Pharmageddon (26 page)

There is a tricky balance here. In
chapter 3
, I praised the efforts of Philippe Pinel and Pierre Louis to monitor the outcomes of treatment in nineteenth century France. The response from French physicians at the time was to rail against attempts to standardize medical practice. At the start of the twentieth century in Massachusetts, the new diagnostic tests introduced by Richard Cabot and others led Alfred Worcester to complain that too great a focus on testing would lead to a neglect of the sacred doctor-patient relationship. Measurement and testing are not problems in their own right; in proper proportion they can be exceedingly helpful. The problems arise if they are captured by an interested party, such as the pharmaceutical industry, just as clinical trials have been.

There is no reason why clinical trials should necessarily lead to a rigid standardization. After all, people do not respond to drugs the way automobiles respond to oil or to the tuning of their running parts. As discussed in
chapter 3
, the data from clinical trials of drugs like the antidepressants, where four of five people apparently responding to drug treatment would have responded if put on placebo, support the exercise of clinical discretion in the use of these drugs rather than a rigid standardization.

Furthermore, rather than leading to standardization, clinical trials might have been expected to lead to a democratization of practice. Medicine is far more specialized than it was in the 1960s and as such it might have been expected that we would question our doctors even less now than we did then, but the proliferation of controlled trials that have been published have provided results open to the public and, with the advent of the Internet, more accessible than in the past. From the 1960s on, patients with complaints from breast cancer to depression could face doctors proposing radical mastectomy or electroconvulsive therapy (ECT) and could talk data to them. No longer could a doctor tell a patient, “Come back and ask me that question when you've had medical training and several years of clinical practice.” At least for a brief time, doctors were transformed from authorities into collaborators whose role was to review the evidence with their patients. The management of breast cancer and treatment with ECT gave rise in the 1960s and 1970s to the notion of informed consent,
⁷
as the results from breast cancer operations and ECT studies in the 1960s came into the public domain.

But where public data like this has democratized authority within medicine, the fact that the data on drugs and the full range of their effects is hidden has also led to a diminution of medical authority, replacing it with something closer to a healthcare totalitarianism. We can question our doctors now about drugs and their effects, but neither we nor they can readily find out what is going on. And against this background the set of guidances drawn up by experts that should enhance clinical practice has become a set of guidelines that have constrained practice, and, paradoxically, led to calls for increased standardization of practice as clinical outcomes fail to improve in response to the efforts of the guideline makers.

But it is another set of measurements that have invaded clinical practice and are leading to a further standardization of care that concern us here. These numbers come from the instruments used in trials, whether from blood tests for cholesterol, peak flow rates, or rating scale changes tracking our moods or sexual functioning. These measurements have been central to the running of modern clinical trials and there seems to be little awareness that taking these measurements out of context and importing them into clinical practice might create a host of unforeseen problems. The measurements have in fact taken on a life of their own.

THE SEDUCTIVE POWER OF SCALES

The seductive power of numbers can be seen by looking at the first measurement technology to have an impact on healthcare—the weighing scale. The earliest weighing scales began to appear in the 1870s.
⁸
Almost instantly people began weighing themselves and doctors began weighing patients. The new numbers persuaded doctors and the early insurance companies that a certain plumpness, which had formerly been seen as a sign of health, was actually a risk factor for future ill health. Physicians and insurers combined to extol the virtues of slimness. Within a remarkably brief period of time, European and American ideas of beauty also changed.
⁹
Gone was the appeal of a Rubenesque woman. Voluptuousness became passé, replaced by the slim flappers of the roaring twenties.

By the 1920s, public weighing scales, commonly with plates fastened to their front giving norms for weight by height, were a regular feature in drug stores or other retailers. Then, in the 1960s, small portable weighing scales that could be placed discretely in the bathroom appeared on the market and soon became a household staple.

In parallel with the sales of weighing scales, a new set of diseases began to gain attention and grow in frequency—the eating disorders.
¹⁰
Anorexia nervosa was first described in the 1870s. In the decades that followed, the new disorder seemed to become progressively more common and by the 1920s was occurring with noticeable frequency, though it is difficult to judge prevalence at the time with any certainty. The eating disorders exploded in the 1960s, spewing out variants such as bulimia nervosa, with estimates that up to a third of women might be affected. Curiously, these were Western diseases, not found in Asian or other populations. That eating disorders were apparently a Western phenomenon gave rise to questions about what it might be about the role of women in the West that would lead to problems like this. Was it wrong for women to work? Had they been abused in childhood? Were cultural expectations for women to be thin simply toxic?

There are biological, social, and psychological factors that contribute to eating disorders and that can help address such questions as why this syndrome primarily affects women rather than men. But the role measurement technology plays should also be considered. The weighing scale in this case offers possibilities to marry biological, social, and psychological inputs and expresses their result simply, concretely, and in a way that people can gauge from day to day, noting whether they are approaching or deviating from the goal they've set themselves. Without the weighing scale to organize these factors, it is hard to see how eating disorders could have mushroomed to affect 30 percent of Western populations.

It is tempting to relegate something as inert as the weighing scale to the margins of consideration. But its power as a behavioral reinforcer should not be underestimated. Stepping on the scales and finding the numbers are over a limit one thinks acceptable can be deeply unsettling. For anyone in training, falling outside a target on their stopwatch might be profoundly dispiriting to the point of abandoning the exercise, while shaving even single digits off a previous set of numbers can provide a huge incentive to keep going.

