MHE3 Reviewing the Medical Literature

1
MHE3 Reviewing the Medical Literature

• Research Methods Lecture 2
• A McCaw-Binns
• Section of Community Health
• UWI-Mona

2
Learning Objectives

• Discuss reasons for reviewing the literature
• Discuss practical aspects of the process
• Databases on CD-ROM
• Source material in printed form
• How to access and search for material

3
Why Review the Published Literature?

• Prevents duplicating work already done
• Determine what others have learned or reported on
  the problem
• Familiarize oneself with methods that might be
  used
• Justifies why the research needs to be done

4
What is needed

• Review of past research with a bias towards those
  aspects of the problem of direct relevance to
  your study
• In this way the hypothesis to be tested can be
  derived in a logical manner
• General statements about the field of research
• Narrow and specific set of statements which lead
  to the development of the aims or hypotheses
• Information gaps/areas for further research

5
Reading the Literature

• Process
• Consult search engines
• Review abstracts
• Select papers of interest and read in more detail
• Critically review important papers
• Summarize in coherent and logical way

6
Sources of information

• Original articles in journals
• Summaries from the medical literature
• Review articles
• Practice guidelines
• Expert opinion
• Consensus statements
• Editorials
• Summary articles in journals

7
Sources of information, contd.

• The internet
• Textbooks
• Consultation with colleagues who are expert
• Attend meetings, lectures, seminars
• Advertisements in journals
• Conversations with representatives from
  pharmaceutical/medical technology companies

8
Index Medicus/Medline

• Largest database over 12 million records
• References and abstracts from 4000 health
  related journals 1966 -- present
• Medicine, nursing and veterinary literature
• Allows sophisticated searching by
  cross-referencing different headings

9
Med-Caribbean

• UWI Medical Library sub data base on Caribbean
  research. Includes
• Listings not in peer reviewed journals (grey
  literature)
• Articles by Caribbean scholars published in
  journals to which the Medical Library does not
  subscribe
• Regional information
• You are invited to contribute

10
Systematic reviews of evidence

• Cochrane Database of Systematic Reviews
• Oxford Research group Cochrane Collaboration
• Only reviews high quality RCTs
• Highest grade evidence availability
• On CD-ROM combined with DARE
• Good bibliography of evidence-based methods

11
DARE Database of Abstracts of Review of
Effectiveness

• NHS Centre for Reviews and Dissemination
• Reviews identified from the published literature
• Some meet Cochrane criteria for classification as
  systematic
• Others not screened and need critical appraisal
  by the user
• Small number of records but growing

12
ASSIA Applied Social Science Index and Abstracts

• Includes social sciences not restricted to
  health
• Covers sociological aspects of health care
• Health promotion
• Socioeconomic status
• Community care
• Patients rights

13
Other databases

• CINAHL Cumulated Index to Nursing and Allied
  Health Literature
• CANCERLIT cancer journals, books, reports
• AIDSLINE HIV/AIDS information
• PSYCHLIT psychiatry/psychology literature
• AMED complementary medicine

14
Internet

• Access to reports, guidelines, discussion lists,
  home pages etc.
• Quality of material very variable (good ? odd)
• Some reviewed by experts, many not
• Access to information haphazard
• Difficult to assess information on the internet
• ?bona fide expert
• ?views/opinions of an individual

15
Using Medline Become familiar with subject
headings

• Biased towards USA words/phrases
• e.g. Prenatal care vs ANC family physician vs GP
• All possible spellings of a word need to be used
• e.g. anemia and anaemia
• Use multiple terms for same condition
• e.g. bed sore/pressure sore/decubitus ulcer
• Sometimes asking for the opposite helps
• e.g patient non-compliance vs patient compliance

16
Tree structure

• Medline gives most precise term it can to a
  subject
• Terms arranged in a tree structure
• Most specific rule crucial to the correct use
  of the tree
• e.g. Cannot retrieve article on the brain by
  searching for central nervous system
• If search for NSAIDS, will not retrieve all
  articles on aspirin, only general articles on
  NSAIDS. Need to search for aspirin

17
Medline options

• EXPLODE
• All sub-groups within a subject area will be
  included. If you EXPLODE NSAIDS, you broaden
  search to include ignored sub-groups.
• NSAIDS 4554 hits
• Aspirin 3223 hits
• NSAIDS exploded 13877 hits
• OR
• Broadens search to include more than one topic
  that could cover the subject
• e.g. Family practice OR primary health care

