PPHC: L6 Observational Studies I

What are the 2 types of observational studies?

Descriptive and analytic

What are 3 examples of descriptive studies?

Case report, case series, and cross-sectional

What are 3 examples of analytic studies?

Cohort, case-control, and cross-sectional

Observational studies

• Individuals are followed ('“observed”) in … settings

• No … assignment by researcher

real-world, intervention/exposure

Descriptive observational studies

• Describes … of an … (i.e. disease)

• Goal: examine … (of exposure or outcome)

• … evaluate an intervention

• Start: … hypothesis

• End: … hypothesis 

occurrence, outcome, patterns, do not, no, possible

Analytic observational studies

• Goal: evaluate … (between exposure and outcome)

• … evaluate an intervention (not assigned by researcher)

• Start: … hypothesis

• End: … hypothesis 

relationship, do, defined, confirm or reject

Case …

• Detailed presentation of 1 case (‘n of 1’)

• What they do

  • Report a new or … condition

  • Describe previously … disease

  • Show unexpected new … effect

  • Report … events

• What they don’t do

  • Measure disease … (describe prevalent disease)

  • Identify … factors

  • Identify … of disease

reports, unique, undescribed, therapeutic, adverse, incidence, risk, cause

Case …

• Detailed presentation of 2 or more cases

• What they do

  • Report a new or unique condition

  • Describe previously undescribed disease

  • Show unexpected new therapeutic effect

  • Report adverse events

• What they don’t do

  • Measure disease incidence (describe prevalent disease)

  • Identify risk factors

  • Identify cause of disease

series

… study is an observational design that surveys exposures and/or outcomes at a … (‘snapshot’)

• Can be analytic (i.e. measure association between exposure and disease)

  • Temporal sequence of exposure and outcome … to determine (don’t know which occurred first) → main limitation (compared to cohort and case-control studies)

• Can be descriptive (i.e. focus on exposure or disease)

  • Measures prevalence not incidence

Cross-sectional, single point in time, impossible

… observational studies describe relationship between exposure and outcome

• Disease can be the exposure OR outcome

• Can evaluate intervention, though not assigned by researcher

Analytic

Analyzing analytic cross-sectional study … measure … of an outcome because we don’t know when outcome occurs

• In analytical cross-sectional study, exposure and outcome are assessed at the same time

• So relevant concept is …

cannot, risk, odds

The … of an event can be defined as the ratio of the number of ways the event can occur to the number of ways the event cannot occur 

odds

What is the formula for odds?

odds = event/(1-event)

We are betting on a horse named Epi who has a 60% probability of winning the race. What are the odds that Epi will win the race?

1.5

What are the 2 measures of association?

Odds ratio (OR) and relative risk/risk ratio (RR)

What is the formula for odds ratio for analytic cross-sectional study?

Odds = odds of outcome among exposed/odds of outcome among unexposed = (a/b)/(c/d) = ad/bc

• OR … 1 = association (exposed group have higher odds of outcome compared to unexposed group)

• Or … 1 = no association

• Or … 1 = inverse association (exposed group have lower odds of outcome compared to unexposed group

>, =, <

OR > 1

• % … = (OR - 1) x 100

• If OR = 1.58 then % … = (1.58 - 1) x 100 = 58% higher odds

OR < 1

• % … = (1 - OR) x 100

• If OR = 0.25 then % … = (1 -0.25) x 100 = 75% lower odds

increase, increase, decrease, decrease

“Big Picture” for analytic epidemiology: is there a relationship (…)

association