The example of eating disorders provides a dramatic metaphor for the problems of maintaining equilibrium once we position ourselves on a scale. In this case the numbers tantalize with seductive possibilities of control. Part of the problem is that numbers can trump non-numerical judgments, in the way a ringing telephone cuts across a live conversation we may be having. Unless we have great willpower, or wisdom, the only way to resist their lure may be to set up competing sets of numbers.

Similarly, as we have developed capacities to measure environmental radiation or chemicals, our difficulties with these hazards seem to increase rather than diminish. We need to be reminded that the rates of cancer from background radiation pose a lesser risk than being hit by a car crossing the street. Or that cabbages come with forty-seven natural pesticides in them, many of which would not get a license for use from regulators, but these are what give them their taste.

Or consider the lure of comparative measures in attempts to map and forecast economies. It is easy to produce figures from manufacturing industry and other areas of the economy and from this to construct indices of the Gross National Product (GNP), which people think is especially “good” if it appears to be growing from year to year. But presently scales like this one leave out such factors as the number of trees cut down, the amount of pollution generated, and changes in the quality of life of a nation's inhabitants—all areas that may critically affect our future well-being but can too easily be overlooked if attention is only paid to shifts in GNP numbers. Perversely as things are measured at present, an oil spill like that of the
Exxon Valdez
adds to measurable economic activity and increases GNP.

Our capacity to generate measures and scales in the health domain sets up formidable challenges that cannot readily be managed by just telling people to keep calm or be wise. Will we be helped or harmed by the increasing power of scanning technologies to pick up ever more subtle changes in breast tissue or thyroid gland configurations or prostate gland markers, for example? What will happen once we have technologies that can map out patterns of brain activation? We don't know. Will having wise heads at the medical helm be sufficient to stop us obsessing about the meaning of variations in our numbers?

One reason we can suspect wise heads at the helm will not be sufficient is that managing the numbers found in scales is not just a matter of advising people to put them in context. Like the brooms brought to unpredictable life in the Sorcerer's cave when in Disney's
Fantasia
the apprentice waves a magic wand, vivified by market forces the numbers can come to life. Scales set up markets. The numbers from weighing scales helped set up all sorts of markets in diets, health farms, and exercise equipment. Our performances as timed on stopwatches feed markets for sneakers, running gear, and lifestyle coaches. Our growing capacities to map individual variation from the genome to brain scanning will set up further markets for people wanting to minimize future risks for themselves or their children. Having identified a vulnerability in ourselves, we will all be in the market for a remedy.

RABBITS FROM HATS

The power that prompting concern in us about some measure of our apparent health can have in the development of new drug markets comes through dramatically in the cholesterol story. For fifty years it has been known that very high cholesterol levels and especially familial hypercholesterolemia, an uncommon genetic disorder that leads to high cholesterol levels, can be a risk factor for heart attacks. These are the people whose cholesterol levels really do count but these could often be picked up without a blood test by the old style clinical gaze alone—in people with this illness there are cholesterol deposits around the eyes.

In the early 1950s, the Framingham study, which followed 5,209 men and women from Framingham, Massachusetts, in an attempt to pinpoint the risk factors for heart attacks and strokes, identified the key risks as obesity, a history of heart attacks or other cardiovascular events, smoking, and a sedentary lifestyle.
¹¹
Raised cholesterol was also a risk factor, but of much less importance; moreover, it was only a risk factor when one or more of these more serious risks were also present. The most important things, then, for people who have not already had a heart attack and even for those who have had a heart attack, are to reduce weight, get fit, and stop smoking rather than to measure their cholesterol levels. In fact most Western countries saw a 30 percent drop in cardiovascular mortality between the 1970s because of increased attention to smoking, diet, and fitness and the 1990s when the statin group of drugs became widely used to lower cholesterol levels. And aside from the selective use of statins after cardiovascular events, study data suggest that, if anything, there is an increase in mortality in people using statins who are not otherwise at risk of a cardiovascular event.
¹²

While, therefore, there are some people, primarily in hospital care, who have already had a heart attack or stroke, who may be helped by cholesterol screening, widespread and indiscriminate cholesterol testing in society in general with the consequent treatment with statins that slightly elevated cholesterol readings almost inevitably lead to, may in fact lead to as much harm as good.

Since the development of the statins in the 1990s, pharmaceutical companies have put a premium on getting doctors to test for cholesterol and on prompting patients to find out about their cholesterol levels. Today, almost all adults in the United States, and increasingly people elsewhere, know their cholesterol levels. This campaign was facilitated by the fact that most routine blood tests that have been run since the 1980s have included those for cholesterol levels. As guidelines on cholesterol management began to recommend ever lower cholesterol levels, results that would not have been seen as problematic only a few years previously, began to trigger panic in doctors. In this way company marketing has effectively achieved the first case of mass screening for which informed consent was not sought.

While the cholesterol-lowering statins grew to become a $30 billion a year market in the late 1990s, it was also becoming clear that simply lowering cholesterol did not provide a person much benefit. Indeed, the drugs could be risky in their own right, and cholesterol itself, scientists were finding, was not without benefits. This however did not put a break on statin sales—the numbers were “refined.” Popular articles and medical reports began to distinguish between high and low density cholesterol and their ratio to each other, as well as triglycerides and fatty acids (which are further lipids found in blood). Where we all might have had an average overall cholesterol level before, it was becoming increasingly unlikely that any of us would be absolutely “normal” on all these measures and could walk out of a doctor's office without being proffered a drug to match our numbers, even though the attention that cholesterol and other lipids now receive in medical encounters is out of all proportion to their clinical usefulness.
¹³