Supporting syntheses of the literature

1
Supporting syntheses of the literature

• SHINE
• Quality The Whole Picture
• 26th November 2004

2
Content

• Definition and purpose
• Types of syntheses
• The role of the information specialist
• Identifying previous reviews, scoping searches,
  retrieval of primary studies
• Challenges of the role

3
Definition and purpose

• Definition
• reduce large quantities of primary information
  into palatable pieces for digestion by decision
  makers (Mulrow, 1995)
• produce reliable results on which to base
  decisions by using robust processes to locate,
  appraise and synthesise research evidence (Song
  et al., 2000)
• Purpose
• Quality improvement

4
Systematic reviews

• a review that has been prepared using some
  kind of systematic approach to minimising biases
  and random errors, and that the components of the
  approach will be documented in the materials and
  methods section. (Chalmers Altman, 1995)
• Organisations undertaking systematic reviews
• Cochrane Collaboration
• Campbell Collaboration
• Centre for Reviews and Dissemination

5
Health Technology Assessments

• a multidisciplinary field of policy analysis
  that studies the medical, social, ethical, and
  economic implications of the development,
  diffusion and use of a health technology
  (INAHTA, 2000)
• Organisations undertaking HTAs
• NHS QIS
• NICE
• INAHTA

6
Guidelines

• Clinical guidelines are systematically developed
  statements to assist practitioner and patient
  decisions about appropriate health care for
  specific clinical circumstances.
  (www.sign.ac.uk)
• Organisations preparing evidence-based guidelines
• SIGN
• NICE
• GIN

7
One topic three syntheses

• Systematic review
• In alcohol dependence, is acamprosate or
  naltrexone more effective for preventing relapse?
• HTA
• Prevention of relapse in alcohol dependence
• Guideline
• The management of harmful drinking and alcohol
  dependence in primary care

8
The role of the information specialist

• To comprehensively identify studies to answer the
  review question(s)
• Stages
• Identifying previous reviews
• Undertaking scoping searches
• Retrieving primary studies

9
Identifying previous reviews

• Preliminary evaluation of the evidence base
• Has this question already been answered?
• If not, can an answerable question be formulated?
• Is there a reasonable evidence base?
• CDSR, DARE, HTA database, Guidelines Finder,
  National Guideline Clearing House
• Ongoing reviews
• Bibliographic databases, journals and websites

10
Scoping searches

• Scoping searches establish the feasibility of a
  review question and investigate how searching and
  appraisal can be operationalised. (Booth
  Fry-Smith, 2003)
• establish the volume and quality of the primary
  literature
• aid development of the review protocol
• indicate how much the review might cost

11
Retrieval of primary studies

• Formulating clear search questions
• Recognising the best evidence to answer the
  question(s)
• Identifying the most likely sources to search
• Searching information sources effectively
• (Snowball, 2002)

12
Formulating clear search questions

• The product
• The well-built clinical question (Richardson et
  al., 1995)
• The well-formulated question (Counsell, 1997)
• The anatomy of a question (Sackett et al.,
  1997)
• The process
• Focusing the question (Oxman et al., 1993)
• Formulating the question (Counsell, 1997)
• Formulating the problem (Cochrane Handbook)
• Slide reproduced with kind permission from Andrew
  Booth

13
Formulating clear search questions

• Ask a poor question and you will get a poor
  review. A clear question also helps the reader
  rapidly assess whether the review is relevant to
  his or her own clinical practice (Counsell,
  1997).
• Clearly framed questions guide much of the
  review process including strategies for locating
  and selecting studies or data, for critically
  appraising their relevance and validity, and for
  analysing variation among their results
  (Cochrane Handbook).
• Slide reproduced with kind permission from Andrew
  Booth

14
Formulating clear search questions

• POPULATION/PATIENT who?
• INTERVENTION/EXPOSURE what?
• COMPARISON compared with?
• OUTCOME with what result?
• PICO, PIOC or PECO
• Slide reproduced with kind permission from Andrew
  Booth

15
Worked example 1

• Population/Patients are persons with alcohol
  dependence
• Intervention who are treated with acamprosate
• Comparison compared with those treated with
  naltrexone
• Outcome(s) less likely to relapse (into heavy
  drinking and dependence)?

16
Worked example 2

• Population/Patients should pregnant women be
  offered
• Intervention first trimester ultrasound scanning
  (nuchal translucency measurement)
• Comparison or second trimester serum screening
• Outcome(s) accuracy of risk assessment for
  trisomy 21 (Downs Syndrome)?

17
Translating PICO into a search strategy
POPULATION Alcoholics OR Alcoholism And INTERVENTION Acamprosate OR Campral And COMPARISON Naltrexone OR Trexan And OUTCOME Temperance OR Sobriety
Slide reproduced with kind permission from Andrew
Booth
18
Is PICO effective (1)?

