Biostatistics and Epidemiological Methods

Relative Risk = ratio of rate of disease in those exposed to that rate among those not exposed.

•  e.g. RR(lung CA) = rate of lung CA in smokers / rate of lung CA in nonsmokers

Attributable Risk = rate of disease in exposed – rate of disease in non-exposed

Controlled Clinical Trial

•  prospective study
• split participants into 2 groups (usually randomly) – give each group different Rx
• assess results and compare between groups.

Prospective Observational Study Design (cohort study design)

•  answers the question of whether a particular factor is a causal factor for a given condition.
• difference between this and controlled clinical trial is that the 2 groups are not selected by the investigator (they are selected based on presence or absence of the factor in question)
• it is prospective
• the groups are usually not comparable in several aspects.

REtrospective Study Design

•  AKA case-control study design
• identify 2 groups (one affected, the other not) that are identical in all other ways
• look back into past to try to identify causal factors.

Cross-Sectional Surveys

•  does not involve the time factor.
• “snapshot” at one moment in time.
• can determine association of 2 factors, but not causation

Advantages of Retrospective Design over Prospective

•  prospective inefficient in diseases with low incidence rates.  (lots of $$, time, energy spent on following up cases that don’t develop disease in question)
• retrospective study has ability to yield results quickly (prospective usually requires extended follow-up)
• prospective studies suffer loss of patients during follow-up (distortion of results)

Advantages of PRospective over REtrospective

•  direct estimate of incidence rate
• problem of historical data (recall bias)
•  cannot identify time sequence of factors.  (sometimes)