Biostatistics in Clinical Trials: From SAP to Regulatory Submission
Biostatistics is the scientific backbone of every clinical trial. The statistical analysis plan (SAP), sample size justification, and final analysis outputs are scrutinised by regulatory agencies and determine whether a trial's results are accepted as evidence of efficacy and safety.
The Role of Biostatistics in Clinical Trials
Biostatisticians are involved at every stage of a clinical trial — from protocol design and sample size calculation, through interim analysis and data monitoring, to final analysis and regulatory submission. A well-designed statistical analysis plan (SAP) pre-specifies all analyses to prevent data dredging and ensures that the primary endpoint analysis is pre-defined, transparent, and reproducible.
Statistical Analysis Plan (SAP)
The SAP is a comprehensive document that pre-specifies all statistical analyses to be performed on the trial data. It includes the primary and secondary endpoint analyses, handling of missing data, subgroup analyses, sensitivity analyses, and the statistical methods to be used. The SAP must be finalised and signed off before database lock to demonstrate that the analysis was pre-specified. Regulatory agencies scrutinise the SAP closely; any deviation from the pre-specified analysis must be clearly documented and justified.
Sample Size Calculation
Sample size calculation determines how many subjects are needed to detect a clinically meaningful treatment effect with adequate statistical power. Key inputs include the expected treatment effect size, the variability of the primary endpoint, the desired power (typically 80% or 90%), the significance level (typically 0.05), and the anticipated dropout rate. Sample size justification is a required element of the clinical trial protocol and is reviewed by regulatory agencies and ethics committees.
Interim Analysis and Data Monitoring
Interim analyses allow sponsors to assess accumulating data during the trial for safety, futility, or efficacy. They must be pre-specified in the SAP and protocol, and conducted by an independent Data Safety Monitoring Board (DSMB) or Data Monitoring Committee (DMC) to protect the integrity of the blind. Alpha spending functions (such as O'Brien-Fleming or Pocock boundaries) are used to control the overall Type I error rate when multiple looks at the data are planned.
Adaptive Trial Designs
Adaptive designs allow pre-specified modifications to the trial based on accumulating data — such as sample size re-estimation, dose selection, or population enrichment. They can improve trial efficiency but require careful statistical planning to control error rates and maintain trial integrity. The FDA and EMA have published guidance on adaptive designs; sponsors should engage with regulatory agencies early when planning an adaptive trial.
CDISC ADaM and Regulatory Submission
Analysis datasets for regulatory submission must be provided in CDISC ADaM format. ADaM datasets are derived from SDTM and are designed to support the analyses specified in the SAP. Key ADaM datasets include ADSL (subject-level), ADAE (adverse events), ADTTE (time-to-event), and ADLB (laboratory data). Each ADaM dataset must be accompanied by a define.xml and a reviewer's guide that documents the derivation of every variable.
Key Checklist
Frequently Asked Questions
What is a statistical analysis plan (SAP) in a clinical trial?
A statistical analysis plan (SAP) is a comprehensive document that pre-specifies all statistical analyses to be performed on clinical trial data. It includes primary and secondary endpoint analyses, handling of missing data, subgroup analyses, and the statistical methods to be used. The SAP must be finalised before database lock to demonstrate that the analysis was pre-specified and not influenced by knowledge of the results.
What is sample size calculation in clinical trials?
Sample size calculation determines how many subjects are needed to detect a clinically meaningful treatment effect with adequate statistical power. It is based on the expected effect size, endpoint variability, desired power (typically 80–90%), significance level (typically 0.05), and anticipated dropout rate. Sample size justification is required in the clinical trial protocol.
What is an interim analysis in a clinical trial?
An interim analysis is a pre-specified analysis of accumulating trial data conducted before the final analysis. It may be used to assess safety, futility, or efficacy. Interim analyses must be pre-specified in the SAP and protocol, and conducted by an independent Data Monitoring Committee to protect the integrity of the blind and control the overall Type I error rate.
What is an adaptive clinical trial design?
An adaptive design allows pre-specified modifications to the trial based on accumulating data — such as sample size re-estimation, dose selection, or population enrichment. Adaptive designs can improve trial efficiency but require careful statistical planning to control error rates. The FDA and EMA have published guidance on adaptive designs.
What biostatistics services does a CRO provide?
A specialist CRO provides biostatistics services including SAP development, sample size calculation, randomisation and blinding, interim analysis support, final analysis, CDISC ADaM dataset production, and regulatory submission support. Experienced biostatisticians work across Phase I–IV trials and multiple therapeutic areas.
Need specialist support?
IDDCR Global Research provides specialist CRO services for pharmaceutical, biotechnology, and medical device sponsors. Contact us to discuss your requirements.