STA424 BIOSTATISTICS & EPIDEMIOLOGY
3 Hours/Week, 3 Credits
Biostatistics overview: roots, development and nature of discipline, current focuses and challenges. Basic quantities: lifetime distribution, survival function, hazard function, interrelationships, mean residual life function, median life time, censoring, truncation, type – I and type – II censoring, random censoring. Parametric methods: likelihood construction for censored and truncated data, inference procedure for exponential and Weibull distributions under complete and censored data, accelerated failure time models. Nonparametric methods: estimation of survival function, hazard function; reduced sample method, product limit method, actuarial method, estimation and standard error. Semiparametric methods: cox proportional hazard model – its assumption, parameter estimation, explanation of the hazard ratio, model diagnostics. Epidemiologic concept: overview of important historical development of epidemiology, basic terminology and principles used in epidemiology. Sources of data of community health: census, vital statistics and morbidity data. Measure of disease frequency: incidence, prevalence, sensitivity and specificity. Estimation of risk and rate. Measure of effect and measures of association: measure of effect, measures of association, standard measures, prevalence ratio, relative risk, attributable risk, odds ratio, standard errors of estimates for different types studies, McNemar test. Study designs: case-control, cohort, prospective, retrospective, longitudinal studies. Matching: purpose and effect of matching, matching in case-control studies, matching in cohort studies. Bias, confounding, and effect modification: concept of confounding – measured and unmeasured confounding, effect modification, marginal structural models, propensity scores methods, instrumental variables.