Book Review

Prescribing by Numbers: Drugs and the Definitions of Disease
By Jeremy A. Greene.
Johns Hopkins Press, Baltimore 2007.
Reviewed by Howard Spiro (howard.spiro@yale.edu).

Bookmark and Share

During the 60 years that I have been a physician, many symptoms have been turned into diseases, in what philosophers call reification. A good example comes in common “heartburn,” which used to be a badge of conscientiousness, but now has been promoted to GERD or Gastro-Esophageal Reflux Disease. Professional hubris and “big pharma” enthusiasm has turned a minor symptom relieved by a little baking soda into a calamity which demands lifelong vigilance against recurrence.

This saga is wonderfully explicated in Prescribing by Numbers, a book which I strongly urge upon all clinically inclined readers. It is a gripping story, enlarged from his Ph.D. dissertation by physician/historian Jeremy Greene, a resident in medicine at Boston’s Brigham and Women’s Hospital. In it he describes how increasing reliance on numbers, like blood pressure or cholesterol level, has turned people without complaints into patients with “pre-disease” that physicians can treat in a goodhearted attempt to prevent its progression into full-blown disease.

Having lost faith in the “annual physical” to find something in the patient’s body that can  be prevented, many clinicians have turned to the patient’s blood, to screen for such harbingers of disease as fasting blood sugar or high cholesterol level.  And once they spot an “abnormality,” they have learned to treat people without symptoms to correct these markers.  That has lead to many people taking a drug  “every day for an indefinite period, if not for the rest of their lives” without assurance that  “they will, in fact, ever receive any benefit from their pharmaceutical regimen,” as Greene puts it.

His conclusions are comforting to old clinicians like me, but may trouble younger ones who have been nurtured on number and protocol. At the very outset, Greene makes it plain that he is not demonizing drug companies, but merely recording how “risk-benefit,” properly a statistical calculation, has led to universal “prevention.”  Nor does he gainsay the remarkable decrease in strokes and heart attacks in the United States, but he does recount how forays of actuaries/epidemiologists/statisticians have supported marketing of medications. In simple terms, statistical probability has trumped personal considerations.

A good example comes in a woman I know, the scioness of at least three generations of women who have lived to almost 100 years, who has been plied with antihypertensive/anti-cholesterol agents to help her live as long as her forebears.  Greene tells how this has come about. In his first example, he relates how studies of chlorothiazide, Diuril, the first of the oral diuretics, changed hypertension from a finding  treated only when it caused trouble to a disease which must always be contained at ever lower levels. Accordingly, after the development of diuretics and other drugs, symptoms became less important than signs like high blood pressure in the prevention/treatment of potential disease, even where the evidence was not all in.

Nowadays, “high blood-pressure” has become a disease in itself, where earlier physicians saw it as a spectrum:  hypertensive patients would even be given sedatives before any treatment, to ascertain  what their blood-pressure would run when they were at ease. Other organs, especially the fundi and kidneys, would be scrutinized to find any damage to sensitive organs for hypertension was then considered only one of many factors, stress among them and family history, all of which combined to lead to heart disease and stroke. (As I was writing this, I was delighted to read in the September 12, 2007 issue of the The Lancet a thoughtful review reinforcing the contribution of “stress” to cardiovascular disease.)

Greene repeats the important observation that actuaries look for useful markers in defined populations to guide insurance companies as to their financial risk.  But clinicians treat individual patients and should be aware how little a single measurement means to the fate/prognosis of an individual patient. From guidelines for treating hypertension, it was not long before “the silent epidemic” of high blood pressure was promulgated, and “the concept of prediabetes invested borderline test results with a sense of pathophysiological urgency.” Later, as Greene establishes, the definition of “pre-diabetes” made multitudinous innocents measure their blood and urine sugars, and take medication to stave off the real thing, diabetes.  

Cholesterol provided another milestone, so that in New Haven everybody seems to know his or her cholesterol level, and some take agents like Lipitor, hoping to eat what we like without guilt. Nor do even the aged escape “prevention,” trudging from echocardiogram to colonoscopy before hurrying off to the gym.

Those of you reading this review may wonder at the enthusiasm of an aged gastroenterologist commenting on a field outside his own, but I recommend this book as a valuable contribution to a skepticism that has been swamped by therapeutic pride. I grant you that the health of the American people has enormously improved over the past 60 years, and I have certainly benefited from the advances. Nevertheless, people are different, and the rush to “evidence-based medicine”  sometimes ignores the disagreements of statisticians, now our sooth-sayers, about the precise quantification of some of their statistical approaches, not so different from the discussions of contentious clinicians in a less numerical age. (Again, in the JAMA of September 12, this problem is discussed.)

Nevertheless, Jeremy Greene warns us in his superb book that things are not always as they are claimed. If you do not have the time to read his whole book, look at his magisterial last chapter on "The Therapeutic Transition."

Published: October 24, 2007