Measures of Disease Frequency
On this page
Direct answer
Incidence counts new cases arising in a population at risk over a defined period; prevalence counts all existing cases at a point or during a period of time. The two are linked by the approximation prevalence = incidence × average duration, which is why chronic diseases and better survival inflate prevalence without touching incidence. Rates carry a time dimension and a population base (numerator within denominator × multiplier such as 1,000), proportions have no time element, and ratios let numerator and denominator be separate populations. Comparing populations demands age standardisation — direct standardisation applies study rates to a standard population, indirect standardisation produces the standardised mortality ratio (SMR = observed/expected deaths). NEET-PG questions reward the formula, the prevalence–incidence relationship, and clean definitions of IMR, NMR and MMR.
What you must remember
- Incidence proportion (risk): new cases ÷ population at risk over a period; incidence rate (person-time): new cases ÷ total person-time at risk, used when follow-up differs between subjects.
- Prevalence: point versus period; a measure of disease burden, not risk; numerator includes old and new cases.
- Prevalence = incidence × duration: longer survival or shorter duration (cure, rapid death) moves prevalence in predictable directions.
- Attack rate: an incidence proportion used in outbreaks (cases among the exposed population); secondary attack rate = secondary cases among susceptible contacts ÷ (total contacts − primary cases) × 100.
- Rate vs proportion vs ratio: rate needs a time unit; proportion's numerator is part of its denominator; ratio compares two distinct quantities (sex ratio, maternal mortality ratio).
- Standardisation: direct method when population-specific rates are known; indirect method (SMR) when they are not — SMR above 1 means more deaths than expected.
- Core indicators: IMR = infant deaths per 1,000 live births (first year); NMR = deaths in the first 28 completed days per 1,000 live births; MMR = maternal deaths per 100,000 live births; CBR and CDR per 1,000 mid-year population.
Worked example worth internalising
A township of 10,000 adults is followed for one year; 250 develop hypertension, and of the 500 who already had it, 15 die of stroke. Incidence of hypertension is 250/10,000 = 2.5%; prevalence at year-end is roughly (500 − 15 + 250)/10,000 ≈ 7.4% — incident and prevalent cases answering different questions. If a second survey counts 30,000 person-years of observation because people joined and left midway, the incidence rate becomes 250/30,000 = 8.3 per 1,000 person-years. Now compare this township with a retirement colony where 40% of residents are over 65: crude rates will mislead, so apply the township's age-specific rates to a standard population (direct standardisation) to get comparable adjusted rates; if only total deaths and the standard mortality structure are available, compute expected deaths and divide observed by expected to get the SMR — 1.3 means 30% excess mortality after age adjustment.
The same arithmetic governs programmatic indicators: an infant mortality rate of about 28 per 1,000 (India, Sample Registration System, recent years) is a probability-like proportion expressed per 1,000 live births, while the crude death rate divides all deaths by the entire mid-year population. Keeping the denominator straight — live births for IMR and NMR, live births in the denominator but deaths per 100,000 for MMR — settles most one-mark questions before they are asked.
Where students slip
Prevalence gets quoted as if it measured risk: it does not, because it depends on duration; a hepatitis B carrier state has high prevalence precisely because it lasts years. The second slip is numerator–denominator discipline in MMR, which is expressed per 100,000 live births while IMR is per 1,000 — an examiner's classic matching exercise. Third, "rate" is used loosely for proportions; strictly, a rate implies time (incidence rate, growth rate), whereas the "attack rate" of an outbreak and the IMR are proportions wearing rate-name uniforms, an accepted epidemiological convention worth knowing rather than fighting.
Frequently asked questions
How are incidence and prevalence related mathematically?
Prevalence approximates incidence multiplied by average duration of disease, so chronicity and improved survival raise prevalence while rapid cure lowers it.
When is person-time incidence preferred over cumulative incidence?
When subjects are observed for different durations — staggered recruitment, losses, late entry — the incidence rate per person-time weights each subject's actual time at risk.
What is the secondary attack rate used for?
It measures spread among susceptible contacts of primary cases within one incubation period, judging both infectiousness and the success of outbreak control measures.
How does indirect standardisation differ from direct?
Indirect standardisation applies standard population rates to the study population's age structure to generate expected events, yielding the SMR; direct standardisation does the reverse, adjusting the study's own rates to a standard population.
What denominators define IMR and MMR in India?
IMR is infant deaths per 1,000 live births in the first year of life; MMR is pregnancy-related maternal deaths per 100,000 live births, both reported by the SRS.