Part - time MSc course Epidemiology & Statistics Module

Download Report

Transcript Part - time MSc course Epidemiology & Statistics Module

The following lecture has been approved for
University Undergraduate Students
This lecture may contain information, ideas, concepts and discursive anecdotes
that may be thought provoking and challenging
It is not intended for the content or delivery to cause offence
Any issues raised in the lecture may require the viewer to engage in further
thought, insight, reflection or critical evaluation
Literature Searches
Electronic Searches
Dr. Craig Jackson
Senior Lecturer in Health Psychology
School of Health and Policy Studies
Faculty of Health & Community Care
University of Central England
[email protected]
Aims & Objectives
Aims
Establish the importance of literature reviews
Assist in the practicalities of performing a literature search
Provide guidelines for the Critical appraisal of papers
Objectives
Introduction to literature searching
Interviewing electronic data sources
Allow critical appraisal of scientific papers
To enable the production of a substantial piece of research within a speciality
field, addressing a specific question and carried out in scientific manner
Information sources among physicians
Medical
Other
%
%
No time to search
25
20
Colleagues
52
67
Meetings
13
30
Textbook / Journal
30
21
Internet
14
18
5
7
21
2
Other
No info needed
Searching Behaviour
Clinicians rely on expert based systems “wetware”
(consulting colleagues, journal reviews, textbooks, and continuing ed.)
Academic researchers rely on computer resources
hardware and software
Expert Based Information
Benefits
• Wisdom through experience
• Clinical nuggets that can’t be derived through the scientific method
• Fill in the gaps in outcome-based knowledge with evidence derived from
clinical experience
Detriments
• Info may be out of date
• Reverse gullibility: clinical experience favoured over patient oriented
evidence
• Knowledge may be based on highly selected population
The “Revolutionary Changes”
• Internet access to literature databases
• Increase in volume of general medical literature
• Systematic reviews are more common
• Specialist medical, psychological, nursing, ergonomic databases
• Practice guidelines based on reviews
• On-line purchase of articles
• On-line full-text journals
P I C O Method for understanding papers
•P
Patient / Population characteristics
•I
Intervention / Exposure / Factor of interest
•C
Comparison / Control Condition
•O
Outcome
PUBmed
www.pubmed.com
Good Research Questions
Diagnostic Questions
Compare with gold standard
Etiologic Questions
Case-Control study
Intervention Questions
RCT
Prognostic Questions
Patient cohort study
Search Strategy
Use MeSH browser (Medical Subject Headings)
If MeSH-term not available, use “free text”
Aim at maximum of 50 studies
Sensitive search – miss little but gain useless studies
Specific search - more exact but few results
Search filters
Quality Rules of Thumb
Better quality if…..
Medicine better versus Internet
Journals with High Impact Factor
English Language Journals
Structured Abstracts
(Intro, Methods, Results, Conclusion)
Purpose
1. Define a problem to be researched
2. Review and summarise relevant literature (timescale)
3. Present this as Literature Review (word count) within dissertation
4. Viva considerations
“The selection of available documents (both published and unpublished) on a
topic …….. and the effective evaluation of these documents in relation to the
research being proposed”
Hart, C (1998)
The literature review justifies the topic chosen
Benefits
Understand previous research on topic by knowing:
1. The main theories in the subject area
2. Evolution / Acceptance of such theories
3. How the theories have been applied
4. Main criticisms that have been made of work on the topic
5. Theoretical limitations
Underpins knowledge needed for dissertation
Examiners like to see. . .
•
Identification of:
Main concepts, theories, theorists and methodological approaches
•
Understanding & Critical Evaluation
•
Literary flow
•
Systematic reasoning
•
Demonstration of the need for this research
•
Demonstration that the research contributes something new
Key Points
Appropriate breadth and depth
Rigour
Consistency
Clarity
Brevity
Sound analysis
Don’t base size on word-length
Process of searching
Select key words in subject area
Search sources using key words
(CD ROMs, on-line bibliographic databases, library catalogues)
Select key papers
Obtain some key papers
(photocopy, ILL, subscription)
Other papers
Obtain abstracts (e)
(make database)
Critically review key papers
Critically review other papers
Develop themes
Write 1st draft of review
Tips
Earlier the better - informs methodology
Thorough – having papers / abstracts to hand at write up helps
Think how relevant paper is to subject – be strict at a later date
Refine earlier attempts at searches – re-run every three months
Journal e-subscriptions & e-alerts
Information overload can be good
Look for
argument
style
problem
events
conclusions
theory
motives
hypothesis
standpoint
politics
justification
concepts
techniques
questions
evidence
definitions
ways of thinking
perspective
interpretation
ethical consideration
conflicts of interest
funding
validity
methodology
POEM analysis
Patient Oriented Evidence that Matters
U=
U
R
V
W
