SDTM, or the Study Data Tabulation Model, is a standardized format for organizing clinical trial data before submitting it to regulatory agencies like the FDA. Developed by CDISC (Clinical Data Interchange Standards Consortium), SDTM provides a consistent structure so that every sponsor, whether a global pharmaceutical company or a small biotech firm, presents its raw trial data in the same way. This standardization allows regulators to review submissions faster and makes it possible to compare data across different studies and programs.
Why SDTM Exists
Before standardized formats, every company organized clinical trial data differently. A lab result might be stored in one format by one sponsor and a completely different format by another. Regulatory reviewers had to spend significant time just figuring out how each dataset was structured before they could evaluate whether a drug was safe and effective.
SDTM solves this by creating a universal blueprint. It streamlines data collection, management, analysis, and reporting. Beyond regulatory review, it supports data warehousing, allows researchers to mine and reuse datasets from past trials, and makes it easier for companies to share data with partners or for due diligence activities like acquisitions. When the FDA or Japan’s PMDA receives a submission in SDTM format, reviewers can apply consistent analytical tools and more easily spot trends, gaps, or safety signals across the data.
Is SDTM Required?
Yes, for most submissions to the FDA. As of March 15, 2022, SDTM-formatted data is required for new drug applications (NDAs), abbreviated new drug applications (ANDAs), and certain biologics license applications (BLAs). For noncommercial investigational new drug applications (INDs), the requirement kicked in on March 15, 2023. The current required version is SDTM Implementation Guide (SDTMIG) v3.3, based on SDTM v1.7, though a newer version (SDTMIG v3.4, based on SDTM v2.0) was published in November 2021.
Japan’s PMDA also requires CDISC standards for electronic study data submissions. Globally, regulators are increasingly aligning around these standards. Adherence is strongly encouraged even in regions where it isn’t yet mandatory, as it aligns submissions with international expectations and reduces friction during review.
How SDTM Organizes Data
SDTM sorts clinical trial observations into three main classes, each covering a different type of data collected during a study.
- Interventions: What was given to the patient. This includes the investigational drug itself (called “exposure”), any concomitant medications the patient was already taking, and self-administered substances like alcohol, tobacco, or caffeine.
- Events: What happened to the patient. This covers adverse events (side effects), medical history from before the trial, and protocol milestones like randomization or study completion. For example, “Subject 101 had mild nausea starting on Study Day 6” is an observation that belongs to the Adverse Events domain.
- Findings: What was measured or observed. This includes lab test results, ECG readings, vital signs, and answers to questionnaires.
Within these classes, data is grouped into specific “domains,” each represented as its own dataset. The Adverse Events domain (AE) captures side effects. The Laboratory Tests domain (LB) holds lab values. The Demographics domain (DM) is a special-purpose dataset included in every study, containing basic information about each subject like age, sex, and race. There are also domains for trial design information, such as the planned sequence of study phases.
Each domain follows a predictable structure with standardized variable names. A variable for the start date of an adverse event, for instance, always has the same name regardless of the sponsor or the therapeutic area. This predictability is what makes automated review tools possible on the regulatory side.
The Define-XML File
SDTM datasets don’t travel alone. Every submission must include a metadata file called Define-XML, which acts as a detailed roadmap for the data. It tells regulators exactly what datasets are included, what each variable means, what controlled terminology was used, and how values were coded. Think of it as a data dictionary packaged alongside the actual numbers. Define-XML is a required standard for submissions to both the FDA and PMDA.
How SDTM Differs From ADaM
SDTM and ADaM are companion standards that serve different purposes in the submission process. SDTM captures the raw, collected data in a tabulated form, essentially a clean record of what was observed during the trial. ADaM (Analysis Data Model) takes that raw data and reshapes it into datasets specifically designed for statistical analysis, such as the datasets behind the tables, figures, and efficacy conclusions in a submission.
Traceability between the two is a core principle. Every variable in an ADaM dataset must be traceable back to its source in the SDTM data. This means a reviewer can follow any statistical result back to the original observation, verifying that no data was lost or inappropriately transformed along the way. Sponsors are expected to thoroughly document how each ADaM variable was derived from the SDTM source.
What SDTM Means in Practice
For pharmaceutical and biotech companies, SDTM compliance is a significant operational consideration. Data managers and programmers must map raw data from case report forms and electronic systems into SDTM-compliant datasets, ensuring the right variables land in the right domains with the correct terminology. This mapping process, sometimes called SDTM conversion, typically happens after data collection is complete but before the statistical analysis begins.
Getting it right matters. Submissions with SDTM errors or inconsistencies can trigger questions from regulators, potentially delaying the review timeline. Companies often invest in specialized software and trained staff to handle the conversion, and CDISC publishes conformance rules that help teams validate their datasets before submission. For anyone working in clinical data management, biostatistics, or regulatory affairs, understanding SDTM structure is a foundational skill.