Critical Appraisal of Cohort Studies

Introduction

• Focus on critical appraisal of cohort studies using the CASP approach.
• Importance of observational studies when RCTs are not feasible or ethical.
• Observational studies examine risk factors or exposures without manipulating them.

Learning Outcomes

• Introduce cohort studies and their significance in healthcare research.
• Teach critical appraisal using the CASP checklist.
• Discuss risk ratio calculation and interpretation.
• End with a quiz for knowledge testing.

What is a Cohort Study?

• Strongest design of observational studies.
• Researchers identify participants without the outcome of interest and classify them by exposure status.
• Example: Study on smoking and lung cancer.
  • Identify non-lung cancer patients.
  • Classify them as smokers or non-smokers.
  • Follow participants over time to compare lung cancer rates.

Risk Ratio

• Definition: Ratio of incidence of disease in exposed vs. unexposed groups.
• Interpretation:
  • RR > 1: Exposure increases disease risk.
  • RR < 1: Exposure decreases disease risk.
  • RR = 1: No difference in risk.
• Example Calculation:
  • Exposed group risk (A/(A+B)) = 0.85
  • Unexposed group risk (C/(C+D)) = 0.05
  • Risk ratio = 0.85 / 0.05 = 17
  • Conclusion: Smokers are 17 times more likely to develop lung cancer than non-smokers.

Importance of Critical Appraisal

• Quality of cohort studies can vary.
• Use the CASP checklist for appraisal.

CASP Cohort Studies Checklist

1. Focused Question:
   • Clear research question regarding lithium and dementia risk.
2. Cohort Recruitment:
   • Assess for selection bias.
   • Population: Over 50 with bipolar disorder from eight states (Medicaid insured).
3. Measurement Bias:
   • Exposure assessed via health administrative data.
   • Outcome measurement (dementia) likely objective.
4. Confounding Factors:
   • Account for confounders like age, gender, ethnicity.
   • Authors adjusted models for potential confounders.
5. Follow-Up Duration:
   • Max follow-up time: 3 years.
   • Important to minimize dropout rates.
6. Hazard Ratio:
   • Measures association over time.
   • Example: Hazard ratio of 0.77 (protective effect of lithium).
7. Confidence Intervals:
   • Indicate precision of results.
   • Narrower intervals = higher precision.
8. Believability of Results:
   • Consider potential biases, chance effects, and biological plausibility.
9. Local Applicability:
   • Assess how results may apply in different contexts.

Conclusion

• The upcoming module will cover case-control studies with a similar appraisal format.
• Quiz available for knowledge testing on this module's topics.