2. Clinical Research Design
• The best design to answer the research question
• Feasibility
• Interest
• Relevant
• Novelty
• Ethical
3. • The best design to answer the
research question
4. Common misconceptions
• Some designs are better than other
• Not true
• Need to select appropriate designs to match problems
• Some methods are better than others
• Not true
• Need to select appropriate methods to match problems
• Topical issues are better or more easily published
• Not true
• Editors and examiners want to see good research
5. Common problems
• Lack of focus
• Decide what your research question is and stay with it
• Lack of a clear research question
• Don’t proceed until you have one
• Going too quickly
• Take your time in the early stages of a project
• You must try to avoid changing topic of changing designs
7. Developing a research question
A clear statement in the form of a question of what you set out to
investigate, eg:
• Does intervention A work better than intervention B?
• What are the most common long-term complications of contracting COVID-
19?
• What is the patient experience of attending a nurse-led care clinic?
• What are the most effective treatments for COVID-19?
9. Research methods
• Quantitative
• Randomised controlled trials
• Define primary outcome measure and data collection method
• How will randomisation be achieved?
• How will blinding be assured?
• How will sample size be calculated?
10. Research methods (contd.)
• Quantitative (contd.)
• Surveys
• Questionnaires
• Postal
• Telephone
• Internet
• Observations
• Sample size considerations
• Sampling procedures
14. Types of longitudinal studies
Trend
• same population (eg population of Hong Kong)
• not same people
Cohort
• same sample (eg nursing class of 2010)
• not necessarily the same people
Panel
• same sample (eg nursing class of 2010)
• same people
15. Types of longitudinal studies
Trend
• very informative about trends
• not very informative about people
Cohort
• quite informative about trends
• quite informative about people
Panel
• least informative about trends
• most informative about people
16. Types of longitudinal studies
Trend
• relatively easy to conduct
• sampling is easy
Cohort
• harder to conduct
• quite vulnerable to attrition
Panel
• Very hard to conduct
• Very vulnerable to attrition
27. Mixed methods (quantitative & qualitative)
• All the problems of survey and qualitative designs
• Which mixed methods design to use
28. Systematic review (& meta-analysis)
• More time consuming than anticipated – always
• Possibility of a nil return
• Retrieving articles
• Evaluating quality of articles
30. Common mistakes
• Wanting to change the world
• You are unlikely to
• Your project will be too ambitious
• Setting out to ‘prove’ something
• This is philosophically flawed
• And research ‘proves’ nothing
• Deciding on your method first
• You need to be clear on a research question
• Your methods should then follow
31. Issues affecting feasibility
• Will you have time?
• How long will data collection take?
• How long will it take to obtain permission?
• Will you obtain permission?
• Is the clinical area committed to the work?
• Do you envisage ethical problems?
• Can you recruit participants?
• What is the incidence/prevalence of the issue?
• Can potential participants give consent?
33. What do you want to get out of a research
project?
• Personal
• Achievement
• Ambition
• Altruistic
• Improve lives
• Professional
• Status
• Recognition
• Academic
• MSc/PhD
35. Is anyone interested in your proposal?
• Are you interested?
• If not, then you will be less committed to the work
• Can you enthuse others about the work (you may need their help!)?
• Is the issue of interest to your hospital?
• It should make life easier if it is
• But beware of ‘stepping on toes’
• Is the issue of national importance?
• This will also make life easier
• There is the potential for financial support
• You should check to see if there are relevant national priorities
37. Is the work novel?
• It does not have to be
• Replication is a vital aspect of clinical research
• Replication provides you with a ‘ready made’ design
• But you will have to justify doing a replication study
• If the work is novel
• You need to be sure it is before you claim it is
• You need to explain why it is necessary
39. Is the work ethical?
• Can you envisage any ethical problems?
• Try to anticipate ethical issues
• What are the local ethics procedures?
• How long do they take?
• Expect at least one revision
51. Conclusion
• There is no best design
• Have a clear research question
• This will dictate
• Design
• Methods
• Analysis
• Ethics
• Every methods has its drawbacks
• You have to be interested in the project
• Others need to be interested to provide support
• Anticipate ethical issues