Evidence-Based Practice Introduction to methods and searching for Librarians

1
Evidence-Based PracticeIntroduction to methods
and searching for Librarians

• Ann McKibbon MLS PhD
• McMaster University
• mckib_at_mcmaster.ca

2
Rules for the day

• Have fun
• Stretch your minds
• Make clinical decisions
• Ask questions and make comments
• Make mistakes and say dumb things
• Develop some new skills
• Stop me for breaks

3
Morning Agenda

• Introduction to EBM
• EBM and the question
• EBM and searching/retrieval
• Categories of care
• How are the similar
• How are they different
• EBM and the article
• Types of articles
• What makes a good one

4
Agenda

• How they differ?
• What makes them strong?
• How do I find them?
• Therapy
• Diagnosis
• Prognosis
• Etiology/harm
• Systematic reviews

5
Current Working (Pragmatic) Definition

• a set of tools, resources, and procedures
• for finding current best evidence from research
• and applying this evidence
• for decision making with respect to
• the care of individual patients (EBM, EBN, EBD,
  EBHC)
• the care of groups of people (EBPublic Health)
• the running of your library (EBL)
• raising your child (EBParenting)
• taking into account the situation, culture,
  resources, and common sense

6
Historical Roots of EBP

• Danielfirst clinical trial with Shadrack,
  Meshack, and Abednigo
• The results? At the end of 10 days, they
  "appeared better and fatter than all the young
  men who had been eating the royal rations"
  (v15), and the king found them in learning and
  wisdom to be "ten times better than all the
  magicians and enchanters in his whole kingdom"
  (vs18-19).

7
Ibn Sina (981-1037)

• Rules of drug evaluation
• The drug must have a specific defined mode of
  action
• It must be tested on a well defined disease
• The time of action must be observed
• The effect of the drug must been seen to occur
  consistently in many cases
• The experimentation must be done on the human
  body (horses or lions may react differently than
  a human)

8
Historical Roots of EBP

• Lindscurvy
• Late 1700s
• 12 sailors
• Dramatic results
• Slow implementation

9
OslerInformation

• It is astonishing with how little reading a
  doctor can practice medicine, but it is not
  astonishing how badly he may do it.

10
Historical Roots of EBP

• First randomized trial
• --not enough drug in UK
• --George Orwell not
• eligible for trial

11
EB Health CareFirst Version

• Patient
• 
  
• Evidence Clinical
  
  Expertise

12
EB Health CareNew Version

• Clinical Circumstances
• 
  Clinical
  Experience
• Patient preferences
  Evidence
• and actions

13
5 Steps of EBHC

• Framing the question appropriate to the needs of
  the patient
• Finding the evidence
• Evaluating the evidence (critical appraisal)
• Making and doing the decision
• Evaluation of the whole process

14
Critical appraisal

• Three-step process
• Find out how good (strong) the evidence is
  (assessment of the methods)
• Find out what the results are and how strong they
  are
• Figure out how our patient (or patient group) and
  settings matches with the study patients and
  setting

15
Critical appraisal

• Go to first page of supplemental package
• Pictorial representation of the EB process from a
  clinical perspective
• Most of the steps are done intuitively and not on
  paper

16
Why our searching skills are important
17
New Awareness of the Importance of Evidence and
Difficulties to Find/Accumulate

• Systematic reviews and meta-analyses
• Presence of search strategies a measure of
  quality
• Expert searchinghave we lost our edge?
• Seeing a true collaboration (work as equals)
• Searching and data management
• Cochrane and Campbell Collaborations, DARE
• Librarians full members with own sections on
  methods
• CADTH, AHRQ and Technology Assessment
• As above (Jessie McGowan)
• Guidelines and Care Maps
• Librarians true partners (Ruth Holst)
• Professional Societies hire librarians/contract
  for services
• CFPC, CMA, ACOG, AAN

18
Need for new information tools and techniques

• Books and book-like products
• Clinical Evidence has much librarian
  inputsearching etc
• PIER from ACP
• Summary Journals
• ACP Journal Club, EBM, EBN
• Web pages and sites
• MEDLINEPLUS and Go Local implementations
• Provide sites and production
• New products
• OVID EBM Reviews and other aggregated services
• PubMED Clinical Queries
• BMJUpDates
• New services

