Putting Together a Research Proposal

1
Putting Together a Research Proposal

• Tate Erlinger, MD, MPH
• Director of Clinical Research
• Office of Research Administration
• Seton Family of Hospitals

2
Part I Starting Out
3
Establish the Rationale

• A rationale for the choice of topic,
• Why is it important or useful within the concerns
  of the discipline in which you are writing?
•  
• It is also sensible to indicate the limitations
  of your aims.  In other words, don't promise what
  you can't possibly deliver.

4
Know the Literature in the Field

• A review of existing published work ("the
  literature") that relates to a topic. 
• Tell how your proposed work will build on
  existing studies and yet explore new territory.

5
Outline your Approach

• An outline of your intended approach or
  methodology (with comparisons to existing
  published work), perhaps including costs,
  resources needed, and a timeline of when you hope
  to get things done

6
Part II Setting up a Specific Project
7
Where does clinical research start?

• Location
• Location
• Location
• Find a rich environment and a good mentor
• Take advantage of opportunities

8
A Research Project
QUESTION
9
Steps to Starting a Project?

• Why do you want to do a project?
• What is your question?
• Who should you ask for help?
• How might you approach the problem?
• Where is the study taking place?
• What is your time frame? (when?)

10
Why do you want to do a project?

• Interest in a topic
• Bolster your resume
• Exploring career paths
• Mandatory
• Other

11
Assemble a group

• Clinical Research is a group processnot an
  individual endeavor
• Identify a mentor/advisor
• Work together to build a team as your project
  progresses

12
What is the question?

• The MOST important part of the process
• How do you choose a question?
• How do you phrase the question?
• How does the question change as you start to look
  at what is feasible vs. what is optimal?

13
Choosing a Research Question

• Think of an area you are interested in and start
  READING
• Reviews
• Meta-analyses
• Editorials
• What are the conflicts in the field?
• What do experts say are the Gaps in knowledge?

14
Drafting a Question

• The aim of this study is to..
• Determine the association between X and Y
• Determine the impact of intervention X on
  outcomes 1, 2 and 3.
• Etc.
• Be very specific
• Clearly articulate all the outcomes,
  interventions or variables of interest

15
After the Question

• Go back to the literature
• Good recent reviews
• Good recent meta-analyses
• Create a data table

16
Revisit the Question

• After reviewing the literature more thoroughly
  has your question changed or been modified?
• Nokeep going
• Yesfine, collect a little more information

17
How, When, Where

• Once the question is posed, think about the How,
  When, Where
• In reality, youve been considering this all along

18
Summary

• The question is the hardest and most important
  thing to define.
• Be specific
• Dont be afraid to refine
• Assemble a team, starting with an advisor
• Take the time to learn your field and gather data
• Have fun!

19
Setting up

• Decide on study design
• Cross-sectional (single point or period in time)
• Descriptive
• Exploratory associations
• Case-control
• Retrospective
• prospective
• Incident vs. prevalent cases
• Selection of cases / controls
• Prospective
• Random sample
• Convenience sample
• Duration of follow-up
• Measurement of exposures (multiple) / outcomes
  (multiple)

20
How large should the study be?

• Sample size / power calculations
• Underpowered studies are only useful for
  feasibility
• ALL studies should have an estimate of
  statistical power
• Use data from prior studies or pilot studies
• Various Methods used to calculate power or sample
  sizedepends on study design and analytic
  techniques being used.

21
Collecting Data

• How will data be collected?
• Paper form (standardized)
• Spreadsheet
• Other?
• Who will collect the data?
• When will data be collected?
• What type of data will you collect?

22
What type of data?

• EXAMPLE chart review
• Create data collection tool
• Avoid free text difficult to analyze
• Multiple choice (or pre-defined choice)
  preferable
• Pilot the data collection tool first
• What were the problems?
• What was missed?
• Was there something that needs further
  clarification?
• Did you capture the data you were really after
  (valid)?

23
Data Storage

• Must transfer data from paper to electronic if
  not already done.
• Avoid multiple data manipulations if possible,
  e.g. transfersbetter to start with electronic.
• Keep tight version control of data sets and
  ownership of data.
• Keep data secure, locked, on protected computer
  and/or locked drawer

24
Data Analysis

• Seek help if needed.
• Cleaned data set will be easiest to work with
• Analyze data according to Pre-defined plan
• DO NOT go fishing
• Multiple comparisons problems
• Spurious results

25
Sections of the Proposal

• Title
• Investigators
• Abstract
• Specific Aims
• Background and Significance
• Experimental Design and Methods
• Preliminary Results (if applicable)

26
Title

• Type the entire protocol title. This title
  should match the one on the Request for Expedited
  Review and Request for Full Board Review Forms.

27
Investigators

• List the investigators as Principal Investigator,
  Co-Investigator, Collaborators, Coordinators,
  etc.

28
Abstract

• Include the hypothesis being tested, summary of
  the study aims and procedures, experimental
  design, and the significance of the research in
  laypersons language.

29
Specific Aims

• Clearly state the hypotheses to be tested and the
  objectives or specific aims. Numbers and types
  of subjects and types of procedures should be
  specifically noted

30
Background and Significance

• Concise
• Demonstrate a knowledge of the field
• Cite relevant papers in the field

31
Background and Significance

• Provide details of clinical study design.
• Include an in-depth narrative of the methodology
  to be employed.
• Provide details of Statistical Considerations.
• This section must include a justification of the
  sample size and a statement regarding power based
  on one or more of the primary outcome measures,
  presumably ones for which there is some
  information about intra-individual variation

32
Preliminary Results

• Be concise.
• If there are none, please state.

33
References

• NIH guidelines for preparing a proposal at
  http//www.niaid.nih.gov/ncn/grants/write/index.ht
  m