Clinical Statistical Programming
Resource Guide
Part 1: Key R Topics
Core R Programming
• Base R (data types, vectors, lists, data frames, loops, functions)
• R Markdown / Quarto for reproducible reports
• Package management (CRAN, Bioconductor, internal packages)
• Writing reusable functions and R packages
Data Manipulation
• tidyverse (dplyr, tidyr, purrr, stringr, lubridate)
• [Link] for high-performance data wrangling
• Merging, reshaping (wide ↔ long), and cleaning clinical datasets
• Handling dates, missing values (NA), and special characters
Clinical Data Standards
• CDISC standards: SDTM and ADaM dataset creation in R
• haven – reading SAS .sas7bdat and .xpt (XPT) files
• xportr – writing FDA-compliant XPT files
• admiral – ADaM dataset derivation (Roche/GSK open-source package)
• metacore / metatools – metadata-driven programming
TLF Production (Tables, Listings, Figures)
• rtables – clinical-grade table engine (Roche)
• tern – TLF functions built on rtables
• gt / gtsummary – summary tables
• flextable – Word/RTF-compatible tables
• Tplyr – grammar-based table creation
• ggplot2 – publication-quality plots (KM curves, forest plots, etc.)
• survminer – survival analysis plots
• forest plot packages (forestplot, ggforestplot)
Statistical Analysis
• Descriptive statistics (summary, frequency, cross-tabs)
• survival package – Kaplan-Meier, Cox regression
• lme4 / nlme – mixed models (MMRM)
• mmrm package – MMRM for clinical trials
• coin, exactRankTests – non-parametric tests
• Multiple comparisons (multcomp, [Link])
• meta / metafor – meta-analysis
Regulatory & Validation
• Validation in a GxP environment (IQ/OQ/PQ concepts)
• Writing validation documentation for R packages
• valtools – R package validation framework
• CSV/XPT traceability and audit trails
• Understanding 21 CFR Part 11 implications
Reporting & Output
• officer – generating Word and PowerPoint outputs
• rmarkdown / knitr – automated report generation
• Quarto – next-gen reporting (HTML, PDF, Word)
• Outputting RTF, PDF, HTML from R
Workflow & Collaboration
• renv – reproducible package environments (critical for GxP)
• Git / GitHub / GitLab – version control
• targets – pipeline orchestration (like Make for R)
• YAML/JSON for configuration files
• CI/CD basics for validated environments
Pharma-Specific Packages Ecosystem
Package Purpose
admiral ADaM derivations
xportr XPT export with metadata
rtables Clinical tables
tern Analysis functions
renv Environment lockfile
mmrm Mixed models
Tplyr Table grammar
gtsummary Summary tables
Part 2: Essential SAS Topics
Phase 1: SAS Basics
• SAS Environment: SAS windows, libraries (WORK, perm), LIBNAME, datasets,
running programs.
• DATA Step Basics: Structure, creating datasets, variables, LENGTH/LABEL/
FORMAT, IF/ELSE, subsetting (WHERE/IF), KEEP/DROP/RENAME.
• Basic PROCs: PRINT, CONTENTS, SORT, FREQ, MEANS, UNIVARIATE.
Phase 2: Intermediate SAS
• DATA Step (Deeper): SET, MERGE, BY-group processing (FIRST./LAST.), RETAIN,
ARRAY, DO loops, functions, INFILE/INPUT.
• PROC SQL: SELECT, FROM, WHERE, ORDER BY, GROUP BY, Joins, Creating tables,
Aggregate functions.
• Formats & Dates: PROC FORMAT, SAS Dates (numeric storage), Functions
(TODAY, MDY, INTCK, INTNX).
Phase 3: Clinical SAS Essentials
• CDISC Basics: Theory, SDTM/ADaM, why standards matter.
• Clinical Trial Basics: Phases (I-IV), key terms, protocols, SAP, TLFs.
• Macro Basics: %LET, %MACRO/%MEND, parameters, conditional logic, usage.
• ODS: ODS PDF/RTF/EXCEL, Titles/Footnotes.
Phase 4: Reporting
• PROC REPORT: Column/Define statements.
• PROC TABULATE: Table structures.
• Graphs: PROC SGPLOT.
Part 3: Clinical SAS Programming Theory
Clinical Trial Basics
• Phases: Phase I (safety), Phase II (efficacy), Phase III (confirmation), Phase IV
(post-marketing).
• Key Terms: Protocol, Investigator, Sponsor, Informed Consent, Randomization,
Blinding, Placebo, Endpoints, AE/SAE.
Key Clinical Documents
• Protocol: Master document.
• SAP (Statistical Analysis Plan): Analysis methods.
• CRF (Case Report Form): Data collection source.
• CSR (Clinical Study Report): Final submission.
Study Populations
• ITT (Intent to Treat), mITT, PP (Per Protocol), Safety Population, FAS.
CDISC Standards
• CDASH (collection), SDTM (raw/tabulation), ADaM (analysis), Define-XML
(metadata), SEND (non-clinical).
• SDTM Domains: DM, AE, CM, LB, VS, EX, MH, DS, PE, TU.
• ADaM Datasets: ADSL, ADAE, ADLB, ADVS, ADTTE, ADEFF, ADCM, ADMH.
Validation & Regulatory
• Validation: Double programming, QC, documentation (Specs, Test Plan, Results).
• Regulations: ICH Guidelines (E6, E8, E9, E3), 21 CFR Part 11, FDA requirements.