Sample Size Re-estimation (SSR) Design

Sample size re-estimation (SSR) is a type of adaptive design which allows modifications in the sample size based on the results of interim analyses of a trial. Sample size alteration (adjustment or re-estimation) is possible in blinding or unblinding manner depending on the different criteria such as treatment effect-size, conditional power, etc.

Sample size

Sample size represents the number of participants in a study. When a researcher decides to conduct a research, he/she needs to select the sample out of the target population. Sample population is the population in which the researcher conducts the study.

Statistical power of a study is one of the deciding factor for the sample size of that particular study. Sample size is directly proportional to the power of the study. (Statistical power is the probability that a statistical analysis will be able to identify and reject the false null hypotheses.)

To identify the clinically significant difference (if truly exists) in a clinical trial, it is important to set pre-determined statistical power for the study. To maintain the power of study, appropriate number of subjects are required. The estimation of correct sample size for a study is critical for the success of the study. Any negligence in its estimation may lead to rejection of an efficacious drug and approval of an ineffective drug.

Need of SSR

During development of therapeutic product, sample size is estimated on the basis of limited information such as effect size between two treatment groups and variability of the primary end point. Hence, there can be requirement of few variations in the set methodology due to uncertainties of key factors. For example -  results of early (small scale) phases may be insufficient for providing appropriate estimates for late (large-scale) phases. Therefore, it may be needed to change the sample size in the mid-course of the study.

A classic design starts with pre-specified sample size and modification in sample size is not allowed post commencement of the trial. Thus, this design takes all the samples at the end of the study for analysis if there is no dropout.

SSR is a design which is flexible in nature and permits alteration in sample size amid the trial to maintain the required statistical power. The change in sample size could be based on information obtained from interim analysis for the effect size or for other nuisance parameters (such as standard deviation, coefficient of variation, drop-out rate).

Advantages of SSR

  • Flexible in nature and allows modification in the sample size
  • Avoids unnecessary exposure of patients to the inferior drug
  • Avoids long term trials in case drug is ineffective/low effective

Disadvantages of SSR

  • Re-estimation of sample size based on the observed difference in a relatively small interim sample may lead to introduction of bias and reliability issue
  • May need repetition of recruitment process (depending on the interim analysis)