Evidence Based Medicine

1
Evidence Based Medicine

• Clinical Practice Guidelines

2
What is Evidence Based Medicine?

• Philosophy of medical practice to aid in the
  approach to decision making in the clinical care
  of patients.
• It is integrating individual clinical expertise
  (internal clinical evidence) with the best
  available external clinical evidence from
  systematic research of the literature

3
Evidence Based Medicine...

• We make decisions based on the best clinical,
  scientific ______________
• Critical appraisal of published literature helps
  determine
• ______________ of evidence
• ______________ of evidence
• Goal of EBM to improve patient outcomes.

4
Internal Clinical Evidence

• Individual Clinical Expertise
• Training
• Experience
• Knowledge

5
External Clinical Evidence

• Literature Research
• Randomized Controlled Clinical Trials
• Meta-Analysis Studies
• Cohort Studies
• Case Controlled Studies
• To answer clinical questions regarding
• Therapy, Diagnosis, Etiology, Prognosis,
  Prevention, Quality Improvement

6
Literature Review Pitfalls

• Not all medical journals publish high quality
  studies. (We critically evaluate validity of
  studies)
• Authors will have inherent bias (by words they
  use, points they stress or dont stress, etc.)
• Editors and reviewers are human (bias)
• Negative studies frequently arent published

7
7 Levels of Evidence for Literature Evaluation

• Evidence from
• Grade l- Large RCT (gold standard)
• Grade 2- Small, not as well designed RCT,
  Meta-Analysis studies
• Grade 3- Cohort studies (retrospective or
  prospective, non randomized, observational)
• Grade 4- Case-control studies (non-rand., always
  retrospective, observational)
• Grade 5- uncontrolled or poorly controlled
  studies, case series (descriptive),
  cross-sectional, Review articles
• Grade 6 -Conflicting studies
• Grade 7 -Expert opinion

8
Randomized Controlled Clinical Studies/Trials
(RCT)

• True experimental study
• Most useful for quantifying data
• Well designed studies will infer cause/affect
  relationship (good thing)
• Randomization will reduce bias and confounding
  variables.
• Most useful study design to determine benefits,
  harm, actual efficacy of drug, etc.

9
Randomized Controlled Trials

• Parallel two groups of patients one give study
  drug, one group give control, watch both groups
  to observe outcomes
• Crossover one group of patients half take study
  drug, then control drug other half take control
  drug, then study drug.
• Before and After (___________) one group of pts-
  take measurements before drug, give drug, then
  take measurements again.

10
Limitations of Randomized Controlled Trials

• Expensive to perform
• Many are funded by Drug companies-- big bias
  potential
• External factors can always influence validity of
  any study
• Statistical verses clinical significance
• Need trials long enough to see outcomes

11
Potential Areas of Bias in any Study

• Credentials of researchers
• Funding of study
• Is study blinded?
• Is study design adequate- internal validity
• Is population representative of real world
  population?
• Randomization of treatment groups

12
Observational Studies

• Cohort Study (follow-up study)
• Most powerful after RCT.
• Follows patients to find out who experiences the
  outcomes two groups those who received
  intervention (drug), and those who didnt.
• Cannot determine cause and effect since bias in
  design
• Good for finding associations
• Can be expensive because of time involved.
• Usually done retrospectively looking back at
  records, using memory for recall, etc.

13
Observational Studies cont..

• Case-Control design
• Start with patients who have outcome, identify
  control group (those without outcome) and look
  backwards for risk factors potentially causing
  outcome. (Search for factors in the past that may
  explain the outcomes)
• Good for rare diseases, rare adverse events
• Retrospective, bias easily introduced

14
Observational Studies cont

• Cross-Sectional
• All outcomes and exposure data present during a
  specified period are evaluated.
• Identifies variables that predict outcomes
• Taking a look at one period in time may not be
  predictive of another time period.
• Useful for outcome assessments, predicting future
  costs, identifying severity of illness

15
Weakest Observational Design..

• Descriptive or Case-Series Report
• No control group
• Difficult to make conclusions or associations
• Most bias, cheapest to perform
• Reporting on a circumstance, event, program and
  trying to make some inferences
• Example Reporting on changes in outcomes before
  and after an intervention program was
  implemented.

