Part 1 obgyn notes Sri Lanka
    NOTES for part 1
    /
    statistics
    /
    Studies

    Studies

    Home

    3 views

    Home

    All pages

    Pages I own

    CLASSIFICATION OF STUDY TYPES (MEDICAL RESEARCH)

    At the highest level, all studies are divided into:

    1. Observational studies – researcher does not intervene
    2. Experimental (interventional) studies – researcher assigns intervention

    1. OBSERVATIONAL STUDIES

    The investigator observes exposure and outcome as they naturally occur.

    A. Descriptive Studies

    (Describe disease occurrence – no comparison group)

    Used to answer:

    👉 Who, where, when?

    ❌ Not why

    1. Case Report

    • Detailed report of one unusual patient
    • Often first signal of:
      • New disease
      • New side effect
      • Rare presentation

    ✔ Hypothesis-generating

    ❌ No comparison, no causation

    Example: First report of thalidomide-related limb defects

    2. Case Series

    • Description of multiple similar cases
    • No control group

    ✔ Identifies patterns

    ❌ Cannot estimate risk or association

    3. Ecological Study

    • Unit of analysis = population, not individual
    • Uses aggregate data

    ✔ Cheap, quick

    ❌ Ecological fallacy (group association ≠ individual risk)

    Example: Country-level alcohol intake vs liver cirrhosis rates

    B. Analytical Observational Studies

    (Include comparison groups → test hypotheses)

    2. CASE–CONTROL STUDY

    Core logic

    • Start with outcome
    • Look backward for exposure

    Structure

    • Cases → disease present
    • Controls → disease absent

    Direction

    • Outcome → Exposure (usually retrospective)

    Measure

    • Odds Ratio (OR)

    Best for

    • Rare diseases
    • Diseases with long latency

    Limitations

    • Recall bias
    • Selection bias
    • No incidence

    3. COHORT STUDY

    Core logic

    • Start with exposure
    • Follow forward for outcome

    Types

    • Prospective cohort
    • Retrospective cohort (records used, but logic is forward)

    Measures

    • Incidence
    • Relative Risk (RR)
    • Attributable Risk (AR)

    Best for

    • Rare exposures
    • Studying multiple outcomes

    Limitations

    • Expensive
    • Long duration
    • Loss to follow-up

    4. CROSS-SECTIONAL STUDY

    Core logic

    • Exposure and outcome measured at the same time

    What it measures

    • Prevalence

    Strengths

    • Quick
    • Cheap
    • Useful for health planning

    Weakness

    • Cannot establish temporality
    • Cannot infer causation

    Memory lock:

    📸 Snapshot in time

    2. EXPERIMENTAL (INTERVENTIONAL) STUDIES

    The researcher actively assigns an intervention.

    5. RANDOMIZED CONTROLLED TRIAL (RCT)

    Core features

    • Participants randomly allocated to:
      • Intervention group
      • Control/placebo group

    Gold standard for

    • Causality
    • Treatment efficacy

    Bias control methods

    • Randomization
    • Blinding (single / double / triple)
    • Allocation concealment

    Measures

    • Relative Risk
    • Absolute Risk Reduction
    • Number Needed to Treat (NNT)

    Limitations

    • Expensive
    • Ethical constraints
    • Not always generalizable

    6. CLINICAL TRIAL (NON-RANDOMIZED)

    • Intervention assigned
    • No randomization

    ❌ More bias than RCT

    ✔ Used when randomization unethical

    7. FIELD TRIAL

    • Preventive interventions
    • Conducted in healthy populations

    Example: Vaccine trials

    8. COMMUNITY TRIAL

    • Unit of intervention = community
    • Not individual

    Example: Water fluoridation program

    3. SECONDARY / EVIDENCE-SYNTHESIS STUDIES

    9. SYSTEMATIC REVIEW

    • Structured, predefined method
    • Identifies, appraises, synthesizes evidence

    ✔ High-quality evidence

    ❌ Depends on included studies

    10. META-ANALYSIS

    • Statistical pooling of results
    • Produces a summary effect size

    ✔ Highest level of evidence

    ❌ Garbage in → garbage out

    4. SPECIAL / OTHER STUDY TYPES

    11. DIAGNOSTIC ACCURACY STUDIES

    • Evaluate tests using:
      • Sensitivity
      • Specificity
      • Predictive values
      • Likelihood ratios

    12. PROGNOSTIC STUDIES

    • Follow patients with disease
    • Assess outcomes (survival, complications)

    13. QUALITATIVE STUDIES

    • Explore experiences, beliefs, behaviors
    • Methods:
      • Interviews
      • Focus groups
      • Thematic analysis

    ✔ Depth, meaning

    ❌ Not numerical

    14. MIXED-METHODS STUDIES

    • Combine quantitative + qualitative
    • Used in health systems research

    FINAL EXAM MEMORY LOCK

    • Rare disease → Case–control
    • Rare exposure → Cohort
    • Causality → RCT
    • Prevalence → Cross-sectional
    • Highest evidence → Meta-analysis

    🧠 MASTER TABLE — CLASSIFICATION OF STUDY TYPES (MEDICAL RESEARCH)