• Statistically significant correlation between use
  of PICO-structured form and complexity of search
  strategy (P 0.002) and between clinical
  requests handled by PICO-structured form and
  fewer items retrieved (P 0.028). However,
  librarians preferred minimally structured forms
  to PICO-structured forms in every dimension
  except informativeness.
• (Booth et al., 2000)
• Slide reproduced with kind permission from Andrew
  Booth

19
Recognising the best evidence to answer the
question

• Therapy
• Diagnosis, screening
• Aetiology
• Prognosis, natural history
• Cost-effectiveness
• Qualitative
• (McKibbon, 1999)

20
Recognising the best evidence to answer the
question

• T In alcohol dependence, is acamprosate or
  naltrexone more effective for preventing relapse?
• D What is the accuracy of screening for alcohol
  problems in primary care?
• A What are the environmental risk factors
  associated with binge drinking in college
  students?
• P What is the long-term prognosis of patients
  with alcoholic cardiomyopathy following total
  abstinence?
• CE Are brief interventions cost-effective for
  problem drinkers?
• Q What are adolescents' perceptions of parental
  alcoholism?

21
Identifying the sources to search (1)

• Subject
• General, eg MEDLINE, EMBASE
• Subject-specific, eg AMED, CINAHL, BNI
• Type of question/type of evidence
• Therapy, eg Cochrane Database of Systematic
  Reviews (CDSR) Cochrane Central Register of
  Controlled Trials (CENTRAL)
• Cost-effectiveness, eg NHS Economic Evaluation
  Database (NHS EED) Health Economic Evaluations
  Database (HEED)

22
Identifying the sources to search (2)

• Databases
• MEDLINE, EMBASE, CINAHL, PsycINFO, AMED, HMIC
• Searchable e-journal packages
• Science Direct, Ingenta, Journals_at_Ovid
• Registers of Research
• mRCT, National Research Register, cancer.gov
• Conference Proceedings
• Index to Conference Proceedings, ISI Proceedings
• Dissertations
• Digital Dissertations, Index to Theses
• Interested parties
• Professional associations, voluntary groups
• Internet .

23
Searching information sources effectively

• Translating the parts of the question (PICO) into
  search terms
• Combining search terms using Boolean logic
• Sensitivity vs precision
• Sensitivity maximising
• Synonyms
• Subject headings and free-text
• Truncation and wildcards (, ?, )
• Proximity operators (adj, near, same)
• Search filters, quality filters

24
A search strategy

1. Alcoholism/
2. Alcohol drinking/
3. alcoholic?
4. alcoholism.tw
5. dipsomania.tw
6. or/1-5
7. Taurine/
8. acamprosate.tw
9. campral.tw.
10. or/7-9

1. Naltrexone/
2. naltrexone.tw
3. trexan.tw
4. or/11-13
5. Temperance/
6. temperance.tw
7. sobriety.tw
8. ((control or reduc or restrict) adj2
   drink).tw
9. or/15-18
10. 6 and 10 and 14 and 19

P
25
Challenges of the role

• Publication bias
• Time vs quality trade-off
• Complexity of review questions
• Widening perspectives

26
References
Alderson P, Green S, Higgins JPT eds. Cochrane
Reviewers handbook 4.2.2. updated March 2004.
Available from ltwww.cochrane.org/resources/handboo
k/hbook.htmgt Accessed 11 October 2004. Booth A
and Fry-Smith A. 2003. Developing the research
question. In Etext on Health Technology
Assessment (HTA) information resources, Available
from Internet ltURL http//www.nlm.nih.gov/nichsr
/ehta/chapter1.htmlgt, Accessed 14 March 2003.
Booth A, O'Rourke AJ, Ford NJ. 2000. Structuring
the pre-search reference interview a useful
technique for handling clinical questions. Bull
Med Libr Assoc 88(3)239-46. Chalmers I and
Altman DG eds. 1995. Systematic reviews. London,
BMJ Publishing Group. Counsell C. 1997.
Formulating questions and locating primary
studies for inclusion in systematic reviews.
Annals of Internal Medicine, 127(5), 380-387.
27
References
INAHTA. 2000. INAHTA International Network of
Agencies for Health Technology Assessment global
networking for effective healthcare, Stockholm,
INAHTA. McKibbon A. 1999. PDQ evidence-based
principles and practice. Hamilton, Ontario BC
Decker Inc. Mulrow C. 1995. Rationale for
systematic reviews. In Chalmers I and Altman DG
eds. Systematic reviews. London, BMJ Publishing
Group. Oxman AD, Sackett DL and Guyatt GH. 1993.
Users' guides to the medical literature. I. How
to get started. The Evidence-Based Medicine
Working Group. JAMA, 270(17), 2093-2095. Richardso
n WS, Wilson MC, Nishikawa J and Hayward RS.
1995. The well-built clinical question a key to
evidence-based decisions. ACP Journal Club,
123(3), A12-A13.
28
References
Sacket DL, Richardson WS, Rosenbery W, Haynes RB.
1997. Evidence-based medicine. How to practice
and teach EBM. 1st ed. London Churchill
Livingston, 1997. Snowball R. 2002. 'Find the
evidence'--reflections on an information skills
course for community-based clinical health-care
staff at the Cairns Library, Oxford. Health
Information Libraries Journal, 19(2),
109-112. Song F, Eastwood AJ, Gilbody S, Duley L
and Sutton AJ. 2000. Publication and related
bias. Health Technol Assess, 4 (10). Villanueva
EV, Burrows EA, Fennessy PA, Rajendran M and
Anderson JN. 2001. Improving question formulation
for use in evidence appraisal in a tertiary care
setting a randomised controlled trial. BMC
Medical Informatics and Decision Making, 1(1),
4.  
29

• Janette Boynton
• Senior Health Information Scientist
• 0141 225 6982
• jboynton_at_htbs.org.uk