Research Methodology and Biostatistics

Introductory Course Workshop

4-6 March, 2005

Pharmasri Anita Das MBBS, MPH

[Link]
Clinical Research Protocol
1. What is a protocol? Definition
2. Why a protocol? Can’t we do without it?
3. How to write a protocol? Protocol development

Pharmasri [Link]
Protocol is essentially….
a written plan of the study

Well known devices for seeking grant funds

Vital scientific function: help organise research in a
logical, focused & efficient way

Pharmasri [Link]
Protocol defined
ICH GCP 1.49
A document that describes the objective(s), design,
methodology, statistical considerations and
organization of a trial.

All research should be conducted according to a written,

clear, comprehensive and consistent protocol.

Pharmasri [Link]
Example of Research Protocol
Title of Project
Statement of Problem
Relevance of problem
Literature review
Statement of Objectives
Outcome variables
Statement of research hypothesis

Pharmasri [Link]
Health Research Methodology: A Guide for Training for Research Methods, WHO 2001
Research Methodology
Summary
Research Design – Research strategies & setting
Sampling
Use of controls
Study instruments
Short description of plans for data collection
Analysis and interpretation plan

Pharmasri [Link]
Health Research Methodology: A Guide for Training for Research Methods, WHO 2001
More Examples
1. TITLE PAGE
2. INVESTIGATORS AND STUDY
ORGANISATION
3. SIGNATURE PAGE
4. SYNOPSIS
5. TABLE OF CONTENTS
6. LIST OF ABBREVIATIONS
7. GLOSSARY OF TERMS
8. INTRODUCTION and BACKGROUND
9. OBJECTIVES
Pharmasri [Link]
ICH E6: Good Clinical Practice, Chapter 6
10. STUDY DESIGN
11. STUDY POPULATION & SUBJECTS
12. STUDY TREATMENT
13. ASSESSMENT OF EFFICACY
14. ASSESSMENT OF SAFETY & AE
15. CLINICAL ECONOMICS
16. STUDY CONDUCT
17. DATA MANAGEMENT
18. STATISTICAL METHODS
19. ETHICS & REGULATORTY
20. ADMINISTRATIVE MATTERS
21. REFERENCES
22. APPENDIX

Pharmasri [Link]
ICH E6: Good Clinical Practice, Chapter 6
 2. Why a protocol?
Can’t we do without it?

Why is a written, clear, comprehensive and consistent

protocol so important to research?
1. Justification: justify the research
2. Evaluation: basis for reviewing the research
3. Specification: basis for agreement
4. Blueprint: basis for planning & implementation

Pharmasri [Link]
1. Justification
All research requires justification.
The protocol is testimony to the fact that the
investigators have sufficiently thought through the
proposed research concerning its justification,
scientific quality and ethical propriety.

Pharmasri [Link]
ICH GCP on trial justification

2.2 Before a trial is initiated, foreseeable risks and

inconveniences should be weighed against the anticipated
benefit for the individual subject and society. A trial
should be initiated and continued only if the anticipated
benefits justify the risks.
2.4 The available non clinical and clinical information on an
investigational product should be adequate to support the
proposed clinical trial.

These are summarized in rationale section of protocol.

Pharmasri [Link]
2. Evaluation
All research requires review by both external and
internal parties.
The protocol provides the basis for review.

Who might want to review a protocol?

Pharmasri [Link]
 Protocol reviewers

#. Who? Uses of protocol

External
1. Funder/Sponsor Show research merit and utility, and researcher
capability.
2. Scientific community Show research is scientifically sound; hence
publishable.
3. Ethics committee Show ethical propriety
4. Regulatory agency Show compliance with applicable regulation
and laws.
5. Marketing authority Show validity of research results
Internal
6. Co-investigators Secure collaboration
7. Internal QC Ensure protocol comply with GCP, regulation,
internal SOP.
8. Each key participant Ensure protocol is sufficient, clear and
consistent to each; ensure protocol a true
Pharmasri reflection of practice.
[Link]
3. Specification
basis for agreement
All research requires collaboration.
A protocol specifies how the research is to be
conducted by all parties involved.
1. Contract between sponsor and investigator on how
the research is to be conducted. (Hence you sign and
date a protocol.)
2. Agreement among investigators on uniform way to
conduct a research.

Pharmasri [Link]
 4. Blueprint

All protocol eventually has to be operationalised to

become actual research.
A protocol provides the basis for planning and
implementing the research.

