CDISC SDTM, ADaM & SEND
SDTM organizes study tabulation data; ADaM supports analysis datasets and traceability; SEND structures nonclinical study data. CDISC also provides a broader family of clinical research standards.
Capability / YouSophia
Health Data Science and Data Standards
Apply scientific and data-driven methods to help teams understand clinical, nonclinical, and other health evidence.
THE WORK AT A GLANCE
CURATEPrepare study dataQuality, terminology and standards
ANALYZEExplore the evidenceStatistical methods and visualization
INTERPRETCommunicate findingsContext, limitations and review
Our focus
Make complex health evidence more useful
We bring a thoughtful, multidisciplinary approach to the needs of public-sector teams and the people they serve.
01 /
Study data analysis
Support analytical questions with well-documented approaches to clinical and nonclinical data.
02 /
Statistical computing
Use statistical programming and visualization approaches to develop analysis-ready data and clear outputs.
03 /
Data harmonization
Bring together data from different studies or sources for consistent analysis.
04 /
Research support
Assist with scientific investigations, evidence synthesis, and interpretation of results.
A visual perspective
The building blocks of health data science
These connected elements help frame the work and the people it is intended to support.
01Study information
02Data preparation
03Scientific methods
04Interpretable evidence
Illustrative framework; activities depend on mission needs.
The evidence lifecycle
Turn health data into reviewable evidence
Health data science combines subject matter understanding, reproducible analysis, and careful interpretation. Each stage helps reviewers understand both what the evidence shows and where uncertainty remains.
01 /Understand
Define the research question, relevant populations, data sources, and limitations.
02 /Prepare
Curate, harmonize, and document clinical or nonclinical datasets.
03 /Analyze
Apply statistical methods, programming, visualization, or simulation appropriate to the question.
04 /Interpret
Review results, assess assumptions, and communicate findings in context.
In practice
What this work can look like.
Activities are shaped around the mission, the available evidence, and the needs of the people doing the work.
ILLUSTRATIVE OUTPUTS
Data dictionaries, analysis plans, reproducible scripts, tables and figures, and scientific summaries.
01Prepare analysis-ready datasets and examine consistency across multiple studies.
02Develop tables, listings, and figures to help reviewers explore study results.
03Support scientific investigations through literature review, data analysis, and documented methods.
Health data science and data standards
Data standards for connected, reviewable evidence.
Standards, models, terminologies, and dictionaries serve different roles. Together they can help teams structure, exchange, interpret, and analyze health and study data. The applicable standard and version depend on the use case and agency requirements.
HL7 & CDA
HL7 develops standards for the exchange and integration of health information. The Clinical Document Architecture (CDA) specifies the structure and semantics of clinical documents.
OMOP Common Data Model
The Observational Medical Outcomes Partnership (OMOP) Common Data Model provides a standardized structure and vocabulary approach for observational health data analysis.
LOINC & SNOMED CT
LOINC identifies health measurements, observations, and documents. SNOMED CT provides clinical terminology for representing health concepts.
MedDRA & ICH E2B
MedDRA is a medical terminology used in regulatory activities. ICH E2B XML R2 and R3 provide formats for electronic exchange of individual case safety reports.
ISO, IDMP & product dictionaries
ISO publishes standards used across many domains; IDMP is a set of ISO standards for identification of medicinal products. The FAERS Product Dictionary and WHO-Drug provide product and drug terminology resources relevant to safety information.
The supplied draft statement of work identifies these standards and resources except OMOP, which was added at your request. Standards selection and implementation should be confirmed for the specific project. FDA Study Data Standards Resources ↗
A closer look
Scientific rigor from question to interpretation
Health datasets can vary by study, population, terminology, and purpose. Thoughtful preparation and transparent methods help teams use that information responsibly.
Study data preparation
Review source formats, metadata, variables, and terminology. Document cleaning, transformations, and assumptions so a dataset can be understood and reused.
Statistical computing and visualization
Support statistical programming, tables, listings, figures, simulations, and exploratory analyses using methods suited to the question and the available data.
Scientific synthesis
Connect analytical findings with relevant literature and domain knowledge. Describe limitations, uncertainty, and areas requiring additional scientific review.
Illustrative scenario
Putting the work in context.
For an analysis spanning several studies, support could begin with a review of data definitions and quality, followed by harmonization, reproducible analysis, and visual summaries for subject matter experts.