To the Editor.—
Rose et al1 present a decision analysis for isoniazid chemoprophylaxis that, as a statistician, I find sound. As a patient, however, I choose to decline isoniazid therapy, despite the authors' apparent recommendation to the contrary. Nevertheless, their analysis has played a crucial role in my decision.
Life expectancy: The authors' "best-estimate" analysis shows that a converter of my age (40 years) who chooses isoniazid therapy increases his life expectancy by six days. Had the increase been a matter of years, I would choose isoniazid, but an increase of six days does not persuade me. Instead, the analysis helps because it tells me to look for individual circumstances that matter more than population averages and to consider other measures of risk.
Discount rate: This rate measures my reluctance to trade a higher chance of death this year, from hepatitis, for a lower chance of dying