Pharmasri [Link]
 People involved
Many professions are involved in
Implementation:
1. CRA
 CRF development and data collection mechanism, monitoring
convention.
2. CDM
 data definition, database development, query generation
3. Site coordinator
 site management, data collection, specimen/material handling etc
4. Statistician
 randomisation, data monitoring, analysis
5. Investigator
 techniques, schedule
6. Reference centres
 techniques, schedule.

Pharmasri [Link]
3. How to write a protocol? Protocol
development
A protocol is a joint product of the different
professions involved
The best way to write an impressive protocol is to
“copy” another one.
Must have SOP and follow guideline.

Pharmasri [Link]
Protocol development- 1
1. TITLE PAGE
Protocol #
Author
Protocol Date
Amendment # and Date
Author’s and Reviewer’s signature and date

2. INVESTIGATORS & ORGANISATION

Principal and co-Investigators
Sponsor and Research organisation
Reference centres, Study Committees etc
Pharmasri [Link]
Protocol development- 2

3. SIGNATURE PAGE
4. SYNOPSIS
5. TABLE OF CONTENTS
6. LIST OF ABBREVIATIONS
7. GLOSSARY OF TERMS
Definition of common and technical terms

Pharmasri [Link]
Protocol development- 3

8. INTRODUCTION and BACKGROUND

Literature review
Rationale and motivational background justifying the
proposed research
Make reference to Investigator’s brochure for details

Pharmasri [Link]
Protocol development- 4

9. OBJECTIVES
A. Primary question: the major question the investigators
are most interested in.
B. Secondary or subsidiary questions :
relate to subsidiary end points.
relate to subgroup hypotheses eg other than overall
contrast between each arm of trial.
C. Ancillary questions: questions that do not bear
directly on the intervention being tested but which are
nevertheless of interest; eg. natural history of disease
in the control group, risk factor analysis.
Pharmasri [Link]
 Protocol development- 5
10. STUDY DESIGN
Study design, # patients & Duration
Schematic diagram of study design

Screen Baseline Randomise Rx

Wash-in/out End

Rx group

Control group
Discussion of study design: Justify decision

Pharmasri [Link]
Protocol development- 6

11. STUDY POPULATION AND SUBJECTS

Inclusion criteria
Exclusion criteria
Subject withdrawal & Drop-out
Screening failures

Pharmasri [Link]
Protocol development- 7

12. STUDY TREATMENT

Study Drug
Dosage and Administration
Concomitant Medication/Treatment
Treatment Allocation and Randomisation
Blinding & Emergency Unblinding procedures
Assessment of compliance
Investigational product handling

Pharmasri [Link]
 Protocol development- 8

13. ASSESSMENT OF EFFICACY

Specify efficacy parameter
Primary end-point(s): Rx effect for primary objective
Secondary end-point(s): supportive measures related to Rx
effect for secondary objective
Global assessment variable: overall safety, treatment effect
and usefulness; requires judgement.
Surrogate variable: predictor of clinical benefit
Specify methods for assessing & recording
Specify who and when assessed.

Pharmasri [Link]
Protocol development- 9

14. ASSESSMENT OF SAFETY & AE

Eliciting, Documenting and Reporting Adverse Events
Assessment of Severity & Causality
Follow-up and Assessment of Outcome
Serious Adverse Events
Treatment of Adverse Event
Overdosage & Pregnancy
Breaking the Study Blind

Pharmasri [Link]
Protocol development- 10

15. CLINICAL ECONOMICS

16. STUDY CONDUCT
Study procedures and Flowchart
Detailed description of stages and visits
Sample handling and Analysis

Pharmasri [Link]
 Flowchart
  Screenin Baseline PCI In Follow Follow Final
g hospital up up visit
Visit 1 2 3 3 4 5 6

Timeline (month) -4 -1 0 12-24 1 3 6

hrs
Procedures              

Check eligibility X            

Screening Coronary angiogram X*            

Written informed consent X            

Medical History   X          

Vital sign   X          

Blood sample for Hematology   X          

(10 ml)
Blood sample for Chemistry   X          
Randomization to treatment   X          

Blood sample for CK   X X X X** X** X**

ECG     X X X** X** X**

Efficacy assessment     X X X X X

QCA     X X** X** X** X

Report AE and SAE     X X X    

CompletePharmasri
CRF [Link]
X X X X X X X
Protocol development- 12

18. STATISTICAL METHODS

Sample size and power considerations
Randomisation
End-points
Analysis sets
Final analysis
Planned Interim analysis

