Where's your evidence reading an academic paper

1
Where's your evidence!- reading an academic paper

• Mark Gabbay

2
Traditional Knowledge Acquisition

• Written
• Textbooks
• Sparsely referenced at best, out of date
• Articles
• Traditional reviews reference serve to underpin
  authors prejudice
• Research reviews or reports
• Verbal
• opinion leaders, teachers, colleagues
• Experience
• Mindlines (Gabbay Le May BMJ 3291013-7)

3
Evidence-Based Health Approach

• Seeks to gather knowledge to inform choice and
  decision options
• Ideally then shared in consultations introduced
  by either party
• Facilitates sifting of knowledge to judge
• relative quality
• applicability to individuals/strategic planning

4
Types of Evidence

• Quantitative
• Qualitative
• Primary
• Secondary
• Anecdotal
• Systematic
• Biased, Opinionated, Conflicicting interest
  motivation etc

5
Quantitative research

• Traditional
• Experimental- laboratory, clinical
• Trials RCTs, non randomised, n of 1
• Epidemiological
• Case control/cohort
• Surveys
• Need to understand alternative methods, relative
  strengths and limitations, and suitable
  settings/circumstances

6
Qualitative

• Different approaches
• Interviews/focus groups
• Ethnography
• Audio/Video
• Mixed methods, data-rich and thick descriptions
  with thematic analysis
• Measures of rigour exist and should be
  considered data collection, analysis,
  presentation, triangulation etc.

7
Hierarchy of Evidence

• a. Systematic reviews/ meta-analyses
• b. Experimental designs
• c. RCTs 
• a. Well-designed, non-randomised trials
• b. Cohort control studies
• c. Case-control studies
• Consensus conference,Expert opinion Observational
  study
• Other types of study eg. Interview based, local
  audit, Quasi-experimental, qualitative design 
• 4 Personal communication    

8
Sources of Information

• Primary- searches of journal databases
• Medline, Psychinfo, EMBASE, CINAHL etc
• Secondary
• Cochrane
• Clinical Evidence
• Bandolier, Effective Healthcare bulletins etc
• SCHARR netting the evidence
• National Electronic Library for Health

9
Drug Misuse Evidence

• Cochrane Collaboration
• Systematic reviews
• NTA evidence summaries
• Original Research
• Find it, appraise it
• Homework for next masterclass

10
Consider design context

• Population
• Inclusions/exclusions- relevant to your
  population? Clearly defined reasonable?
• Sample size calculation?
• Setting relevant?
• Defined Intervention?
• Congruence- professional and client
• Monitoring systems? Dropouts?
• What outcome should be measured?
• Clinically relevant/important
• Measure reliable/valid?

11
Urgghhh STATS!

• Means to compare samples of populations
• Measure of potential chance rather than real
  differences
• Types of Data
• Nominal
• Ordinal or Ranked
• Numerical- discrete or continuous
• Normal Distribution or not
• parametric v non-parametric