Book Review

Prescribing by Numbers: Drugs and the Definition of Disease

N Engl J Med 2007; 357:516August 2, 2007

Article

Prescribing by Numbers: Drugs and the Definition of Disease
By Jeremy A. Greene. 318 pp., illustrated. Baltimore, Johns Hopkins University Press, 2007. $49.95. ISBN: 978-0-8018-8477-1

At the dawn of the era of personalized medicine, we are poised to understand how patterns of genes and proteins can influence the risk that a disease will develop in an asymptomatic person. We will determine the risk by calculating risk scores and categorizing patients on the basis of those scores. The idea of quantifying risk by means of numeric formulas and then labeling asymptomatic persons as diseased is controversial, but it is not a new concept. In Prescribing by Numbers, Greene reviews the history of the treatment of hypertension, diabetes, and hypercholesterolemia and addresses the idea of treating asymptomatic persons on the basis of quantitative measures to prevent future health problems.

Greene describes the relationship between advances in treatment, the incentives of manufacturers, and the effect on the public of increased attention to prevention. Science drives the development of technology, which in turn drives the development of science. With the power and benefits of new technology, the opportunity to be more aggressive than ever before in recruiting at-risk patients for treatment or prevention has changed standards of clinical practice. Clinical benefits can be assumed, but quantification is difficult and raises many questions.

The risk–benefit trade-offs of the quantitative approach are complex, and Greene's historical revelations are timely. GlaxoSmithKline's drug Avandia (rosiglitazone maleate) is now in the news because of worries about cardiovascular events related to its use in patients with diabetes. Twenty-seven years before this finding, the use of tolbutamide — the first oral medication for diabetes — was curtailed when a clinical trial showed an increased incidence of cardiovascular events. Controversy between the manufacturer and the Food and Drug Administration surrounding the labeling of this drug continued for almost 15 years. Today, we continue to treat the symptom of high blood glucose in an effort to reduce the chance of future cardiac events, but we are not always certain of this result.

Another revelation in this book concerns the history of commercial influences on clinical development and the conflicts that can result. Given current discussions about the influence of industry in health care, it is perhaps worth acknowledging how far we have come. In 1958, an entire issue of the Journal of the American Medical Association (JAMA) was devoted to articles that detailed the efficacy and safety of chlorothiazide. Readers might not have known that the editor-in-chief of JAMA was also the president of the Pharmaceutical Manufacturers Association (now called the Pharmaceutical Research and Manufacturers of America).

Not discussed in Greene's book are the benefits that patients have derived from the quantitative approach to disease categorization. Complex forces clearly will shape personalized medicine, and the trade-off between risks and benefits may not always be clear. We can look to previous experiences from our evolving clinical guidelines to anticipate challenges to come.

Kevin A. Schulman, M.D.
Duke University, Durham, NC 27715