This module introduces evidence-based optometry which involves learning how to ask answerable research questions, conduct literature searches on electronic databases and critically evaluate scientific publications. Epidemiology is defined followed by an explanation of how to decide upon which study design is most likely to answer the research question. The advantages and disadvantages of cross-sectional studies, case-control studies, cohort studies and randomized-controlled clinical trials are discussed in terms of their susceptibility to bias and other forms of data distortion.
Qualitative research, questionnaire studies, meta-analysis and decision analysis are introduced. Differences between discrete and continuous data are explained in the context of deciding whether to use parametric or non-parametric statistical tests to analyse results. This involves learning about the normal distribution and how to select the correct measures of central tendency (mean, mode and median) and dispersion (standard deviation, coefficient of variation, and standard error of the mean, confidence limits, range and interquartile range). The calculation of relative risks and odds ratios, together with their respective confidence limits, as a means of assessing disease risk in epidemiological studies is also explained. Evaluation of diagnostic tests is covered in terms of assessing discriminative ability (sensitivity, specificity, predictive values and likelihood ratios, ROC curves), validity and repeatability (Bland-Altman plots, coefficient of variation, kappa).
Five lectures are devoted to hypothesis testing and cover testing normality (Kolmogorov-Smirnov one-sample test), the comparison of two (t-tests, Mann-Whitney U test, Wilcoxon signed-rank test) or more (analysis of variance, Kruskal-Wallis test, Freidman test) groups, correlation (Pearson, Spearman and Kendall correlation coefficients), linear regression (slope, intercept, coefficient of determination) and analysis of frequencies (Chi-square goodness of fit and contingency tests, Yates’ correction, Fisher’s exact test, phi coefficient). Descriptions are provided for performing these epidemiological and statistical calculations using SPSS and Microsoft Excel.
Finally, clinical research ethics covers landmark documents, key ethical theories, laws protecting research participants, NHS ethics, University ethics and preparation of a research protocol and consent form for ethical approval.