19
New Respect for our Abilities/Collections

• Grants
• Preparation, editing of grants
• Teaching of grant writing skills (Wessel Pitt
  modules)
• Institutional Review Boards
• Membership
• Standards (Kate Oliver)
• New service opportunities
• Scherrer and publication process
• Code YellowLibrary 911
• Teaching
• EBM workshops in Hamilton and Denver
• Limited only by our
• Imagination
• Ability to recognize and create opportunities

20
Clinical Research

• Move now to looking at clinical research
• the kind of research that is strong enough and
  applicable enough for use in making/changing
  clinical decisions

21
Clinical Research

• Question basedonce question formed
• Methods
• Observation
• Manipulation/experimental
• Which is stronger to do?
• Which is easier to do?
• Do we need both?
• Why?

22
Observational or Experimental?

• What is the process that elderly people go
  through as they come to terms with living in an
  assisted living arrangement?
• Does yoga improve fatigue levels in people
  recovering from cancer?
• How effective is this appetite reducing drug in
  relation to exercise?
• Do suduku puzzles protect me from developing
  Alzheimer disease?
• What is my life expectancy now that you have told
  me I have ALS?
• Is this blood test as effective as stomach biopsy
  at telling me that I have celiac disease?

23
Publishing Wedge (therapy)

• Idea
• Idea development
• Laboratory
• Animal
• Early human Phase I
• Middle human Phase II
• Late human Phase III

24
Categories of clinical literature

• Original studies
• Therapy
• Diagnosis
• Prognosis
• Etiology
• Clinical prediction guides
• Differential diagnosis
• Qualitative studies

• Syntheses
• Systematic reviews/ meta-analyses
• Clinical practice guidelines
• Economic studies
• Decision analyses

25
Similarities across categories

• Done to answer legitimate and important problems
  and issues
• Meet standards
• Ethical
• Moral
• Record keeping and reporting
• Publishing standards and procedures

26
Similarities across categories

• Question based
• Preplanned
• Comparative
• Patients/participants
• Results and statistics
• Settings
• Cultures/health care systems
• Conflict of interest/disclosure statements
• Funding sources

27
Clinical Question

• For young children, is a smoke detector/fire
  alarm that uses a recording of his or her
  mothers voice more effective at awakening the
  child and shortening time to evacuation as
  compared to a standard tone-based smoke detector?

28
Settings

• Important for assessment of match between patient
  and article
• Most often primary care, tertiary care, hospital,
  office (UK surgery), nursing home, university
  medical center, chronic care facilities, home
  care
• Compare the typical headache seen in above
  settings

29
Cultures/health care systems

• For profit
• Not for profit
• Managed care
• Health maintenance organizations
• Medicare/Medicaid
• Veterans Affairs Hospitals
• Socialized medicine
• Two tier vs. three tier

30
Health care systemsassignment

• What health care system does each country have?
• What type of research comes from each country?
• Canada
• Israel
• Sweden
• United Kingdom
• Australia
• The Netherlands
• United States

31
Funding Sources

• Complex and costly issue
• Competing demands with for profit funders vs.
  not for profit funders
• Kjaergard and Als-Nilesen showed that
• in pharmacological and nonpharmacological
  randomised controlled trials from 12 specialties
  financial competing interests were significantly
  associated with authors conclusions
• personal, academic and political competing
  interests were not significantly associated with
  authors conclusions.

32
Conflict of Interest

• Researchers or authors of the study or its report
  stand to gain (probably financially) from certain
  results of the study.
• Watch for this. Should be included in the
  articleusually at the end and in a small font.

33
Differences

• All articles same for a few features
• Question based
• Preplanned
• Comparative
• Patients/participants
• Settings
• Cultures/health care systems
• Funding sources
• Then different.

