What is Evidence Based Medicine?
“Evidence-based medicine is the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients.” See Sackett DL et al. http://www.cebm.net/
The practice of Evidence Based Medicine (EBM) involves making critical use of up-to-date medical research in order to make informed decisions with respect to the treatment and care of individual patients.
The procedure followed in EBM closely mirrors the statistical problem solving cycle (see below) and consists of the following stages:
- Posing a focused clinical question;
- Searching for the evidence;
- Appraising the evidence;
- Applying the evidence.
Statistics is an essential tool to aid the practice of EBM. It provides the relevant techniques required in order to analyse and interpret data from research and thus draw appropriate conclusions (inferences) from the results. Inferences are generally based on a sample of individuals drawn from the population of interest, since it is not usually feasible to study the whole population. Provided the sample is representative, these inferences will relate to the population of interest. An understanding of statistics also allows us to critically appraise the literature and apply the results of medical research to clinical decision-making.
The statistical problem solving cycle
Data are numbers in context and the goal of statistics is to get information from those data, usually through problem solving.
 |
Figure 1: The statistical problem solving cycle |
A procedure or paradigm for statistical problem solving and scientific enquiry is illustrated in the diagram. The dotted line means that, following discussion, the problem may need to be reformulated and at least one more iteration completed.
Population and sample: definitions
Population: the entire group of interest e.g. people who are obese.
Parameter: a particular characteristic of the population that we are interested in e.g. the population mean or proportion. This is usually unknown and so needs to be estimated from sample data.
Sample: a subset of the population of interest taken in order to estimate the population parameter. Measurements or values (observations) are taken from the sample in order to learn about the population. E.g. we could take a sample of people from a particular city in order to estimate the proportion of people in the city who are obese. It is important to ensure that the sample is a good and unbiased representation of the population of interest.
Statistic: a quantity calculated from the sample and used to estimate a population parameter e.g. the sample mean, or proportion.
Variable: in order to estimate the prevalence of obesity we would measure weight and height. These would be described as variables as their values differ from one individual to the next.
The table below provides some simple examples of population parameters with the corresponding sample statistics (estimate of the population parameter).
| Name | population parameter | sample statistic |
|---|---|---|
| Mean | $$\mu$$ |
$$\bar{x}$$ |
| Standard deviation | $$\sigma$$ |
$$s$$ |
| Proportion | $$\pi$$ |
$$p$$ |
| Table 1: Parameters and their corresponding sample statistics | ||