In complex or longer duration clinical studies, it is common to submit the data up to a predetermined date, often known as the Data Cutoff Date. This can refer to a data cleaning date, database lock date or the date when specific events or milestones are achieved. Since data continues to be entered for ongoing subjects, using a cutoff date allows analysis to be focused up to a projected time. A data cutoff (DCO) is commonly used in interim analysis of oncology studies. Below, we will discuss how data cutoff is implemented in SDTM database, since SDTMs are the source data for submission. 

What is a data cutoff? 

The process of creating a subset of data and restricting it up to a target date is called the Data Cutoff. Analysis can be as simple as excluding any data after the cutoff date or it can require complex derivation to handle data beyond cutoff.   

Ways to implement DCO in database 

a.  SDTM built using restricted raw (CRF and vendor) data: 

• In this approach, raw datasets are created by applying data cutoff to source datasets (CRF + vendor). SDTMs are built using the restricted raw database. 

• Pros: 
  • Increases transparency: DCO application is clear and easy to review in restricted datasets, which enhances transparency between source and cutoff applied datasets. 
• Cons:  
  • Extra step: Additional step is required to validate restricted raw datasets to confirm correct DCO application, before using them as source to SDTM. 
  • Not cost efficient: Validating source datasets may increase cost and complexity. 

 b.  DCO accounted within SDTM programming: 

• DCO is applied in the SDTM programs at the development stage. SDTM output datasets are restricted by using the DCO date.  

• Pros:  
  • No extra source datasets: Since DCO application is validated along with SDTM data validation, no need to create or validate additional raw datasets. 
  • Cost Efficient: This method is the most cost effective as it reduces the steps for applying DCO in a separate database. 
• Cons:  
  • Lack of traceability: There is no source data to compare against. It may be harder for sponsors to verify details of DCO. 

How to cut the database? 

It is important to understand how to use the DCO date to restrict data for an ongoing study.  

Types of Dates for DCO applications: 

• Assessment dates – these are the date of visits or assessments 
• Start and End dates – these are the dates of events or interventions 

Applying a data cutoff requires special considerations for the different data types: 

1. Survival (Death) and follow-up assessments: DCO needs to be applied correctly to handle data up to a cutoff date in time sensitive analysis such as survival and follow-up.  
2. Adverse Events: Both the start and end dates and outcome of the adverse events should be carefully reviewed on how to restrict the data, to ensure data is captured correctly for analysis.  
3. Concomitant Medications: Start and end dates of concomitant medications, as well as ongoing/concurrent flags should be reviewed for DCO application, to ensure data fulfills the requirements for analysis. 
4. Disposition: End of treatment or study termination dates are important dates to determine treatment completion, discontinuation or end of study participation. It is important to know how to restrict these records to ensure accurate analysis. 
5. Exposure dates: The start and end dates of exposure to the study drug or intervention plays an important role for calculating treatment duration and study drug compliance. It is necessary to have a clear understanding of how to restrict the data for these dates. 

Best Practices: 

• Confirm DCO date from all stakeholders, including sponsor and study statistician. 
• Document DCO rules, which can be reviewed by study team to ensure that data is restricted correctly for analysis. Creating DCO specifications can help programmers, validators and reviewers to ensure successful implementation of DCO in SDTMs. 
• A standard macro can be created to apply DCO to all SDTMs, ensuring data is captured correctly.  

Submission documentation:  

• The process of DCO application should be clearly documented in the reviewers’ Guides. 
•  The trial summary (TS) data should be updated for the Data Cutoff Description (DCUTDESC) and Data Cutoff Date (DCUTDTC). 

The data cutoff must be accurately applied in SDTMs, as it can significantly impact trial outcomes. Sponsors and study teams must collaborate to define and document the process of DCO application and implementation to maintain data integrity and traceability for regulatory submissions. 

Questions about how to apply a cutoff date in SDTM with ongoing studies? Contact us to speak with one of our Data Standards Specialists. 

Saloni Shah, Associate Director, Data Standards, has more than 18 years of experience leading statistical programming teams and providing technical oversight in the CRO and pharmaceutical industry, with a focus on oncology therapeutic areas. An expert in CDISC standards, she specializes in SDTM, ADaM and eSubmissions, offering guidance on data integration for Integrated summary of Safety (ISS) and Integrated Summary of Efficacy (ISE), as well as the creation of define and reviewer’s guides. Throughout her career, Ms. Shah has successfully led statistical programming teams from project initiation to completion, ensuring high-quality data standards and seamless regulatory submissions across  multiple therapeutic areas.