Observational Studies/Epidemiology

  • Observational studies include: cohort; case-control; and cross sectional studies
  • They are non-randomised & non-experimental
  • Cohort study
    • Prospective 
    • Used to determine aetiology & natural history of disease
    • Define cohort –> assess risk factors to follow –> follow forward in time
    • Used to calculate risk ratio
    • Risk ratio indicates increased/decreased risk of disease associated with factor of interest
Time sequence can be reviewedCostly 
Provides info on wide range of disease outcomesIf outcome is rare – large sample sizes are required
Incidence/risk of disease can be measuredLong follow up time – increased patient drop out
Can gather info on range of risk factorsDisease outcomes/aetiology may change over time
Can be used to study rare exposure
Changes over time can be studies
Reduced recall bias & selection bias
  • Case control study
    • Retrospective
    • Review cases & controls to establish risk factors
    • Cases & controls can be matched to use participants who are similar in each group
    • Good for rare diseases
    • Used to calculate odds ratio 
    • Odds ratio indicates the odds of disease in individuals exposed and unexposed to a factor
Quick, cheap, easy to performIncreased recall bias
Good for rare diseasesNot suitable for rare exposures
Range of risk factors can be studied
No loss to follow up

Risk ratio & odd ratio are similar when prevalence of disease is rare

  • Cross sectional
    • Observe subject group once & review relationships between observed variables –> prevalence

Summary measures of disease

  • Cancer registration
    • Continuous & systematic collection of data on all cancer patients in a defined population
    • Held by Public Health England
    • Cancer registry will analyse, interpret & report the data periodically
    • Collected from: death certificates; hospitals; healthcare facilities/labs in the area
    • Data collected inc: ID; baselines characteristics; type & stage of cancer; treatment received; progress; survival; history
    • Used for: healthcare planning & monitoring & aiding research

Screening Tests

  • Sensitivity ability to detect disease correctly in those with it
  • Specificity – ability to identify no disease correctly in those who do not have it
  • Sensitivity & specificity are constant regardless of how prevalent the disease it
  • Importance of each may depends on the context of the test….
    • Screening tests in primary care – prioritise sensitivity – ability to correctly identify disease
    • Diagnostic test in secondary care – prioritise specificity – ability to exclude those without disease
  • Positive predictive value – proportion of test positive results that are true positives
  • Negative predictive value – proportion of test negative results in those that are true negatives
  • PPV & NPV will depend on tests ability to detect disease & underlying population
    • Increased prevalence – increased PPV
    • Increased prevalence – decreased NPV

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s