MedTech Terms
    The authoritative reference
    All terms
    Clinical & TrialsClinical Evidence

    Sample Size and Statistical Power

    Pre-trial calculation of the number of subjects needed to detect a specified treatment effect with acceptable Type I and Type II error rates.

    Reviewed by Christian Espinosa, Founder, Blue Goat CyberLast reviewed May 9, 2026

    Definition

    Sample size and power calculations determine how many subjects a clinical trial must enroll to detect a pre-specified effect size with acceptable false-positive (Type I error, alpha, typically 0.05 two-sided) and false-negative (Type II error, beta, typically 0.20 — i.e., 80% power) rates. Inputs include the expected effect size, the variance of the outcome, the test statistic, allocation ratio, and any planned interim analyses. Under-powered trials are a leading cause of inconclusive pivotal studies.
    What the regulation says
    FDA Guidance: 'Statistical Guidance on Reporting Results from Studies Evaluating Diagnostic Tests' and ICH E9 'Statistical Principles for Clinical Trials' both require pre-specified sample-size justification in the protocol.

    What this means in practice

    MedTech pivotal trials commonly target 80-90% power for the primary endpoint. Adaptive designs and Bayesian designs allow for sample-size re-estimation under pre-specified rules.
    Common pitfalls
    • •Powering only the primary endpoint and finding the trial cannot answer key secondary endpoints used for label or HEOR.
    • •Optimistic effect-size assumptions that produce a too-small sample size.

    Primary references

    3 sources
    Link health: 2 verified 1 bot-blocked· last checked 2026-05-09
    ICH·1FDA·1NIH·1
    1. 1
      ICH E9 Statistical Principles
      Verified
      ICHich.org
    2. 2
      FDA Guidance — Adaptive Designs for Clinical Trials (2019)
      Bot-blocked
      FDAfda.gov
    3. 3
      ClinicalTrials.gov
      Verified
      NIHclinicaltrials.gov

    Inline markers like [1] jump to the matching reference above.