QUINTILES BIOSTATISTICS 2005

1
QUINTILES BIOSTATISTICS 2005
NCSU Statistics Career Information Fair September
12, 2005
2
WELCOME
Ryan Thompson Lucille Sidberry Stacy
Woodard Pamela Masters
3
Biostatistics Job Roles

• Biostatistician
• Plans study and analysis, co-ordinates production
  and review of analyses and outputs, writes
  report, interaction with customer
• Statistical Programmer
• Develops (primarily SAS) programs to produce
  statistical analyses and outputs
• Independent verification of all analysis output
• Harmonized Job Descriptions
• Across and between job role grades
• Across sites
• Global Operating Procedures and Work Instructions

4
Process Overview
5
Biostatistician Responsibilities

• Key Responsibilities Include
• Development/Review of Protocol and CRFs
• Statistical Analysis Plan and Table Shells
• Development of QC Plan with Lead Team Members
• Review Data/Edit Checks/Consistency Checks
• Perform and Oversee Analyses
• Organization of Work, Monitoring Timelines and
  Resources, Timely Resolution of Data/Analysis
  Issues
• Interpretation/Writing of Study Report
• Continuous Communication with Project Manager,
  Team Members, and Customer

6
Statistical Programming

• Key Features
• Based on Statistical Analysis Plan and Table
  Shells
• Development Completed Prior to Database Lock
• Reviewed by Biostatistician and QC Specialist
• Validated Software (e.g., SAS)
• Documented Programs
• Internal Documentation of QC and Biostatistical
  Review
• Flexibility for Customer-specific Macros/Programs

7
Statistical Analysis Timelines

• Additional Methods for Managing Timelines
• Creation of Model Reports to ensure objectives
  are being addressed satisfactorily
• Early Data Reconciliation (AEs and SAEs)
• Development and validation of analysis programs
  using actual data and dummy treatment assignments
  prior to database lock
• Building analysis programs in a modular plan when
  awarded multiple projects
• Writing text (especially methods) in modular plan
  when awarded multiple studies or pivotal studies
  and ISS/ISE

8
Process for Managing Statistical Work

• Initial discussions prior to award to clarify
  project expectations, objectives, and strategic
  importance
• Post-Award Kick-off meeting attended by Lead
  Analysis Team Members and Sponsor
• Internal Kick-off meeting with all Analysis Team
  Members and Management Representative to provide
  consistent messages regarding scope, objectives,
  timelines, budget, and project-specific
  guidelines/procedures
• Weekly (on average) Analysis Team Meetings to
  review project status, objectives, milestones
  (completed and pending), and issues.
• Teleconferences with Sponsor (frequency dictated
  by project timelines and needs) and Lead Analysis
  Team Members to review project status and
  communicate project issues and obtain resolutions.

9
Process for Managing Statistical Work

• Email discussion groups can be set up to
  facilitate nonverbal communications between
  Sponsor and Analysis Team. This will allow
  issues to be distributed as they develop and not
  to be limited to scheduled teleconferences.
• As the Analysis Team proceeds with development of
  study documents (Statistical Analysis Plans,
  Analysis Specifications), these documents will be
  transferred to Sponsor reviewers. Follow-up
  teleconferences or meetings to finalize these
  documents will precede the initiation of program
  development.
• Quintiles Analysis Team will develop programs
  based upon approved study documents and will test
  these programs prior to production.
• If requested, draft analyses using dummy
  treatment codes can be produced for Sponsor
  review/approval prior to database lock. This
  provides an opportunity for expediting the final
  output production and review process post
  treatment unblinding.

10
Process for Managing Statistical Work

• Final analysis datasets, outputs, and programs
  will be produced, QCd, and undergo statistical
  review prior to final delivery to Sponsor.
  Formatting of these deliverables will have been
  defined at the initiation of the project (to
  allow efficient program development).
• Upon final approval from Sponsor, the Quintiles
  Analysis Team will return any original study
  documents (e.g., original CRFs) to Sponsor and
  will archive all study documents, programs,
  files, and outputs following the Quintiles
  Project Archiving SOP. Any information which
  Sponsor wishes to archive in-house can be
  provided upon request.

11
Courses to Consider

• ST552 - Linear Models and Variance Components
• ST732 - Applied Longitudinal Data Analysis
• ST744 - Categorical and Censored Data Analysis
• ST745 - Analysis of Survival Data
• PSY436 - Introduction to Psychological
  Measurement
• PSY535 - Tests and Measurements

12
Examples of Recent Indications Analyzed

• Anxiety
• Depression
• Schizophrenia
• Bipolar Disorder
• Epilepsy
• Migraines
• Alzheimers disease
• ADHD

• Glaucoma
• Retinal Detachment
• Osteoporosis
• Breast Cancer
• Diabetes
• Irritable Bowel Syndrome
• Hepatitis
• Sepsis