R x V
W
usefulness
relevance of info
validity of info
work required to get info
3 criteria:
1) address a question that researchers encounter
2) measure outcomes that researchers care about:
symptoms, morbidity, quality of life, or mortality
3) they have the potential to change practise
Slawson & Shaughnessy 2002
Electronic searches are getting easier
Despite the hassle….
Fast
Cheap
Reliable
Common
Simplistic
Interface improvements
Hard but not impossible to compete with the qualities of a real book
Varieties of sources
Papers
Databases
Search engines
Datasets
Journals
Personal
communication
M.U.D’s
Google “ask”
The Useful Web
No customs / No postman
Good at what it does - it’s always there
Instant / very fast
Regulation
Quality control
Transient nature
No international referencing format
Mostly free
Not easily transportable
Interactive
Bits not atoms
Vast
Not always easily accessible
Bits not atoms
Vast
The Useless Web
Maximizing the web
Requirements
Browser you feel comfortable with (16 to choose from)
Netscape Navigator v Internet Explorer
Reliable search engine(s)
individual taste
Adobe Acrobat / Reader
allows viewing of PDF documents (Portable Document Format)
Time
to experiment
to find own style
Quality hits
Surfing is 90’s
Searching time
• Most of useful work done in first 50th % of interface
• If not found early, info is of little use
• Suggests www works
• Information paradox
Useful Health Research websites
www.occenvmed.com
Occupational & Environmental Medicine
www.bmj.com
British Medical Journal
www.fathom.com
Fathom
ehp.niehs.nih.gov/
Environmental Health Perspectives
www.unc.edu/~rowlett/units/index.html Units of measurement
www.researchpaper.com
Research paper.com
www.ohsonline.com
Occupational Health
annhyg.oupjournals.org
Annals of Occupational Hygiene
www.thelancet.com
The Lancet
www.newscientist.com
New Scientist
www.niss.ac.uk
Ovid database
www.scit.wlv.ac.uk/ukinfo/uk.map.html Map of UK Universities
www.occmed.oupjournals.org/
Occupational Medicine
www.joem.org/
Journal of Occupational and Environmental Medicine
www.fb4d.com
Free books for doctors
www.ncbi.nlm.nih.gov
Pub Med
Recommended Search Engines
http://search.netscape.com
Netscape
http://uk.yahoo.co.uk
Yahoo
http://www.excite.com
Excite
http://www.lycos.co.uk
Lycos
http://www.looksmart.com
Looksmart
http://www.google.com
Google
http://hotbot.lycos.com
Hotbot
www.copernic.com
META-SEARCH ENGINE
That which is not…
• Journal Publications still the standard medium
• Peer review still best process
• Web not a shortcut to quality publication
• Owning a (hard) copy of a journal paper is still essential
• Web is there to facilitate access to journals and communication
www.craphound.com
Accessing Worthwhile Research
PROBLEMS
Getting started . . . . . insurmountable task
Too much vs too little
Technology
Judging relevance
Unfamiliar literature
Awareness of available resources
Non-local materials
Keeping track of findings
Keywords and Synonyms
“WOMEN AND WORK IN HISTORY”
Keywords:
Women, Females, Girls, Gender
Synonyms & near synonyms:
Child, Adolescents
Related items:
Work, Employment, Labour, Workforce, Industry, Management
Broader terms:
Early, Medieval, Modern, Contemporary Britain, UK, England, Scotland
Boolean Logic
Basis of most modern search processes
George Boole (19th C)
1. Mapping everything into Bits & Bytes
2. Pre-cursor to Logic gates - essential to computing
3. Excludes terms from a search
The NOT Gate (Inverter)
Takes an Input and produces an Output which is the opposite
NOT GATE
A
Q
A
Q
0
1
1
0
Boolean Logic
The NOT operative
Excludes a term from a search
“Pollution NOT oil”
“Labour NOT politics”
“Disease NOT cancer”
A powerful operating term
Boolean Logic
Variety of search terms
Think carefully about the search terms used
Consider vocabulary -
Scientific
Medical
Classical
Variations
Reiterate search results and re-try
Constant Refinement Process
Use “set-building” for searches
(spelling) (name)
(context) (hyphens)
Boolean Logic
Priority of Boolean process
OR before AND
then finally NOT
“boys OR men AND hobbies NOT football”
Use ( ) to group OR searches
“Stress AND Sex”
“Stress AND Women”
“(Women OR Girls) AND Stress NOT Trauma”
Boolean Logic
Building sets
Returns
1.
“teenagers OR adolescents”
2.
“#1 AND (aggressive OR violent)”350
3.
“#2 AND (treatment OR rehabilitation)”
215
4.
“#3 NOT (females OR girls)”
196
5.
“#4 AND (advantages)”
121
Logical relationships become clear
Search is built step-by-step
Mistakes (sudden drops in citations) easier to see
500
eXcite for webservers
More advanced form of searching than Boolean Logic
Concept-based
Query is simply a description of an information need
Search for documents that are a best match for the words in query
Also search for documents about the same concepts that query describes
Sometimes bring back articles that don't mention any of the words in query
The description / query can be as detailed as you want
Common words are ignored:
a
and
the
Meta-search Engines
Single search on a number of different search engines simultaneously
DOGPILE
INFERENCE FIND
MAMMA
METACRAWLER
PROFUSION
SAVVY SEARCH
Limitation Awareness
“No research project is perfect. What is important though, is that the
researcher recognises this, justifies why design and analysis
decisions were taken, and interprets the results in the light of these.”
Hicks 1999