How to write a study protocol
Viviane Bremer
EPIET, Lazareto, Menorca
September 2011
1
What is the study protocol?
• Every step of a study
• Answers relevant questions
- public health problem important?
- study question relevant to problem?
- objectives consistent with study question?
- study design achieves objectives?
- sufficient power?
- public health impact of the findings?
2
Why write a study protocol?
• Check
- if objectives can be achieved
- feasibility of the study
• Prevent failure to collect crucial information
• Lays down the rules for all partners (quality)
• Obtain approval of ethical committee(s)
• Application for funds
3
Study protocol
How to start ?
• Formulate the research question
• Get
– good examples of protocols
– ideas from similar published studies
– ideas from colleagues
• Use a checklist of items to include
• Get the requested format
(grant application)
4
Protocol outline
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Presentation
Background and justifications
Objectives
Methods
Ethical considerations
Project management
Timetable
Resources
References
Appendices
5
1. Presentation
• Title
– short, accurate, concise
•
•
•
•
Investigators
Main centres
Steering committee (scientific board)
Summary of the protocol
6
2. Background and justification
• Statement of problem, study justification
 importance of subject area
• magnitude, frequency





gaps in existing knowledge
principal questions to be addressed
contribution of results to existing knowledge
use of results
dissemination of results
• Review relevant literature
7
3. Objectives
• Should answer the study question
• Should be S.M.A.R.T.
– Specific
– Measurable
– Achievable/Accessible
– Realistic/Relevant
– Time-based/Timely
8
3. Objectives
Principal objective
• Must be achieved
• Dictates design and methods
Secondary objectives
• Of interest, but not essential
9
Hypotheses
• Translation of the objectives
in terms that allow statistical testing
10
3. Objectives
example
• Non S.M.A.R.T objective
– To identify risk factors for HCV infection
Principal objective
• To determine if sharing a haemodialysis machine with
a HCV infected patient
is a risk factor for HCV infection
Secondary objective:
• To identify failures in procedures
designed to prevent cross-infection
via haemodialysis machines
11
Hypothesis
“The incidence of HCV infection in
haemodialysis patients is higher in patients
sharing machines with HCV infected patients
than in patients not sharing machines with
an HCV infected patients”
12
Protocol outline
1. Presentation
2. Background and justifications
3. Objectives
4. Methods
5. Ethical considerations
6. Project management
7. Timetable
8. Resources
9. References
10. Appendices
13
4. Methods
• Procedures to achieve objectives
– what will be done?
– how?
• Information used to judge validity
14
4. Methods
• Study design
 cohort, case control, cross-sectional…
 brief justification
• Study population





definition
selection
criteria for inclusion and exclusion
mechanisms of recruitment
accessibility, follow-up, representativeness
15
4. Methods
• Sampling design (ref: lecture sampling)
 frame: district, household, persons,…
 method: random, cluster, stratified,…
 randomisation procedures
 replacement procedures (in case of refusal)
• Sample size (ref: lecture sample calculation)
 sample size, power calculations
based on principal objective
 feasibility
16
Sample size
e.g. Options for case-control study on sporadic campylobacter
and food exposure
Prevalence
of exposure
among
controls
Case:
control
ratio
Alpha
error/
power
Odds
ratio to
detect
Number
of
cases
Number
of
controls
10%
1:2
5%/ 80%
2
223
446
10%
1:2
5%/ 80%
3
81
162
10%
1:2
5%/ 80%
4
48
96
10%
1:2
5%/ 80%
5
34
68
10%
1:2
5%/ 80%
6
27
54
17
4. Methods
• Selection and definition
 exposures:
risk factors, protective factors, confounding factors
 outcomes:
definition of case, of control group
• Items to be measured
– scales used
• e.g: legionella <=> car driving
- Driving exposure: definition, quantification, categories
- Legionella infection: case definition, control group definition
18
CC study of sporadic cases
of Salmonella Enteritidis infections
• Exposure
– consumption of custard
• Case
– a person living in South-West Wales with
a laboratory confirmed infection due to S. Enteritidis
in June and July 2011
• Case finding
– through Public Health Laboratory; weekly notifications
• Control
– persons living in SW Wales
in same neighborhood as cases
• Control finding
– random selection of people using telephone directory
19
Methods
Data analysis plan
• Structured in terms of objectives
• Hypotheses tested, dummy tables
 Comparison of groups
 risk factors
 protective measures
 Assessment of dose-response relationship
for key exposures
 Assessment of possible confounding factors / effect
modifiers
• Statistical tests used, adjustment, stratification
20
Methods
Data analysis plan
• Define
– indicators you will need to reach objectives
– data you will need to collect
• Better estimates of sample size for analysis
of sub-groups
21
Methods
Data analysis
Dummy table:
Food specific attack rates of Salmonella infection
in a day care centre, Paris, May 2009
ill
n
total
n
AR
%
RR
95%CI
ice-cream yes
ice-cream no
fruit cake yes
fruit cake no
pudding
pudding
yes
no
22
4. Methods
Data collection
• How
 interview, observation, record review
• By whom
 interviewers: selection, training
 level of supervision
• Tools (ref: lecture “questionnaire design”)
 questionnaires, recording materials
 questionnaires: self or interviewer administered,
face-to-face or telephone interview
• Blind data collection
• Procedures for taking samples
23
4. Methods
Data handling
• Coding (anonymisation)
 during data collection, afterwards?
 by whom?
• Processing
 software, hardware
 entry
• during the study, afterwards?
• single entry, double entry?
• Validation and data cleaning
24
4. Methods
Pilot studies, pre-testing
• No study without pre-test
 Feasibility of sampling
 Data collection, measurement methods
 Questionnaire
• Describe how to test
25
4. Methods
Limitations
• Identification of potential sources of biases




selection bias
information bias
misclassification bias
interviewer bias
• How to deal with them
 possibilities for correcting
 how they will affect the results
26
Protocol outline
1. Presentation
2. Background and justifications
3. Objectives
4. Methods
5. Ethical considerations
6. Project management
7. Timetable
8. Resources
9. References
10. Appendices
27
5. Ethical considerations
• Informed consent
– translated in local lay language
• Confidentiality
– coding data collection instruments
without identifier
• Data storage and protection
• Ethics committee
28
6. Project management
• Participating institutes and persons
• Responsibilities and tasks of each partner
• Data ownership
• (Authorship)
29
7. Timetable
Planning/organisation of the study
• questionnaire design, recruitment, purchases
• permission
• obtain funding
Pilot study
• time to do adjustments
Final study
• data collection
• analysis
• presentation of results and write up
30
8. Resources
• Extent of this section depends
on target audience
• Specify
 available sources
 requested sources
• Keep budget
 reasonable
 detailed
 well justified
31
9. References
• Limit number of references to key articles
• Follow recommended style
• Vancouver
www.library.soton.ac.uk/infoskills/vancouver.shtml
www.transfusion.ca/new/bulletin/vancouver-style.html
32
10. Appendices
• Methodological appendices
• List of definitions
• Questionnaires
• Introductory letters to study participants
• Informed consent forms
…..
33
Common problems
• Too ambitious: too many questions
• Insufficient attention to literature
• Poor justification
 why is it important to answer this question?
 what impact does it have on public health?
• Poorly formulated objectives
• Inappropriate analysis
• Inadequate description
• Absence of pilot
34
Your task next week
• Create a study protocol in small groups
• Present it to the plenary
• Read the references during the weekend