Chapter 79: Principles of Statistical Methods for Research in Regional Anesthesia

NECESSARY BASICS

Anesthesiologists treat thousands of patients in their busy clinical careers; they also save thousands of lives. However, anesthesiologists also have a place in research where their work is far reaching and can impact the lives of millions of people worldwide. This chapter is dedicated to those anesthesiologists who actively undertake research and who keep abreast of the research literature in order to better serve their patients.

The chapter begins with some very basic principles of statistics that form the foundation for those methods most frequently used in regional anesthesia research. These principles, albeit more theoretic and seemingly abstract, are included in the hopes of offering more than a cookbook approach to the statistical procedures. Several statistical packages are available “to crunch data,” yet a minimal number of formulas and calculations are included to illustrate what happens to data when submitted to a statistical program. It is hoped that this chapter will foster more effective dialogue between anesthesiologist and statistician. By the end of the chapter, the reader should have a better understanding of what a statistician needs to know about studies and why this information is crucial to reaching valid research conclusions.

What Is Statistics?

Sir Ronald A. Fisher (1890–1962), the father of modern statistics, considered the science of statistics to be mathematics applied to observational data: “Statistics may be regarded as (i) the study of populations, (ii) as the study of variation, (iii) as the study of methods of the reduction of data.”¹ His definition has three important implications for research:

1. Investigators would like to apply their research findings to vast populations, but it is seldom feasible to study an entire population, so they must study samples from it.

2. Each sample studied will be slightly different; i.e., there is variation among samples. Thus, there will be differences among samples even from studies that have used the same design and methods, and even among samples taken from the same population.

3. Investigators summarize and test the data from their study sample in order to reach reasonable conclusions about the parent population that they can communicate to colleagues, to journal editors, and to the public.

Types of Data

The types of data collected in a study determine the types of statistical analyses. Table 79–1 describes the three types of data, what they represent, their typical level of measurement, and their properties.²