EVIDENCE BASED PRACTICE
BY MOHEB KHOUZAM MD
EVIDENCE BASED PRACTICE 1- TYPES OF STUDY DESIGNS 2- RECOMMENDATIONS
EVIDENCE BASED PRACTICE
1-TYPES OF STUDY DESIGNS
Intervention Observational Others
1-TYPES OF STUDY DESIGNS
A. INTERVENTION STUDIES
Intervention Study
Level I Level II
Randomized Controlled Trials
Level II 1Controlled Trials
WithoutRandomization
1-TYPES OF STUDY DESIGNS
Intervention Prospective Random method Experimental and control groups
RANDOMIZED CONTROLLED TRIALS (RCT)
1-TYPES OF STUDY DESIGNS
CONTROLLED TRIALS WITHOUT RANDOMIZATION
Intervention Prospective No randomization Experimental & control groups
1-TYPES OF STUDY DESIGNSB-OBSERVATIONAL STUDIES
Observational Studies
Level II-2aCOHORT
(Relative Risk)
Level II-2bCASE CONTROL
(Odds Ratio)Level II-3 Level III
1-TYPES OF STUDY DESIGNSCOHORT STUDIES
Exposure status known Follow-up for a period of time Determine presence or absence of disease
(i) Prospective: groups of exposed & unexposed followed-up for a period to determine outcome
(ii) Retrospective: both exposures & outcomes occurred. Rate of disease in exposed / Rate of disease in
unexposed gives Relative Risk
1-TYPES OF STUDY DESIGNSCASE CONTROLLED STUDIES
Retrospective Group of subjects with specific outcome (cases) Group of subjects without specific outcome
(controls) Over a period of time Here starting point is disease status Compare extent of exposure to a variable of
interest Odds of exposure in cases/odd of exposure in
controls= Odds Ratio
1-TYPES OF STUDY DESIGNS
B- OBSERVATIONAL STUDY
Level II-3
Cross Sectional(Prevalence)
Uncontrolled Investigational
1- TYPES OF STUDY DESIGNSCROSS SECTIONAL STUDIES Observational Assess the status of individuals with respect to
presence or absence of both exposure & outcome at a particular time.
Prevalence is the proportion of individuals with a disease at a particular time
Incidence is the number of new cases occurring over a specified period of time
1-TYPES OF STUDY DESIGNS
UNCONTROLLED INVESTIGATIONAL STUDIES
Observational Report the results of treatment or
interventions in a particular group No control group
1-TYPES OF STUDY DESIGNS
B- OBSERVATIONAL STUDY
Level III
Descriptive(Case Report/Series)
Expert Opinion
1-TYPES OF STUDY DESIGNS
DESCRIPTIVE STUDIES
Provide limited information on relationship between exposure and outcome of interest
Case study describes clinical characteristics or interesting features in a single patient
Case series describes the same in a series of patients
1-TYPES OF STUDY DESIGNSEXPERT OPINION
Findings from expert panels and committees and the opinions of respected experts in a particular field
1-TYPES OF STUDY DESIGNSA, B, C-OTHERS
Other Studies
Decision Analysis Meta-Analysis Cost EffectivenessCost-Benefit
1-TYPES OF STUDY DESIGNS
Decision Tree, branches are key problems or decisions Driven by key assumptions
DECISION ANALYSIS
DECISION ANALYSIS
Hysterectomy
Total
Subtotal
Op compl
UneventfulUneventful
Op compl
DeathL/T dis
Cure
Death
L/T dis
Cure
CureCureHysterectomy
TotalOperative complications
Death
Long Term disability
Cure
Cure
Cure
Cure
Death
Uneventful
Uneventful
Subtotal
Operative complicationsLong Term disability
META-ANALYSIS
Effect on endometrial cancer: smoking vs. no smoking
Weighted average of a variable/intervention on a defined outcome
•Beers•Dipalo•Henry•Zo
1-TYPES OF STUDY DESIGNS
Dipalo
Henry
Zo
1-TYPE OF DESIGN STUDIESCOST-BENEFIT ANALYSIS
Compares costs associated with an intervention with the net monetary benefits from the use of that intervention
1-TYPES OF STUDY DESIGNSCOST-EFFECTIVENESS ANALYSIS
Compares the net monetary costs of an intervention with some measure of clinical outcome
2-RECOMMENDATIONSBASED ON QUALITY & QUANTITY OF EVIDENCE
Recommendations
A B EC D
2-RECOMMENDATIONS A There is good evidence to support
recommendations B There is fair evidence to support
recommendations C There is insufficient evidence to support
recommendations, but recommendation is made on other grounds
D There is fair evidence against recommendations
E There is good evidence against recommendations
There is no exact correlation between strength of recommendations and level of evidence