Correctness, Consistency, Completeness, and Appropriateness in Statistical Analysis Plans: A Guide for Clinical Programmers
Introduction
Statistical Analysis Plans (SAPs) play a crucial role in clinical trials, outlining the methods of data analysis and ensuring that results are reliable and valid. As part of the SAP review process, clinical programmers need to conduct various checks to ensure that the document meets necessary standards. This article will guide clinical programmers through the key aspects of SAP review, focusing on correctness, consistency, completeness, and appropriateness. With a better understanding of these checks, clinical programmers will be better equipped to provide valuable input during the SAP review process and ensure that the document is of high quality.
Correctness
The first check in the SAP review process is correctness, which involves ensuring that statements in the SAP are accurate. This includes checking for typographical or grammatical errors, as well as verifying the correctness of scientific descriptions. Clinical programmers should also ensure that the correct SAP template is used, if applicable, and that the document follows the standardized structure. In addition, the author should not be allowed to delete or add sections on the first level unless specified, nor should they be allowed to add a Table of Contents (TOC) for all the Tables, Figures, and Listings (TFLs) in the SAP text.
Consistency
The second check in the SAP review process is consistency, which involves ensuring that the SAP is in line with the protocol and any project-specific standards. Clinical programmers should compare the SAP text with existing documents (e.g., protocol, project standard definitions, company data standards, or analysis programs) to ensure consistency. In the event of inconsistencies between the protocol and the SAP, the clinical programmer should note these inconsistencies and comment on them as part of the SAP review process. It is important to remember that inconsistencies may be allowed between the SAP and project standards, but clinical programmers should be aware of them and provide comments.
Completeness
The third check in the SAP review process is completeness, which involves ensuring that the TFLs described in the SAP text are completely covered by the List of Data Displays and the mock example shell TFLs. Clinical programmers should compare the SAP text with the List of Data Displays and the mock example shell TFLs to ensure that they are consistent and complete. Any inconsistencies should be identified and addressed, as they could lead to incorrect programming and potentially impact the study results.
Degree of Technical Details
The fourth check in the SAP review process is ensuring that the right degree of technical details is provided. This check requires more statistical programming expertise and some statistical knowledge. Clinical programmers should ensure that they have sufficient technical details to perform the programming, including definitions of baseline variables, analysis visit windows, data imputation rules, derived variables, summary statistics, denominators for percentages, and statistical models.
Appropriateness
The final check in the SAP review process is appropriateness, which involves ensuring that the planned analyses are suitable for addressing the study objectives. This check requires some statistical and regulatory knowledge, and not all clinical programmers may possess these skills. However, a careful review by a clinical programmer can ensure a deeper understanding of the statistical environment and that the final SAP meets the reporting requirements.
Conclusion
In conclusion, clinical programmers play a critical role in the SAP review process, ensuring that the document is correct, consistent, complete, and appropriate. By conducting these checks and working closely with the trial statistician, clinical programmers can help ensure that the SAP is of high quality and that the study results are reliable and valid. With a proactive approach and a willingness to ask questions, clinical programmers can make a valuable contribution to the success of clinical trials.
