Research Synthesis of Population-Based Prevalence Studies

1
Research Synthesis of Population-Based Prevalence
Studies

• ORC Macro
• Benita J. OColmain, M.P.H.
• Wanda Parham, M.P.A.
• Arlen Rosenthal, M.A.
• Adrienne Y. Semidey, M.A.
• National Eye Institute
• Rosemary Janiszewski, M.S.

2
Research Synthesis

• Synthesis of research findings offers the
  potential to
• Identify areas of agreement
• Increase statistical power
• Identify areas needing more research
• Fosters a greater sensitivity to the currently
  available scientific evidence

3
Purposes of Research Synthesis

• State-of-the-art literature review
• Presents and analyzes available data, framing
  results so they can be translated into practice
  and policy
• Logical framework for systematic and objective
  review of data (scientific results)
• Establish what is known prior to further research

4
Research Synthesis Essential Steps

• Specification of objectives, hypotheses, scope,
  and methods of systematic review
• Compilation of a comprehensive set of relevant
  primary studies
• Assessment of the methodological quality of the
  studies
• Identification of a common set of definitions of
  outcome, explanatory and confounding variables
• Extraction of a standardized set of outcome
  measures and of study subjects and
  characteristics

5
Meta-analysis

• Subset of research synthesis
• Statistical analysis of a large collection of
  analysis results from individual studies for the
  purpose of integrating the findings (Glass, 1976)
• May produce more generalizable results
• Helps to explore variation between studies and
  assess factors which might modify effects

6
Purposes of Meta-analysis

• Statistical methods for combining results
• Fixed vs. random effects models
• Gain in statistical power for estimates
• Reduces the effect of random error
• More precise summary estimate
• Resolve issues related to conflicting results
  from studies
• Formal/informal models for heterogeneity

7
Meta-analysis Essential Steps

• Use appropriate and clearly stated models
• Explore and allow for all important sources of
  variation
• Report confidence intervals for pooled estimates
• Evaluate tests of heterogeneity
• Clear presentation of results to enable critical
  appraisal and replication

8
(No Transcript)
9
Guiding Principles

• In a research synthesis/meta-analysis, individual
  study results are the raw data
• Systematic reviews are retrospective
  observational studies and thus are potentially
  subject to the same biases as other retrospective
  observational studies
• Need an a priori protocol for data selection and
  analysis

10
Population-Based Epidemiological Eye Studies

• Conducted over the past 20 years
• Framingham, 1980
• Assess prevalence of eye disease based on results
  of clinical examinations
• Evaluate vision impairment and blindness
• Study risk factors associated with eye disease

11
Objectives of Research Synthesis of
Population-Based Eye Studies

• Estimate prevalence of blindness and visual
  impairment (all cause)
• Estimate prevalence of
• Age-related macular degeneration
• Diabetic retinopathy
• Glaucoma
• Cataracts
• Refractive error (myopia and hyperopia)
• Estimate attributable blindness and visual
  impairment due to these diseases

12
Methodological Approach Research Synthesis

• Literature review
• Identify appropriate studies for each disease
• Determine inclusion/exclusion criteria
• Review case definitions for each disease
• Develop consistent criteria across studies
• Use uniform age categories for prevalence
  estimates (5 year CDC standard)

13
Preliminary List of Studies
14
Methodological ApproachMeta-analysis

• Obtain raw prevalence data from each study
  according to standardized case definitions and
  uniform age groups
• Determine appropriate models
• Review graphical and statistical output
• Collapse cells where data is sparse
• Assess heterogeneity

15
Current Status

• Completed meta-analysis for age-related macular
  degeneration
• Derived age-specific prevalence rates
• Geographic and neovascular AMD
• Drusen gt125
• Blindness due to AMD
• Meta-analysis for diabetic retinopathy in
  progress

16
Next Steps

• Complete meta-analysis for other diseases
• Determine current burden of
• Each disease
• National and State level by race and gender
• Blindness and vision loss due to each disease
• Future projections
• Publish results