34
Therapy
35
Therapy/Treatment

• More of these types of studies than any other
• Methodology is pretty well established for
  conducting trials and presenting results
• Indexing (and retrieval) excellent

36
Alternate Names

• Therapy/Treatment (medicine)
• Intervention (nursing)
• Prevention and control
• Prevention
• Primary prevention
• Secondary prevention
• Tertiary prevention
• Quality improvement
• Management???

37
How to do a Therapy Trial

• 2 or more groups
• Each group gets 1 intervention
• All groups followed over time
• At the end of the trial groups compared to assess
  outcomes

38
Things to Look for in a Therapy Trial

• Common sense
• Allocation concealment
• Random allocation
• Blinding
• Follow-up
• Sensible and important outcomes

39
Allocation Concealment

• Done before and during randomization
• No one with any influence on who goes into which
  group can have any knowledge of which group is
  next
• Differences among outcomes if this is not
  donemore favourable outcomes

40
Random Allocation

• You can randomize
• parts of people (e.g., arms, warts)
• whole people
• families
• hospitals or wards (cluster)
• towns

41
Methods for Random Allocation

• Best
• Computer methods that do not allow for
  manipulation of randomization
• An agency that has no involvement in patient
  recruitment such as a pharmacy department
• An external trials office that entails calling
  into a central registration office and providing
  details of patient before the randomization is
  done
• Sealed, opaque, consecutively numbered envelopes
  with external checking

42
Blinding

• Individuals involved in a study (e.g., patients,
  investigators, research staff) do not know who is
  assigned to treatment or control groups.
• Why Individual expectations can influence study
  outcomes

43
Types of Blinding

• Single, double, triple blinding although many
  more could be listed
• Masked, dummy also used
• Patients
• Care providers
• Study personnel
• Data collectors
• Outcome assessors
• Data analysts
• Report writers
• Sponsors

44
Placebo

• To aid in blinding some trial participants may
  get something that is not really a treatment or
  for real
• Sugar pill in vitamin C trials
• Sham ultrasound in treating BPH and both real and
  sham kept equivalent by heating pads placed
  over treatment sites
• Can be for assessors in addition to patient as in
  fake blood for patients in a scope vs full
  surgery trial

45
Follow-up

• Concerns the number of participants who
  completed the study
• Look for withdrawals, drop outs, or those who
  were lost
• 80 is magic number
• Common sense

46
Clinical Question

• Pain is a complex challenge at the best of times
  and is especially difficult for children. A
  hospital committee has been formed that wants to
  look at alternatives to simply increasing the
  doses of pain meds. Someone has suggested that
  video games may distract kids with substantial
  burns from thinking about their pain. Does the
  following article support this assertion?

47
MeSH

• Research Comparative study
• Clinical protocols Placebos
• Feasibility studies Clinical trials as topic
• Pilot projects Clinical trials as topic,
  I-IV
• Research design Multicenter studies
• Double-blind method Randomized controlled
  trials Meta-analysis as topic
• Patient selection Treatment outcome
• Random allocation Single-blind method
• Sample size

48
Publication Types

• Clinical trial
• Clinical trial, phase I
• Clinical trial, phase II
• Clinical trial, phase III
• Clinical trial, phase IV
• Randomized controlled trial
• Controlled clinical trial
• Multicenter study
• Meta-analysis

49
Subheadings

• Therapy (explodable)
• Surgery
• Radiotherapy
• Diet therapy
• Psychology
• Therapeutic use (explodable)
• Administration and dosage

50
Textwords

• Random Double blind
• Double-dummy Mask
• Sham Placebo
• Control trial Efficacy
• Effectiveness

51
Diagnosis and Screening
52
Disease/condition Present?

• Diagnosis
• This patient has signs and symptoms that suggest
  diabetes. Does she have it?
• Signs and symptoms are present that warrant
  action.

• Screening
• We are going to check all students in this school
  to see if the head lice has spread from room 2.
• No symptoms are present but because of the
  population we are going to assess all of them.