    TABLE 1. High-Level Classification of All Studies

    Main Category
    Definition
    Key Feature
    Examples
    Observational
    Researcher does not intervene
    Exposure & outcome occur naturally
    Case–control, Cohort
    Experimental (Interventional)
    Researcher assigns intervention
    Active manipulation
    RCT, Field trial
    Secondary / Evidence synthesis
    Uses existing studies
    Summarizes evidence
    Systematic review
    Special / Other
    Specific research purpose
    Diagnostic / Prognostic / Qualitative
    Test accuracy

    TABLE 2. OBSERVATIONAL STUDIES — OVERVIEW

    Subtype
    Main Purpose
    Comparison Group
    Key Question Answered
    Descriptive
    Describe disease occurrence
    ❌ No
    Who? Where? When?
    Analytical
    Test hypotheses
    ✔ Yes
    Is exposure associated with outcome?

    TABLE 3. DESCRIPTIVE OBSERVATIONAL STUDIES

    Study Type
    Unit of Analysis
    Core Features
    Strengths
    Limitations
    Classic Example
    Case Report
    Single patient
    Unusual/novel finding
    Hypothesis-generating
    No comparison, no causation
    Thalidomide limb defects
    Case Series
    Group of similar cases
    Pattern recognition
    Identifies trends
    No risk/association
    New disease clusters
    Ecological Study
    Population (aggregate data)
    Group-level exposure & outcome
    Cheap, quick
    Ecological fallacy
    Alcohol intake vs cirrhosis

    TABLE 4. ANALYTICAL OBSERVATIONAL STUDIES — CORE COMPARISON

    Feature
    Case–Control Study
    Cohort Study
    Cross-Sectional Study
    Starting point
    Outcome
    Exposure
    Exposure + outcome
    Direction
    Outcome → Exposure
    Exposure → Outcome
    Snapshot
    Time frame
    Usually retrospective
    Prospective or retrospective
    Single time point
    Main measure
    Odds Ratio (OR)
    RR, AR, Incidence
    Prevalence
    Temporality
    ❌ Not clear
    ✔ Clear
    ❌ Not clear
    Causality
    ❌
    ❌ (stronger than CC)
    ❌

    TABLE 5. CASE–CONTROL STUDY (FULL EXAM DETAIL)

    Aspect
    Details
    Core logic
    Start with disease, look back for exposure
    Groups
    Cases (disease) vs Controls (no disease)
    Best for
    Rare diseases, long latency
    Measure
    Odds Ratio
    Cannot measure
    Incidence
    Major biases
    Recall bias, Selection bias
    Cost & duration
    Cheap, quick

    TABLE 6. COHORT STUDY (FULL EXAM DETAIL)

    Aspect
    Details
    Core logic
    Start with exposure, follow to outcome
    Types
    Prospective, Retrospective
    Measures
    Incidence, RR, AR
    Best for
    Rare exposures, multiple outcomes
    Strengths
    Temporal relationship
    Limitations
    Expensive, long duration, loss to follow-up

    TABLE 7. CROSS-SECTIONAL STUDY

    Aspect
    Details
    Core logic
    Exposure & outcome measured simultaneously
    Main measure
    Prevalence
    Strengths
    Quick, cheap, health planning
    Weaknesses
    No temporality, no causation
    Memory hook
    📸 Snapshot in time

    TABLE 8. EXPERIMENTAL (INTERVENTIONAL) STUDIES

    Study Type
    Unit
    Key Feature
    Example
    RCT
    Individual
    Randomization + control
    Drug trials
    Non-randomized clinical trial
    Individual
    No randomization
    Ethical constraints
    Field trial
    Healthy individuals
    Preventive intervention
    Vaccine trials
    Community trial
    Community
    Group intervention
    Fluoridation

    TABLE 9. RANDOMIZED CONTROLLED TRIAL (RCT) — GOLD STANDARD

    Component
    Details
    Allocation
    Random
    Control
    Placebo / standard care
    Bias control
    Randomization, blinding, allocation concealment
    Measures
    RR, ARR, NNT
    Strength
    Best for causality
    Limitations
    Expensive, ethical issues, generalizability

    TABLE 10. SECONDARY / EVIDENCE-SYNTHESIS STUDIES

    Study Type
    Core Feature
    Strength
    Limitation
    Systematic Review
    Structured literature synthesis
    High-quality evidence
    Depends on included studies
    Meta-analysis
    Statistical pooling
    Highest level of evidence
    Garbage in → garbage out

    TABLE 11. SPECIAL / OTHER STUDY TYPES

    Study Type
    Purpose
    Key Measures / Methods
    Diagnostic accuracy
    Test performance
    Sensitivity, Specificity, LR
    Prognostic study
    Disease outcomes
    Survival, complications
    Qualitative study
    Experiences & beliefs
    Interviews, focus groups
    Mixed-methods
    Quant + Qual
    Health systems research

    🧠 FINAL EXAM MEMORY LOCK (TABLE)

    Exam Scenario
    Best Study Design
    Rare disease
    Case–control
    Rare exposure
    Cohort
    Causality
    RCT
    Prevalence
    Cross-sectional
    Highest evidence
    Meta-analysis