Track 15 · Applied Statistics for Professionals · Lesson 12 of 12
Rates and proportions in public health
Calculate and interpret prevalence, incidence rates per population and per person-time, screening accuracy and predictive values, and directly standardised rates that compare populations with different case mixes.
A regional health committee sees that one district has twice as many people living with a chronic condition as another, and that one hospital's crude recovery rate is lower than its neighbour's. Before moving budgets, it needs to know: are these counts or rates, new cases or existing ones, and are the populations being compared actually comparable?
Public-health statistics are mostly rates: a count of events divided by the size of the population that could have experienced them, often over a stated period. Counts alone mislead because populations differ in size; rates alone mislead when populations differ in age, severity or other risk factors. This lesson builds the main measures and the standard fix for the second problem.
Included with a Formula Quest membership
It has 8 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Distinguish incidence from prevalence and link them through duration
- Calculate incidence rates per 1 000 or 100 000 person-years
- Compute sensitivity, specificity and predictive values, and explain why PPV depends on prevalence
- Directly standardise rates to compare populations with different age or severity mixes
- State the limits of standardised rates and screening statistics