53
Diagnostic Decision

• Three choices after assessing patient
• Do nothing for a whileI am not sure at all
  if the patient has a conditionlow probability of
  a positive diagnosis
• Treat right awayI am sure beyond any doubt
  that the patient has the conditionthe diagnostic
  tests will give me no more information
• Start doing diagnostic tests

54
Diagnostic Decision

• Probability of disease
• 0 testing threshold treatment threshold 100
• Wait Test Treat

55
Diagnostic Decision

• Wait Test Treat
• Probability of disease
• 0 testing threshold ? treatment threshold 100
• Positive test results moves ? ? ? ?
• Negative test results moves ? ? ? ?

56
Diagnosis studyexample

• Problem Is this incontinence urge or stress? One
  treated with drugs, other behaviour
• Possible solution questionnaire vs urologist?
• Tested 301 women some incontinence.
• Test results after every women got both
• 75 of time positive when have incontinence
• 77 of time negative when no incontinence

57
Diagnosis/Screening

• Does this person have or not have a specific
  disease or condition?
• Can questionnaires in family medicine settings
  screen for
• eating disorders
• Depression
• domestic violence
• Alzheimer disease
• drinking problems

58
Old Test vs New Test

• Need patients to have spectrum of disease (none
  to severe)
• Everyone gets both tests
• Old test is often invasive, time consuming,
  costly, or has risks involved
• Can only do if gold test available or can be
  rigged

59
Things to look for in a diagnosis study

• Old test vs new test
• Blinding of assessment of results of both tests

60
Blinding of Test Reading

• Absolutely crucial for evaluation of diagnostic
  tests
• Exceptions are things like laboratory tests that
  do not involve personal biases
• Blinding is almost NEVER indexed by NLM nor is it
  included in the abstract of the articles. Often
  hard to find in the body of the article (methods
  section)

61
A good test.

• is positive when it should be positive AND
  negative when it should be negative...
• Sensitivity and specificity
• Positive and negative predictive values
• False positive and negative reactions
• Positive and negative likelihood ratio

62
Calculating diagnosis numbers
Disease is present Disease is absent
Test shows positive (disease may be present) A B
Test shows negative (probably no disease) C D
A C B D
63
Sensitivity and Specificity

• Sensitivity---
• ---test is positive when it should be positive
• Specificity---
• ---test is negative when it should be negative

64
Predictive Values

• Positive---
• ---proportion of people with positive results
  who actually have the disease
• Negative---
• ---proportion of people with negative results
  who do NOT have the disease

65
False Positive/Negative Results

• False positive---
• ---test is positive when the person does not
  have the disease (labeling)
• False negative---
• ---test is negative when the person does have
  the disease (lose time)

66
Likelihood Ratios

• The likelihood ratio for a positive result (LR)
  tells you how likely you are to have a positive
  test if you HAVE the disease
• The likelihood ratio for a negative result (LR-)
  tells you how likely you are to have a negative
  test if you do NOT have the disease.

67
Diagnostic Decision

• Wait Test Treat
• Probability of disease
• testing threshold ? treatment threshold
• 0 100
• Positive test results moves ? ? ? ?
• Negative test results moves ? ? ? ?
• Likelihood ratios are the SLIDERS on this scale!

68
Concepts

• Accuracy
• Bayes theorem
• Diagnosis
• False negative
• False positive
• Negative predictive value

• Positive predictive value
• ROC curves
• Likelihood ratios
• Sensitivity
• Specificity
• Nomogram
• Screening

69
MeSH

• Sensitivity ad specificity
• Predictive values of tests
• ROC curves
• Diagnostic errors
• False negative reactions
• False positive reactions
• Observer variations
• Likelihood functions
• Diagnosis, differential
• Reproducibility of results

70
Publication Types
71
Subheadings

• Diagnosis (explodable)
• Diagnostic use (explodable)

72
Textwords

• Sensitivit
• Specificit
• Predictive value
• False positive
• False negative
• Likelihood ratio
• Accura

73
Prognosis
74
Prognosis/Natural History

• What is going to happen to me over the next
  period of time now that I have been diagnosed
  withPatient preference for this type of
  information and not therapy.
• Should I treat (or choose to be treated) rather
  than what is best to treat with. (Scoliosis and
  prostate cancer)