18
Medline options

• AND
• If search produces too many results, the AND
  option will narrow search, retrieving fewer
  results and being more specific
• e.g. aspirin AND myocardial infarction
• LIMIT
• Narrows results further
• e.g. English language only, specific age groups,
  human only
• Major headings only gives fewer results but
  increases relevance of articles to area of
  interest

19
Too many results?

• Reduce by limiting search to
• English only
• By year(s)
• Reviews only
• Level of emphasis (e.g. major headings only)
• Use more precise terms
• Link to other topics using AND facility

20
Too few results?

• Widen search by using alternative sources
• Search more years
• Broaden search using OR option
• Broaden search using EXPLODE option

21
How to retrieve results?

• Headings only
• Abstracts only
• Abstracts with author affiliation
• Printed
• Kept on a file

22
Request List of Abstracts

• Abstracts
• Summaries of papers which outline
• Author
• Title
• Journal published
• Background and methods
• Key findings
• Review abstracts and select papers to read in full

23
Review articles

• Summaries of current medical evidence on a topic
• Systematic reviews provide reliable answers to
  important questions
• Less subjective/objective conclusion
• Clear guidelines needed in the future to improve
  their quality

24
Standard process for listing articles

• Journal articles listed by author, title,
  journal, year, volume , page number, e.g.
• Graham W, Wagaarachchi P, Penney G, McCaw-Binns
  A, Antwi K Y, Hall M and the CBCA Study Group.
  Criteria for clinical audit of the quality of
  hospital-based obstetric care in developing
  countries. Bulletin WHO 2000 78 (5) 614-620.
• McCaw-Binns A, Moody CO. The development of
  primary health care in Jamaica. West Indian Med
  J 2001 50 Suppl 4 6-10.

25
Information of Interest

• Previous research with direct relevance to
  research topic
• Distribution of the problem
• Who is affected, when and where
• Size and intensity of the problem
• How widespread, severe
• Consequences
• disability, death, waste of resources

26
Information of Interest

• Factors which contribute to or are associated
  with the problem
• If planning an intervention, describe possible
  solutions that
• may have been tried
• what worked, what did not and how or why
• General agreements/disagreements among previous
  researchers

27
Summarize important articles

• NOTE
• Author(s), title, source (journal, book etc)
• Summary of contents relevant to your study
• Methodology (comment if necessary)
• Key findings
• Key weaknesses, data gaps
• Areas for further research
• How you may use study information in your
  research
• Scan references for leads to other information
  sources

28
Critical review of the literature

• STUDY DESIGN
• Hypothesis
• Study population
• Inclusion/exclusion criteria
• Sample selection
• Sample size
• How many individuals are in the study and control
  groups
• Are there sufficient persons to answer the
  research question

29
Writing a Literature Review

• Write a 1-2 page coherent discussion, using the
  relevant references for each category of
  information
• Article may be cited gt once if it addresses more
  than one of your areas of interest /need
• e.g. Disease prevalence, risk factors,
  methodology
• Avoid redundancy
• gt1 article may be cited to support a point
  without repetition

30
Bias

• Distortion of information so that it reflects
  opinions or conclusions that do not represent the
  real situation
• If you have reservations about selected
  references, of if you find conflicting opinions
  in the literature, discuss these openly and
  critically

31
Biases to be aware of

• Playing down controversies and differences in
  ones own study results
• Restricting references to those that support the
  view of the author
• Drawing far reaching conclusion from preliminary
  data
• Making sweeping generalizations from one case or
  a small study
• LEAPS OF FAITH!

32
Ethical considerations

• Careless presentation and interpretation of data
  may mislead readers, even lead to wrong decisions
  affecting peoples health
• If researchers present findings without citing
  the source, can be charged with plagiarism
• Appropriate referencing should occur when you
  prepare proposals and write research reports

33
Ethical issues Avoid plagiarism

• Theft of anothers work or ideas
• You cannot use another writers exact words
  without quotation marks and a complete citation
• You cannot edit/paraphrase another persons work
  and present the revised version as yours
• You cannot present another persons ideas as your
  own, even if you use totally different words to
  press those ideas

34
Summary

• Review the literature for information, ideas,
  identification of information gaps, areas for
  further research
• Summarize past work with a critical eye to study
  design, implementation, analysis, interpretation
• Document previous work adequately and avoid
  plagiarism and bias