Ethics / Lecture Discussion 303
Hugh Blumenfeld, MD PhD
Medicine is increasingly driven by guidelines and evidence from studies designed to show extremely small changes in outcomes. When applied to populations, these changes can positively impact thousands of lives. However, most healthcare providers are unfamiliar with the small magnitude of the benefits that many common interventions offer to individual patients, or the small percentage of patients who see them. Too often, we overestimate both their efficacy and the risks of forgoing them. Based on these misapprehensions, we offer interventions that are more likely to harm than help as though they are the only reasonable choice, and pass judgments on those who “refuse” or are “non-compliant,” eroding patient trust and fueling provider frustration. Statistics have always been morally suspect, and we must be especially careful applying them to patients. Wielding “evidence,” we categorize risks as diseases even when they have no symptoms and our medications don’t change outcomes for the vast majority of those who take them; we offer screenings to detect diseases earlier and end up treating more people without achieving longer lives; and in behavioral health, we define illnesses by lists of symptoms without corresponding pathophysiology and treat them with drugs that often perform marginally better than placebo. To proceed on sound ethical foundations we must distinguish between preventive, curative and palliative interventions, applying appropriate statistical concepts to each; we must know enough math to understand and communicate true risks and benefits; and we must help patients make healthcare decisions from a combination of enlightened self-interest and altruism.