Pharmasri [Link]
Protocol development- 13

19. ETHICS & REGULATORTY CONSIDERATIONS

Institutional Review Board/Ethics Committees
Ethical conduct of the study
Subject information and Consent
Patient protection procedures

Pharmasri [Link]
Protocol development- 14

20. ADMINISTRATIVE MATTERS

Study documentation, CRF and Record keeping
Finance and Insurance
Study Termination
Confidentiality and Publication policy

21. REFERENCES

Pharmasri [Link]
 Protocol development- 15

22. APPENDIX
A. DECLARATION OF HELSINKI
B. STUDY PROCEDURES
C. QUESTIONNAIRE and CRF
D. INVESTIGATIONAL PRODUCT LABELS
E. ELEMENTS OF INFORMED CONSENT
F. SAMPLE INFORMED CONSENT FORM
G. PATIENT INFORMATION SHEET
H. LETTER OF INDEMNITY
I. INVESTIGATORS’ CURRICULUM VITAE

Pharmasri [Link]
 Case 1: To blind or not to blind

This is a multi-center, open label, parallel group randomized trial designed to

demonstrate the equivalence of Staysafe/ANDY-disc® with the standard
treatment Ultrabag® for a treatment period of 12 months. 260 patients on
CAPD from 5 participating sites who meet inclusion/exclusion criteria will be
enrolled into the trial. The primary efficacy variable is the incidence of
peritonitis, and secondary efficacy variables are technique failure and technical
problem encountered in using the products under investigation.
The investigational products here are Staysafe/ANDY-disc® and Ultrabag®.
The nature of the treatment (CAPD) and the 2 products are such that it is not
possible to blind study site personnel, study subjects, monitor, data
management and sponsor personnel to the assigned treatment.

Questions for discussion

a) What is at stake in the above situation?
b) Why is the issue not a trivial matter?
c) What would be your advice?
Pharmasri [Link]
Case 2: Sample size is not always what it
seems
For the above trial, the investigators intended to enroll 260
patients from 5 centres. Section 16.1 of the protocol on Sample
size and power considerations reads as follows:
“Sample size was estimated based on the primary efficacy
parameter, which is incidence of peritonitis rate. Based on
literature data and our practice, an estimate of the incidence of
peritonitis rate on CAPD is 0.05 episodes per patient month. We
accept as equivalent 25% difference (delta) in risk of peritonitis at
1 year. For a power of 0.8 and alpha of 0.05 (one sided), the
required sample size is 264 patients in each treatment groups
followed-up for 1 year. No allowance is made for drop out.”
Questions for discussion
Do you accept the statistician’s calculation above?

Pharmasri [Link]
 Case 3: P value and p value
A trial was designed to compare the effect of a new
drug A, the standard drug B and placebo P. The p value for
treatment effect of A versus P was 0.001, for B versus P was 0.02 and for A
versus B was 0.45.
Questions for discussion
a) What is your take on the above results?
b) Drug A is efficacious?
c) Drug B is also efficacious?
d) There is no difference in treatment effect between drugs A and B in
that study?
e) However, since the p value for A vs P is much smaller than p value
of B vs P, A is likely to be more efficacious than B, though of
course the study did not show that?

Pharmasri [Link]
 Case 4: It is not always possible to trial
Smoking is suspected to increase the risk of certain indolent
benign cancer (patients with the cancer don’t die quickly). The
best design is to randomise subjects to smoking or non-smoking, then follow
them up for many years to determine occurrence of the cancer. This is clearly
not ethically acceptable, even if one could afford to do the trial.
Questions for discussion
To address the above question, discuss the following design options:

Study a large group of smokers to determine how many of them have

a. the cancer. If the incidence is higher than generally expected, we may
conclude that smoking increases the risk of the cancer.
b. We can also follow up a group of smokers and a control group of non-
smokers. If the incidence is higher among smokers than non-smokers,
we may conclude that smoking increases the risk of the cancer
c. We can also assemble a group of patients with the cancer and a control
group without cancer. If the proportion of smokers is higher among
cancer group than non-cancer group, we may conclude that smoking
increases
Pharmasri the risk of the cancer
[Link]
References
Fundamentals of Clinical Trials
by Lawrence M. Friedman, Curt Furberg, David L.
Demets. Springer-Verlag; 3rd edition (November
1, 1998)
Health Research Methodology: A Guide for
Training for Research Methods, WHO 2001.
WHO Publications (Western Pacific Regional
Office)
Good Clinical Practice, Chapter 6. International
Committee on Harmonization E.
[Link]

Pharmasri [Link]