75
PrognosisExample

• Do we know the natural history or prognosis for
  patients who have have been diagnosed with
  Parkinson disease?
• 297 patients (181 men) with 1731 visits over 6.4
  years (mean) showed a variable course of the
  disease and its symptoms

76
Prognosis vs Natural History

• Traditional difference
• Natural history is what happens to untreated
  disease over time.
• Prognosis is what happens to treated disease

77
Prognosis vs Natural History

• Natural History
• Biological onset at cellular level
• Early diagnosis possible (screening)
• Usual diagnosis
• Outcomes

• Prognosis
• Usual diagnosis
• Outcomes

78
Prognosis Methodology

• Inception cohort study
• Group of persons assembled early (or at least at
  a uniform point) in the timing of their disease
  and followed over time.

79
Prognosis Methodology

• Inception cohort study
• Follow up 80 or better
• The follow-up time should be consistent with the
  demands of the disease
• Common sense rules

80
Prognosis Numbers

• Raw rates of disease progression (14 had a
  subsequent myocardial infarction within 5 years
  of follow up)
• Relative risks
• Hazards ratios
• Odds ratio
• Standardized mortality ratios

81
Prognostic vs Risk Factors

• Prognostic factors
• Some aspect that an individual has that can
  modify how that disease will play oute.g., age
  of person with CAD, breast cancer stage
• Risk factors
• Some aspect that an individual has that may
  affect whether that person is more or less prone
  to acquire a disease or condition, e.g., family
  history of CAD, Al and Alzheimers

82
Prognosis Terms

• Adjustment Natural History
• Cohort study Prevalence
• Inception Prognosis
• Incidence Prognostic factor
• Longitudinal studies Prospective studies
• Morbidity Risk factors
• Mortality rates, ratios

83
MeSH

• Cohort studies Survival XXXXXX
• longitudinal studies
• prospective studies
• follow-up studies
• Prognosis
• Morbidity
• incidence
• prevalence
• Mortality
• Survival analysis
• Disease progression
• Time factors
• Age factors
• Sex factors

84
Publication Types
85
Subheadings

• Mortality
• Epidemiology (sometimes)

86
Textwords

• Natural history Prognos
• Inception cohort Clinical course
• Predict Predictive value
• Outcome Prognostic factor
• Course

87
Alternative Source

• For short-term prognosis information, randomized
  controlled trials often have information that can
  be used when desperate. This is especially true
  for placebo-controlled or usual-care arms of
  studies.

88
Etiology/Causation/Harm
89
Etiology/Causation/Harm

• Etiology or causation is the study of what
  causes, or what increases or decreases the risk
  for a disease or condition. This can be either
  positive (protective) or negative (harmful).
• e.g., social support at work is associated with
  fewer short-term psychiatric absences
• giving up driving is associated with depressive
  symptoms in older adults.

90
Etiology/Causation/Harm

• Generally looking at
• Exposures to causative/protective agents
• Outcomes
• Time (can be now, in the future, or looking back)
• Other factors that can affect outcomes
• Need a lot of creativity and common sense
• Many long-term etiology studies come from
  countries with socialized medicine.

91
Q Does the use of cell phones while driving
cause accidents?

• Exposures?
• Outcomes?
• What groups?

92
Q Does the use of cell phones while driving
cause accidents?

• Randomized controlled trial
• large group of persons who
• are told whether they will
• use cell phone during driving for
  long time..

93
Q Does the use of cell phones while driving
cause accidents?

• Cohort study
• a group of persons who uses phones while
  driving is compared with another group who do not
  use phones while driving

94
Q Does the use of cell phones while driving
cause accidents?

• Case-control study
• people who have had automobile accidents are
  compared with people who have not had accidents
  and both groups are studied to see who used
  phones during driving

95
Q Does the use of cell phones while driving
cause accidents?

• Statistically adjusted groups (cross sectional
  study)
• Data on some persons with automobile accidents
  were collected and compared with data from some
  persons with no accidents and both groups have
  cell phone use rates determined.

