Learn — Epidemiology, biostatistics & screening

2 notes · 19 questions on the topic

Epidemiology and study design

Epidemiology is the study of the distribution and determinants of health-related states and events in specified populations, and the application of that study to the control of health problems. Study design is the frame chosen to answer a particular question, and it matters because the design decides which measure of association can be calculated at all, so naming the design correctly is half of every calculation question in the paper.

Descriptive epidemiology answers who, where and when through case reports, case series, cross-sectional or prevalence surveys and ecological studies; it generates hypotheses about causes but can never test them, because exposure and outcome are measured at the same moment.
A case-control study starts from the disease and looks backwards at exposure; it yields the odds ratio, is quick, cheap and the design of choice for a rare disease or a long latent period, and it is vulnerable to recall bias and to the selection of controls.
A cohort study starts from the exposure and follows people forwards to the outcome; it yields incidence, and therefore relative risk and attributable risk, is the design of choice for a rare exposure, establishes temporality clearly, and is expensive and vulnerable to loss to follow-up.
The randomised controlled trial is the strongest design for causation: randomisation removes confounding both known and unknown, blinding removes observer and subject bias, and the analysis is by intention to treat; a community trial randomises whole villages or schools rather than individuals.
Incidence is new cases arising in a defined period divided by the population at risk, and prevalence is all existing cases at a point divided by the population examined; prevalence is approximately incidence multiplied by duration, so a disease that is quickly cured or quickly fatal has a low prevalence however common it is.
Relative risk equals the incidence in the exposed divided by the incidence in the unexposed and measures the strength of an association, while attributable risk is the incidence in the exposed minus the incidence in the unexposed and measures how much disease would be prevented by removing the exposure, which is the number a public health programme needs.
In a two-by-two table with cells a, b, c and d, the odds ratio is ad divided by bc and approximates the relative risk when the disease is rare; sensitivity is a divided by the total with disease and a highly sensitive test is used to rule out, specificity is d divided by the total without disease and a highly specific test is used to rule in, and the positive predictive value falls as prevalence falls, which is why mass screening of a low-prevalence population generates mostly false positives.
Before an association is called causal, chance, bias and confounding must be excluded, and then the Bradford Hill considerations applied: temporality, strength, dose-response, consistency, biological plausibility, coherence, specificity and reversibility; temporality is the only one that is indispensable.

Socio-economic indicators

Socio-economic indicators are the measures of income, education, employment, housing, sanitation, sex ratio and dependency that describe the conditions in which people are born, live and work, and they are grouped with mortality, morbidity, nutritional and health-service indicators in the standard list of health indicators. They matter because they are the causes behind the causes: a hamlet's female literacy rate and sex ratio predict its home deliveries, its undernourished children and its school dropouts more reliably than any clinical measurement made there.

The Human Development Index, published by UNDP since 1990, is the geometric mean of three dimensions — life expectancy at birth, education (mean years of schooling of adults and expected years of schooling of children) and gross national income per head at purchasing power parity — and runs from 0 to 1. The Human Development Report 2025, using 2023 data, placed India at 0.685 and rank 130 of 193 countries, and the report edition must always be quoted with the figure because both the value and the rank change with every edition.
The Physical Quality of Life Index of Morris D. Morris combines infant mortality, life expectancy at age one and literacy, each scaled from 0 to 100 and averaged, and it deliberately leaves income out; contrasting it with the HDI, which puts income back in, answers most indicator questions.
Related composites to name are the inequality-adjusted HDI, the Gender Development Index, the Gender Inequality Index (built on maternal mortality, adolescent birth rate, seats in parliament, secondary education and labour force participation) and the Multidimensional Poverty Index, whose Indian version by NITI Aayog uses twelve indicators across health, education and standard of living.
Sex ratio in India is the number of females per 1,000 males: Census 2011 recorded 943 overall and a child sex ratio of 918 in the 0 to 6 year age group, and NFHS-5 (2019-21) reported a sex ratio at birth of 929 female births per 1,000 male births for the five years before the survey.
Literacy from Census 2011 was 73 per cent overall, 80.9 per cent for males and 64.6 per cent for females, and the male-female gap is itself the indicator that predicts antenatal registration, immunisation and age at marriage in a block.
Individual socio-economic status is graded with the modified Kuppuswamy scale in urban families, which scores education, occupation and total monthly family income and is revised each year with the consumer price index, and with the modified B. G. Prasad scale in rural families, which uses per capita monthly income and is likewise revised.
Other indicators the paper asks for are the dependency ratio, per capita income, the proportion below the poverty line, family size, housing and overcrowding, and access to a safe water supply and a sanitary latrine; the health indicators they sit beside include an infant mortality rate of 24 (SRS 2024) and a maternal mortality ratio of 87 per 100,000 live births (SRS 2022-24) for India in the current Sample Registration System bulletins, and a total fertility rate of 2.0 in NFHS-5, already below the replacement level of 2.1.

Free today: the 10-question daily quiz · the NORCET 11 recall paper · what ₹999 unlocks — once, no auto-renewal.