Observational Study Design

Observational Study Design

1. Cross-sectional Study (Prevalence)

   • Limitations
     • NOT USEFUL: Interpreting temporal or etiologic associations between exposure or risk factor and a health outcome
   • Strengths USEFUL:
     • Hypotheses generation and ideas for further research
     • old - new cases comparision
     • Analysis of general and special or representative populations groups over a specific period of time and location
   • Types of bias
     • Information, response, selection bias and confounding by time
   • Examples
     • National Health Examination Surveys Health questionnaire surveys of the elderly or adolescents

2. Longitudinal (Prospective or Cohort) Study

   • Strengths USEFUL
     • Follow-up of 2 categorize (nom-ramdoly) groups; expose or not to a risk factor(s) of interest
       • Calculation of Relative risk (RR)
       • Examination of the difference between incidence rates: indvs. exposed to the risk of factor VS those who weren't
     • Delineation of dose-response relation
   • Limitations
     • Considerable time, money and effort
     • Changes in person's exposure status
     • Losses to follow-up higher that expected
   • Types of bias
     • Information bias, potential misclassification of exposure and outcome, ascertainment bias
   • Variants
     • Concurrent
       • Initially classified and followed for a period
     • Non-concurrent
       • Classified in the past and followed up to the presence
     • Nested Case-control Study*
       • Case-control study in the architecture of a prospective study
         • N# of disease-related events accumulated in a Cohort study
   • EXPOSURE --> OUTCOME

3. Case-control (Retrospective) Study

   • Strengths USEFUL
     • Comparison of 2 distinct population groups
       • Cases: disease or health outcome of interest VS. Controls: indvs. without the disease or with a related clinical condition
       • Examination of specific risks factor (s); Primary exposure variable
   • Types of bias
     • Survival bias (selective survival of the cases under study) and Recall bias, confounding factors
   • Examples
     • Association between coffee consumption and occurrence of acute myocardal infarction
   • Limitations
     • Odds ratio are used as proxies for Relative risk (RR) since Incidence rates of disease are NOT able to be calculated
   • OUTCOME --> EXPOSURE

4. Case-crossover Study

   • Strengths USEFUL
     • Assessment of the role of possible precipitating factors involve in infectious and chronic diseases
     • Evaluate brief exposure with transient effects on acute health outcomes
     • Each case serves as it's own control

5. Randomized Control Trials

   • Random assignment of a large sample to "exposure" or treatment status or placebo
   • Types of bias
     • Minimization of selection and confounding bias
   • Study staff and patients are "blind" to the intervention and outcome