96
Etiology study types

• Type Time Quality Number
• RCT future
• Cohort now
• Case-control past
• Statistically past - infinity
• adjusted groups
• (cross sectional studies)

97
Etiology Issues

• Association is not necessarily causation
• Ethics
• Logistics
• Blinding
• 2 CsCommon sense and Creativity

98
Association vs Causation

• Just because two things occur at the same time
  does not mean that they are causal--
• Higher intake of ice cream and higher rates of
  drowning occur in the summerlinked? Dont think
  so...
• Poor quality health and low socioeconomic
  status

99
Ethics

• Sometimes you cannot allocate persons to
  exposures
• Smoking
• Divorce
• Genetic disposition
• High socioeconomic status
• Can do this however for drug adverse effects,
  social programs

100
Logistics

• Groups need to be as similar as possible to
  account for confounding
• For power line studies..
• Smoking and drinking issues
• Another set of intertwined issues?
• Need for creative problem solving

101
Blinding

• Blinding is crucial especially for the
  case-control studies.
• Blinding must be 2-fold here
• forming groups without knowing exposures (cell
  phone use in cars)
• assessing exposure without knowing
  disease/exposure status (automobile accidents)

102
Relative Risk (RR)

• Used for RCTs and cohort studies (prospective)
• Comparison of rates of developing the disease/
  condition in the 2 groups of people with and
  without the risk factor (We know exposures)
• Weight gain and coronary heart disease in women
  (400 women in our sample)
• if gain gt 15 lb 106/200
• if gain lt 15 lb 56/200
• RR (106/200)/(56/200) 1.9

103
Concepts

• Association
• Blinding
• Causation
• Case-control study
• Cohort study
• Confounders
• Cross sectional studies

• Odds ratio
• Prospective studies
• Relative risk
• Retrospective studies
• Risk
• Risk factors

104
MeSH

• Case-control studies
• Retrospective studies
• Cohort studies
• Longitudinal studies
• Prospective studies
• Follow-up studies
• Cross sectional studies
• Risk
• Risk assessment
• Risk factors
• Odds ratio

105
Publication Types
106
Subheadings

• Etiology (explodable)
• Epidemiology (for distributions, causes, and
  attributes of disease)

107
Textwords

• Cohort
• Case control
• Risk
• Odds ratio
• Causation or causal
• Relative risk
• Etiol or Aetiol

108
Systematic Reviews and Meta-Analyses

• Two types of review articles exist
• Narrative reviews
• Systematic reviews
• Systematic reviews
• Meta-analyses of study data
• Meta-analyses using individual patient data

109
Systematic Reviews

• Medicine did not invent or develop
• Fully developed by psychology, education, and
  related disciplines
• Early 1900s
• Pearson and enteric fever in the British Army
• NIH (Hygiene Laboratory) report
• People
• Eugene Glass was one of the first developers
• Tom Chalmers was one of the first medical
  developers (On Golden Pond)
• Archie Cochrane plus Tom and Iain Chalmers

110
Systematic vs Narrative

• Team approach
• Narrow specific purpose
• Methods drive the process
• Inclusion/exclusion criteria
• Clinically useful

• One major author
• Broad purpose may not always be stated
• No methods on how articles picked
• No inclusion/ exclusion criteria
• Ecuationally useful

111
Systematic Reviews

• Must have
• Purpose why done
• Search strategy in detail
• Inclusion and exclusion criteria for study
  selection

112
Why do one?

• Too much data
• Too little data
• Resolve discrepancies
• tighten up estimates of effects of treatments,
  exposures, etc
• Analyze patient subgroups
• Plan for new studiessimilar or next generation
  studies
• Provide data for certain types of studies such as
  economic studies, decision analyses, or clinical
  practice guidelines

113
Steps

• Problem formulation
• Identify and select articles (searching and
  retrieval)
• Data extraction
• Analysis and decision if meta-analysis is
  appropriate (clinical and statistical test)
• Presentation of results

114
Statistical Concepts

• Homogeniety
• Heterogeniety
• Weighting
• Pooling
• Effect sizes

115
Statistical Concepts

• Same as for other study types but often qualified
  with combining phrases
• combined odds ratio
• pooled relative risk
• weighted hazards ratio
• typical mortality rate
• summary estimates sensitivity

