This page provides key concepts from this module. It is intended to be a synopsis of the content in the module.
Introduction to Quality Filtering and Evidence-Based Medicine and Health
Not all studies are created equally.
Quality filtering is a process that sifts the more substantial studies from the less informative ones.
Librarians perform quality filtering when they search and choose articles that best answer the information needs of the requester.
Clinical guidelines recommend a particular protocol for prevention or treatment.
Randomized clinical trials are considered the best or the "gold standard"; however, it is not often feasible or ethical to conduct randomized clinical trials.
Not everyone agrees that evidence-based medicine, clinical practice guidelines, and meta-analysis add anything new.
Several different types of studies are used in health services research.
In most health services research questions, investigators need data from huge groups of people.
The sample provides answers that are applied to the larger group in which everyone is really interested, the population.
Researchers try to reduce the variation that occurs when a sample is selected because large differences between the sample and the population make the study results less accurate and less applicable to groups outside the study.
Determining the best sample size and composition is critical in health research.
It is important to read the descriptions of random sampling in the Methods section of an article.
Researchers often assign the people in their sample into groups. These groups may be ready made or not. In experimental studies, some groups may be exposed to different treatments, screening or diets.
It is important that the people in each group must be alike except for the factor being tested.
Choosing people for comparison is a crucial feature in study design.
Be sure to examine the table containing information on the study's groups to make sure that they are similar in socioeconomic and other related factors.
Accurate and complete measurements are essential; unfortunately, there are are many ways to mismeasure variables when people are the object of investigation!
Recall bias is common in studies that compare remembered experiences of ill and healthy people.
Measurement errors occur under certain circumstances.
Masking (also called "blinding") helps to minimize measurement errors in clinical trials.
It is difficult to measure people accurately.
Researchers perform analysis to identify three major characteristics of analysis
Strength of the association: how strong was the association between variables?
Statistical significance: what is the likelihood of getting the results from our sample if there was no relationship between variables in the larger population from which the sample came?
Adjustment: were the groups in the study different in any way that could affect the results?
The Analysis section of research papers is often difficult to understand without a statistical background. Ask for help when you need it from someone trained in statistical methods.
Interpretation concerns the conclusions made about the people in the study. Researchers assess the strength of the association between variables (indicated by the relative risk or other measures) and the cause and effect relationship between them.
Several factors support a causal relationship
The risk factor occurs more often in people with the specific outcome.
The risk factor being studied precedes the effect.
Changes in the risk factor produce the effect.
Other support for causation include
Strength of the association between the factors
Consistency of the association
Evaluating all of these factors help ascertain what type of relationship exists between the variables (the risk factors and outcome).
Extrapolation (to larger groups)
Readers of research studies must decide if the results can be used for groups or in amounts different from the study.
Beware of research articles that end with sweeping generalizations to vastly different groups of people. The data should not be overgeneralized!