Community Medicine and Public Health: measures, designs and screening arithmetic
Community medicine items are the most reliably computable in the paper. Expect a two-by-two table, a study design to classify, or a prevention level to assign. Epidemiology, biostatistics, screening, outbreak investigation and the immunisation schedule carry the marks.
Core rules
- Incidence counts new cases over a period and measures risk; prevalence counts all existing cases at a point and is driven by incidence and duration.
- Relative risk comes from cohort studies; the odds ratio approximates it in case-control studies and is the only measure available there.
- Attributable risk is incidence in exposed minus incidence in unexposed, and describes the excess caused by exposure.
- Levels of prevention: primordial and primary before disease, secondary at the asymptomatic detectable stage, tertiary to limit disability.
- Sensitivity and specificity are properties of the test; positive predictive value depends on prevalence.
Worked example
In a population of 1000, a test is applied and 100 people truly have disease. Of these, 90 test positive. Of the 900 without disease, 90 test positive. Sensitivity = 90/100 = 90%. Specificity = 810/900 = 90%. Positive predictive value = 90/(90 + 90) = 50%. Now apply the same test in a low-prevalence clinic where only 10 of 1000 have disease, keeping sensitivity and specificity at 90%: positives are 9 true and 99 false, so PPV falls to 9/108, about 8%. The test has not changed; the population has.
Traps that cost marks
- Quoting PPV as if it were a fixed test property.
- Confusing relative risk with odds ratio when the outcome is common.
- Describing screening as primary prevention; it is secondary.
- Confusing attack rate in an outbreak with point prevalence.