116
MEDLINE Difference

• Simplistic difference (and hard to differentiate
  between the two)
• Reviewdoes a summary of existing knowledge
• Meta-analysisproduces new knowledge

117
Meta-analysis MeSH

• A quantitative method of combining the results of
  independent studies (usually drawn from the
  published literature) and synthesizing summaries
  and conclusions which may be used to evaluate
  therapeutic effectiveness, plan new studies,
  etc., with application chiefly in the areas of
  research and medicine.
• Clinical trials overview, data pooling,

118
Meta-analysis Previous Indexing

• Follow up studies
• Outcome and process assessment
• Research
• Research design
• Statistics

119
Meta-analysis Publication Type

• Works consisting of studies using a quantitative
  method of combining the results of independent
  studies (usually drawn from the published
  literature) and synthesizing summaries and
  conclusions which may be used to evaluate
  therapeutic effectiveness, plan new studies, etc.
  It is often an overview of clinical trials. It is
  usually called a meta-analysis by the author or
  sponsoring body and should be differentiated from
  reviews of literature.

120
Review Literature MeSH

• Published materials which provide an examination
  of recent or current literature. Review articles
  can cover a wide range of subject matter at
  various levels of completeness and
  comprehensiveness based on analyses of literature
  that may include research findings. The review
  may reflect the state of the art. It also
  includes reviews as a literary form.

121
Systematic Reviews Terms

• Effect sizes
• Heterogeniety
• Homogeniety
• Meta-analysis of study data
• Meta-analysis of individual pt data
• Pooling

• Narrative review
• Summary estimates
• Systematic reviews
• Typical
• Weighting

122
MeSH

• Randomized controlled trials
• Clinical trials
• Controlled trials
• Meta-analysis

123
Publication Types

• Meta-analysis
• Review iff
• ANDed with MeSH or textwords

124
Subheadings
125
Textwords

• Meta-analy
• Metanal
• Metaanal
• Systematic review or overview
• Overview (careful)
• Quantitative review or overview

• Methodologic review or overview
• Heterogeniety
• Homogeniety
• Medline
• Psychinfo
• Psycinfo
• Embase

126
Food for Thought

• Sometimes a rare condition that is being written
  up as a case report will include a substantial
  amount of background material. Dont necessarily
  cross out a case-report if you are desperate for
  review type material.

127
Alternative Sources

• PubMed Clinical QueriesReview articles
• Cochranenote only for RCTS
• Cochrane reviews
• Dare
• Clinical Trials
• DARE (Database of Reviews of Effectivenesslibrari
  an run project)
• Campbell Collaboration
• TRIP databaseone stop shopping

128
Systematic Reviews Quiz

• 1 Systematic reviews are the same as ordinary
  reviews only bigger?
• Based on Petticrew M. Systematic reviews from
  astronomy to zoology myths and misconceptions.
  BMJ. 200132298-101.

129
Systematic Reviews Quiz

• 2 Systematic reviews contain only randomized
  controlled trials.

130
Systematic Reviews Quiz

• 3 Only medical topics can be studied in
  systematic reviews.
• More money on schools improved educational
  outcomes?
• Do men or women make better leaders?
• Does the sexual orientation of the parent matter
  to the childs well being?
• Are fathers more likely to treat their children
  differently than mothers?
• Is job absenteeism an indicator of job
  dissatisfaction?
• Are jurors influenced by the race of the
  defendants?
• Is there a relation between poverty and violence?

131
Systematic Reviews Quiz

• 4 Systematic reviews can be done well without
  experienced information personnel or good library
  support.

132
Systematic Reviews Quiz

• 5 Systematic reviews have no relevance to the
  real world

133
Systematic Reviews Quiz

• 6 Systematic review necessarily involve
  statistical syntheses.

134
Systematic Reviews Quiz

• 7 Systematic reviews have to be done by experts
  with a lot of training in methodologies.

135
Systematic Reviews Quiz

• 8 Systematic reviews are a substitute for doing
  good quality individual studies.

136
Systematic Reviews Quiz

• 9 You can be part of a team doing a systematic
  review.