New Clinical SAS Programmer Topic
The process chart and topics below useful to new Clinical SAS® Programmers. Other topics within the pharmaceutical section are more advanced topics such as CDISC. See also Statistical Analysis and New to SAS® Programming.
SAS Development for New Clinical Study Checklist, MindMap
Objectives:
1. Understand structure of clinical trials and how data is collected over time.
2. Understand how raw clinical data from case report forms are stored in SAS datasets.
3. Understand how to clean clinical data.
4. Know how to create descriptive statistics.
5. Know how to apply statistical modeling on clinical data.
Pharmaceutical Terms:
Baseline, SAP, Table Lookup, Visit Windows
Double-Blinded Clinical Study: Both patient nor the site knows which drug is taken. Most clinical studies are double-blinded with the sponsor blinded until the study is unblinded.
Triple-Blinded Clinical Study: Patient nor the site or the sponsor knows which drug is taken. Few clinical studies are classified as triple-blinded.
Study Day: Day 1 will be defined as the first date on which study drug was administered.
Positive study days will be counted forward from Day 1. Day -1 will be the date immediately
preceding Day 1, and negative study days will be counted backward from Day -1. Day 0 is not a value value.
Baseline: For all other parameters, the baseline measurement will be the pre-dose value collected
on Day 1 or if not available, then the last value collected before Day 1. Baseline is not Day 0 since study day 0 is not valid.
Change: Change from baseline at a particular post-baseline time point will be computed as the
value at the post-baseline time point minus the baseline value.
Duration of Exposure: Duration of exposure (weeks) will be computed as the date of the last
dose of study drug minus the date of the first dose of study drug, plus 1 day (that is, the study day associated with the date of the last dose of study drug) divided by 7 days per week.
Visit Windows: Measurements will be associated with a visit for summarization according to the
study day associated with the date on which the information was collected. Target dates and the
acceptable range of study days for each visit are presented in Table 2. If multiple visits occur
within a visit window, the visit occurring closest to the target day will be selected for
summarization. If there is a tie, the earliest visit will be chosen.
Sample Clinical Trials Study
Timeline: Start study, End study, SDTMs/ADaMs, Database lock and TLGs
SAP - Statistical Analysis Plan, Template , PhUSE
CRF - Case Report Forms
Three types of data collected
1) One record per patient, ex. demog
2) Measurements during protocol visits, may need to sort and subset to get one record per patient such as first dose date or lab baseline flag, ex. vitals, labs, ex, etc.
3) Measurements any time during the study, ex. adverse events, con meds
Three types of joins
1) One to one record, ex. demog with first dose date from ex
2) One to many records, ex. demog with vitals
3) Many to many records using Proc SQL, maybe required if one visit date is used as reference to anther visit date by visit name, ex. adverse events with con meds
SDTM Specification, ADaM Specification
TLGs Table Shells - Tables, Lists and Graphs QC Checklist
Two types of analysis
1) Efficacy - based on the primary and secondary endpoints, ex. change in lab measurements from baseline, survival rates
2) Safety - based on adverse events and subject disposition
Five types of Tables
1) Data Listing, ex. demog characteristics
2) Counts and Frequency of categorical data, ex. proportion of patients with adverse events
3) Summary Table of continuous data, ex. descriptive statistics (mean, sd, min, max) by visit
4) Statistical Analysis Table to model data, ex. descriptive statistics by visit with p-value, survival analysis using Kaplan-Meier (Proc LIFETEST)
5) Graphs, ex. lab scatter plot
1. Clinical Trials Terminology for SAS Programmers, Sy Truong
2. SAS® PROGRAMMER TO CLINICAL SAS PROGRAMMER, Gayatri Karkera, Neha Mohan
3. Success As a Pharmaceutical Statistical Programmer, Sandra Minjoe, Mario Widel
4. THE ROLE OF SAS PROGRAMMERS IN CLINICAL TRIAL DATA ANALYSIS, Ming Wang
5. SAS® Programming for the Pharmaceutical Industry, Brian C. Shilling, Carol Matthews
6. The 5 Most Important Clinical SAS Programming Validation Steps, Brian Shilling
8. Training Statistical Programmers on SAP Review Skills, Sascha Ahrweiler
9. Intro to Longitudinal Data: A Grad Student “How-To” Paper, Elisa Priest,Ashley Collinsworth
10. Longitudinal Data Techniques: Looking Across Observations, Ronald Cody
11. Statistics for Clinical Trial SAS Programmers 1: paired t-test, Kevin Lee
12. Clinical Trial Reporting Using SAS/GRAPH® SG Procedures, Susan Schwartz
13. CDISC Introduction Presentation
14. Introduction - Introduction to the CDISC Standards, Sandra Minjoe
15. Toward a Comprehensive CDISC Submission Data Standard
16. CDISC: Why SAS® Programmers Need to Know, Victor Sun
17. Oncology Trials 101 - The Basics and Then Some, Dave Polus
SAS Programming in the Pharmaceutical Industry book
Common SAS Programming FAQ Index
SAS solution found? Please log minutes saved.
