Balancing Randomized Trials With Anecdote

Anecdotes may be misleading. They are uncontrolled. Findings may be due to chance, not cause and effect. The anecdote may incorrectly dissuade us from proven therapies. Dispassionate and objective, the randomized, controlled trial has rightfully gained priority as the highest level of evidence when selecting therapy for our patients. Its methodology minimizes the effects of bias and is responsible for medicine's progress from unproven treatments to the current era of evidence-based medicine. However, when it comes to drug toxicity, anecdotal case reports have repeatedly demonstrated that medicines previously found safe in randomized, controlled efficacy trials have important adverse effects (1). Those of us who care for patients who cannot tolerate lipid-lowering therapy are often confronted by anecdotes that contradict the findings of well-designed trials. Our statin myopathy clinic, which now exceeds 600 patients (45 of whom have experienced statin-induced rhabdomyolysis), provides such an experience.

Studies designed to assess the efficacy of statins are not sensitive to adverse effects and have taught us little about statin toxicity. We still do not know whether statins directly induce muscle atrophy (2) or whether lowering lipid levels by any means in patients with vulnerable muscle causes myotoxicity. The older, sicker patients receiving multiple medications, which we commonly see in practice, are systematically excluded from efficacy trials. Moreover, these trials have too few patients and insufficiently long follow-up to reveal infrequent toxicities. Trials with run-in phases that lead to the exclusion of more …