16
Clinical Practice Guidelines

• One of the tools used in Evidence Based Medicine.
• Purpose of CPG to increase quality and value of
  care.
• Provides valid information which can improve
  decision making, reduce errors, help correct
  over-use and under-use of health care system, and
  reduce errors

17
Factors Causing the Implementation of CPG

• Increased rate of growth of health care costs
• Large variations of health care practice in
  different geographical areas of the U.S.
• Inappropriate use of health care services, ie.
  Lab tests, diagnostic surgical procedures, Rx
  meds, hosp. Admissions, length of stay.
• Uncertainty of health outcomes from use or
  non-use of various services or treatments

18
Agencies involved in EBM and CPG

• U.S. Dept of Health and Human Services
• NIH (National Institutes of Health)
• FDA (Food and Drug Agency)
• http//www.guideline.gov
• CDC (Center for Disease Control)
• AHCPR (Agency for Health Care Policy and
  Research) http//www.ahcpr.gov
• HCFA (Health Care Financing Admin)

19
Key Goals in Developing Guidelines

• Increasing access to health care
• Cost Containment
• Increasing quality of health care

20
EBM Steps and Decision Process

• Do systematic search
• Gather evidence (literature, experience, opinion,
  etc)
• Critically evaluate evidence
• If evidence of benefit and value, proceed
• If evidence of no value or no benefit, or harm,
  do not recommend or proceed.
• Weigh benefits, risks, costs and develop
  recommendations, or guidelines.

21
Role of Pharmacists

• Evaluating guidelines to recommend for practice
• Drug utilization review and evaluation
• Development and implementation of prescribing
  guidelines or protocols for restricted use drugs
  via PT committee.
• Work with HMOs, State Medicaid programs
  developing therapeutic guidelines for drug use
  and management

22
Examples

• Do we need to use the latest quinolone antibiotic
  on the market for our formulary, or can we use an
  existing, less expensive one to cover our typical
  bacteria seen in our setting/institution/community
  ?What literature evidence is there for better
  therapeutic effect or better patient outcomes?

23
Another Example

• Vitamin E is found in some case-controlled
  studies to have benefit for cardiac protection in
  preventing an MI? Is this enough evidence for us
  to recommend and promote the use of Vitamin E in
  our population/community/setting? Or to put it
  into our cardiac rehab guidelines for high risk
  patients?

24
Disease-oriented Outcomes vs. Patient-oriented
Outcomes

• Disease-oriented outcomes
• physiologic, intermediate, surrogate endpoints
  (blood sugar, blood pressure, flow rate, coronary
  plaque thickness) that may or may not reflect
  improvement in patient outcomes
• Patient-oriented outcomes
• outcomes that matter to patients and help them
  live longer or better (reducing mortality,
  morbidity, lower cost, reduce symptoms)

25
Examples of Outcomes
26
How Do We Evaluate Guidelines?

• A. Recommendation based on consistent and good
  quality patient-oriented evidence
• RCTs, meta-analysis, good cohort studies
• B. Recommendation based on inconsistent or
  limited quality patient-oriented evidence
• Un-controlled trials, trials with inconsistent
  findings, lower quality cohort or meta-analysis
  or systematic reviews, case-controlled studies

27
How Do We Evaluate? Cont...

• C. Recommendation based on consensus, usual
  practice, disease-oriented evidence, case series
  for studies of treatment or screening, and/or
  opinion.

28
Evaluating Studies used in making
Guidelines/Recommendations

• Introduction
• Is study objective clearly stated?
• Is study objective reasonable?
• Is study objective specific enough to be
  measurable?
• Is the study setting appropriate for this type of
  study?

29
Evaluating Studies.

• Methods Section
• Is study observational or experimental?
• What is the study design?
• Is study randomized?
• Is study prospective?
• Is study blinded?
• Is study controlled?

30
Evaluating Studies.

• Study Groups
• Are they randomly selected into study?
• Are they equal in ages, diseases, other factors?
• Is the blinding procedure adequate?
• Are the inclusion/exclusion criteria appropriate
  and sufficient to describe the target population?
• Are the endpoints measurable with the study
  population?

31
Evaluating Studies...

• Methods
• Are statistical tests used for analyzing results
  appropriate for the type of data collected?
• Is there internal validity? (Is study design
  appropriate for study population, setting,
  methods?)

32
Evaluating Studies...

• Results
• Are drop-outs accounted for?
• Are drop-outs included in statistical analysis?
• Are the correct endpoints measured?
• Are endpoints measured correctly?
• Are descriptive statistics provided for all
  indices of measurement stated in the methods
  section?
• Are P-values and/or CI given?
• Is there external validity?

33
Evaluating Studies...

• Discussion
• Are authors comments justified by the results?
• Are conclusions consistent with the study
  objectives?
• Is there statistical importance to this study?
• Is there clinical importance to this study?

34
Evidence Based Table