The ideal confirmation test will have both high sensitivity and high specificity.

Having a positive or negative test isn’t final. A positive or negative predictive value predicts the likelihood that your test result is actually true.


Note: Risk=Probability

Attributable risk:

Standard deviation:

Skewed distribution:

Test reliability:

Test validity:

Correlation Coefficient:

Confidence Interval:

Comparison of data

  1. “Chi-squared test: Used to compare percentages or proportions (nonnumerical or nominal data)
  2. T-test: Used to compare two means
  3. Analysis of variance (ANOVA): Used to compare three or more means.” From Crush Step 3


Incidence: (New cases of a disease) ÷ ( population at risk) in a given period of time

Prevalence: (Total cases of disease) ÷ ( population). Can be at single time point (“ point prevalence”) or over a period of time (“ period prevalence”)

Number needed to treat/ harm: # of pts that must be treated to prevent/ cause 1 pt to have the measured outcome; 1/( risk difference), e.g., 1/( 0.1) → NNT of 10



Types of studies in epidemiology

Case-Control Study

“A case-control study is designed to help determine if an exposure is associated with an outcome (i.e., disease or condition of interest). In theory, the case-control study can be described simply. First, identify the cases (a group known to have the outcome) and the controls (a group known to be free of the outcome). Then, look back in time to learn which subjects in each group had the exposure(s), comparing the frequency of the exposure in the case group to the control group.

By definition, a case-control study is always retrospective because it starts with an outcome then traces back to investigate exposures. When the subjects are enrolled in their respective groups, the outcome of each subject is already known by the investigator. This, and not the fact that the investigator usually makes use of previously collected data, is what makes case-control studies ‘retrospective’.” http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1706071/

A case-control study is also observational.

In summary, a case-control study is an observational, retrospective study that starts with a known outcome — disease or cases for the case group and no disease for the control group–and looks back in time to find differences in exposure or risk factors between the case and control group. A good example is comparing a group of patients with lung cancer (case group) with a control group without lung cancer to find out if smoking exposure is associated with lung cancer. Association in a case-control is measured with an odds ratio (not RR).

Cohort study

Like the case-control, a cohort is also an observational study. Unlike a case-control, cohort studies are usually forward-looking – that is, they are “prospective” studies, or planned in advance and carried out over a future period of time. The goal is to identify differences in outcome between groups characterized by exposure/ risk factor; e.g., patients who smoke & those who don’t smoke are followed over time to determine if lung cancer incidence is different between groups.

Cross-sectional study

Assess simultaneously for outcome & exposure at a single point in time. For example, How many people in telephone survey are smokers? How many have lung cancer? May use RR or OR.

Non-randomised controlled studies

Randomized control trial
– Allows for inferred causality
(exposure leads to outcome) rather than just association (exposure & outcome are somehow linked)

Systematic Review



Considerations in Study Review

Bias/ study design:
Selection bias:
Information bias:




  • Leon Gordis, Epidemiology, 5th Edition.
  • Pocket Primary Care
  • Adam Brochert, Crush Step 3
  • http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1706071/
  • http://www.aafp.org/fpm/2004/0500/p47.html (A Simple Method for Evaluating the Clinical Literature)
  • https://onlinecourses.science.psu.edu/stat507/node/71 (Sensitivity, Specificity, Positive Predictive Value, and Negative Predictive Value)
  • http://handbook.cochrane.org/chapter_9/9_2_2_4_measure_of_absolute_effect_the_risk_difference.htm
  • http://childhoodcancer.cochrane.org/non-randomised-controlled-study-nrs-designs