COURSE CONTENT

Descriptive statistics. Variable, types of variables, statistical tables, graphical methods, numerical descriptive measures of central tendency and variability, coefficient of variation, Gaussian distribution, transformations, normal values, evaluation of laboratory findings (sensitivity, specificity, positive and negative predictive value), ROC curve (Receiver Operator Curve).

Parameter estimation. Point estimation, confidence interval and standard error in estimation (i) the mean, variance and percentage in a population and (ii) the difference of the means and percentages and the ratio of the variances in two populations.

Statistical tests. Null and alternative hypothesis, type I and II error, power of a test, p-value of a test, the concept of statistical significance, hypothesis testing (i) about the mean, variance and percentage in a population and (ii) for the difference of means and percentages and the ratio of variances in two populations, pairwise observations.

Analysis of variance. Analysis of variance for independent samples, analysis of variance table, multiple comparisons.

Qualitative data analysis. Contingency tables, χ² test as a criterion of association and goodness of fit of qualitative characteristics, simple logistic regression model, relative risk (RR), relative odds ratio (OR).

Statistical correlation and dependence. Pearson’s r correlation coefficient, least squares method, prediction, simple linear regression model, coefficient of determination.

Non-parametric controls. Advantages and disadvantages of non-parametric tests, Kolmogorov-Smirnov one-sample test, Wilcoxon signed rank tests, Mann-Whitney U, Kruskal-Wallis, Spearman’s ρ correlation coefficient.

Survival analysis. General, censored data, survival tables, survival function, hazard function, survival curves, Kaplan-Meier method, comparisons between survival curves, simple Cox regression model.

The laboratory exercises of the Medical Statistics course include the analysis of medical data with the help of computers and appropriate statistical packages.