Statistics in biomedical research I

Overview

Statistics are increasingly prevalent in medical practice. Nowadays much concern is devoted in hospital utility statistics, audit, resource allocation, number of new cases (for example AIDS), and so on. Journals nad magazines for doctors are full of statistical material of this sort, as well as the finding of individual research studies. Statistical issues are implicit in all clinical practice when making diagnosis and choosing an appropriate treatment. It is increasingly common to see statistical result from research papers quoted in promotional materiál for drugs and other medical therapies.

The article focuses on particular steps of a research project – planning, design, data collection and data storage, data processing, data analysis, presentation, interpretation and publication of results. We deal with planning, design and data storage in the first part of Statistics in biomedical research. Particular attention is paid to the role of prior estimates of a sample size. Using examples it is showed a relationship among a mean difference, a variability, a sample size and a statistically significance in continuous data. It is demonstrated an association among an occurrence of a certain quality in two samples, sample sizes and a significance in qualitative data. It is explained the difference between clinical and statistical significance. There are practical advices how to store data correctly for statistical processing.

Keywords:
statistics – research project – planning – sample size estimation – data storage