A data safety monitoring board reviews efficacy outcomes every 3 months during a large stroke prevention trial enrolling 2,400 patients. The biostatistician notes that repeated looks at accumulating data—currently showing a trend toward reduced ischemic stroke incidence (p=0.08 at interim analysis)—increase the chance of a false positive result unless the significance threshold is adjusted downward. No safety signals have emerged. Which statistical concept is being described?

  1. A)Alpha spendingGABARITO
  2. B)Regression to the mean
  3. C)Effect modification
  4. D)Verification bias
  5. E)Allocation concealment

Explicação

When investigators perform multiple interim analyses, the overall probability of a type I error increases unless alpha is controlled across repeated looks. Alpha spending approaches distribute the allowable false positive rate over the course of the trial. Thi... Ver explicação completa e trilha adaptativa →

Fazer o diagnóstico grátis de USMLE