How to read

1
How to read

• Mark Gabbay

2
Reader System, MacAuley

• R- Relevance
• E- Education
• A- Applicability
• D- Discrimination (quality)
• E- Evaluation (is it robust evidence)
• R- Reaction (now what?)

3
Relevance

• Primary Care, possibly specialist related to
  substance misuse
• Clues
• Journal
• Title
• Author affiliation

4
Education

• Concentrate on what will alter practice
• Relates to work undertaken
• Challenges what we already do
• If not, stick to abstract/summary
• Applicability
• Relevant to your setting

5
Discrimination

• Academic source?
• More likely to have internal peer review
• Journal?
• high quality reputation and peer review system
• Statistician author?
• Well referenced?
• Be wary of unsupported assertions

6
Evaluation

• Hierarchy of Evidence
• Systematic review/meta-analysis
• Well conducted Randomised Controlled Trial
• Trial
• Cohort study
• Before and after or case control
• Case reports, expert/consensus view

7
Reaction

• Circulate to all staff and discuss ASAP
• Summarise and disseminate
• File for easy access
• Note ref and brief summary, index card or
  computer
• Bin

8
Critical Appraisal

• 2 stages
• Is the information supplied?
• No- Some- Detailed
• Judgements about whats reported
• Appropriate but not ideal
• Gold standard
• Hmmmmm

9
Primary or Secondary research

• Primary- derived from original data
• Secondary- collations of appraised original
  research
• Cochrane collaboration
• Evidence based reviews-York
• Bandolier- Oxford
• SMMGP

10
Choice of criteria

• Depends on methodology used
• Review/meta analysis
• Trial
• Cohort
• Case control etc
• Depends on whats being explored
• Intervention
• Prognosis
• Epidemiology
• Economic- cost effectiveness
• Qualitative research

11
Whats the point?

• Develop personal approach and system
• Formal appraisal
• Informal checklist to use when reading
• Confidence in accessing and understanding
  research, and sorting out how good evidence is
• Defence against inadequately referenced
  guidelines and protocols etc

12
Trials Criteria

• Population studied
• Sample size
• Power calculation done? Target attained?
• Inclusion/Exclusion criteria
• How representative was sample?
• How generalisable to your population?
• Pragmatic v Explanatory design
• Setting Specialised or Primary Care

13
Allocation of treatments

• Sources of Bias recognised and accounted for?
• Randomisation described
• Appropriate, free from fiddling
• Blinding
• Differences between groups
• Described
• Balanced, stratified for key variables?

14
Outcome measures

• Gold Standard?
• Comparable with other research?
• Validity and reliability?
• Better alternatives?
• Clinically relevant/important?
• Relevant to your practice/population?

15
Overall approach

• Clear aims and objectives given
• Hypotheses stated
• Beware fishing trips
• Beware subsample reports that were not in
  original design and power calculation

16
Results

• Fully reported- any obvious gaps?
• Data provided to support findings?
• Statistics- obvious tests not reported in favour
  of obscure ones?
• Fishing trip?
• Lots of tests need stricter limits
• P value and CIs

17
Discussion

• Contextualised with other research
• Problems acknowledged, design decisions and
  consequences discussed
• Generalisability/limitations of findings
• NNT, costs etc

18
Reviews

• Searches comprehensive
• Suitable electronic databases
• All relevant included
• Adequate time cover
• Appropriate search terms
• Secondary referencing
• Hand searches
• Personal contact
• Publication bias and unpublished data

19
Review 2Critical appraisal used

• Appropriate to research type
• Results given
• Summaries of papers and scores
• Inter-rater reliability, approach when
  disagreement
• Meta-analyses explained if used
• Complex stats, so read up if interested

20
Other research

• Health economics
• Data collection and analysis
• Qualitative
• Sampling
• Analysis
• Method used
• Triangulation
• Context and relevance

21
Sources of Information

• Books
• Obviously The Evidence-Based Primary Care
  Handbook, Mark Gabbay Ed RSM press 99
• Ridsdale or Greenhalgh (also BMJ series 95/6)
• Sackett et al Evidence based medicine
• Websites
• University of Oxford Centre for Evidence Based
  Medicine
• University of Sheffield ScHARR netting the
  evidence guide

22
This years papers

• One year follow-up of opiate injectors treated
  with oral methadone in a GP centred programme.
  Hutchinson S, Taylor A, Gruer L, et al.
  Addiction 2000 95 (7) 1055-68.
• Overdose deaths attributed to methadone and
  heroin in New York City, 1990-1998. Bryant WK,
  Galea S, Tracy M et al. Addiction 2004 99
  846-854.
• Methadone treatment practices and outcomes for
  opiate addicts treated in drug clinics in general
  practice results from the capitals National
  Treatment Outcome Research Study. Glossop M,
  Marsden J, Stewart D et al. British Jounral of
  General Practice 1999, 49 31-4
• Impact of a positive hepatitis C diagnosis on
  homeless injecting drug users a qualitative
  study Charlotte NE Tompkins, Nat MJ Wright and
  Lesley Jones British Journal of General Practice
  2005 55 263268
• Buprenorphine versus methadone maintenance
  therpay a randomised double-blind trail with 405
  opioid-dependent patients. Mattick RP, Ali R,
  White JM at al. Addicition 2004 98 441-452
• Papers 3-7 in the Course Reader
• Listed on page 28 in the Course Handbook

23
The Marking Grid Page 14 in Tutor Handbook