are some updates on flu vaccine news – first the good news:
Over the last decade there has been a needed discussion about the relationship between politics and science. This has mostly been spawned by the perceived “Republican War on Science,” at the center of which is the global warming debate. In reality, both ends of the political spectrum (as evidenced, for example, by the Huffington Post) tend to trump science with ideology. That is the nature of politics. But at least the issue has been raised.
Briefly, defenders of science have pointed out that science should inform politics, not the other way around. Ideologues should not be allowed to put their thumb on the scale of science in order to get the result their ideology demands. Further, the optimal policy emerges from an honest assessment of the relevant science. Values still come into play for many issues, so science alone is not enough, but the science has to be right.
Within medicine this issue often involves the regulation of the standard of care and public health policy. An example of the former is the law passed last year is Connecticut that essentially exempts professionals who treat “chronic Lyme disease” from the standard of care – the department of health cannot act against their license for treating this controversial condition with unproven therapies.Rather than allowing experts to determine the standard of care, which is an ever moving target, this law locks into place a very controversial, and in my opinion dubious, practice.
Governments and environmental advocates are promoting compact fluorescent lightbulbs (CFLs) as a way of reducing electricity use, saving money, and reducing our carbon footprint. CFLs are not a perfect technology – when turned on they take a moment to fully brighten and they contain a small amount of mercury which requires special procedures for disposal. CFLs are likely also to be a transitional technology, as more energy efficient light sources (such as LEDs) are already coming onto the market. But CFLs are a safe and energy efficient alternative to incandescent bulbs.
It seems, however, with any new technology comes a wave of internet fearmongering, and CFLs are now a prime target. YouTube videos are circulating claiming that CFLs cause headaches, mercury toxicity, a host of symptoms from electromagnetic sensitivity, and something called “type 3 diabetes.” Let’s take a look at the claims and the science.
Mercury in CFLs
There is a small amount of mercury in each CFL, necessary for the function of the bulb, about 4mg on average, with some newer bulbs having as little as 1.4mg. There is no exposure to mercury from using CFLs, as long as they are not broken. Even if a bulb is broken the exposure to mercury is negligible, far less than eating a tuna fish sandwich. But still, there are recommended procedures for cleaning up and disposing of a broken bulb to further minimize exposure, such as not using a vacuum, and ventilating the area. These procedures represent the cautionary principle in action, but make it easy to fearmonger about the risks of the mercury in the bulb.