Skip to Main Content
Responsive image
Libraries Research Guides

Evidence Based Practice

Clinical Question Types

Etiology - determining if there was an association between an exposure and a disease or condition (also called "harm")
Prognosis - prediction of the clinical course or the natural history of a disease or condition
Diagnosis - using a tool to arrive at a diagnosis of a disease or condition
Treatment - intervention for therapy (including adverse effects studies), prevention, rehabilitation, quality improvement, or continuing medical education (also called "therapy")
Clinical Prediction Guide - prediction of some aspect of a disease or condition
Qualitative - how people feel or experience certain situations

(Adapted from Wilczynski, N. L.et al. (2005))

Study Design and Types of Evidence

When dealing with EBP, you will often hear about the “best” evidence. This means two things:

  1. Not all evidence is created equal.
  2. Not all levels of evidence are available for every question.

Imagine you had a patient with recurring, painful acne. What kind of research would you seek to help deliver a treatment: a systematic review of several randomized controlled trials of treatments for this condition, a single case study of a patient experiencing similar symptoms, or your colleague’s third cousin, who once mentioned that “chocolate causes acne”? 

These are all forms of evidence—even the last one!—but something like a systematic review, which amasses data from numerous studies to create generalizable results, will provide the broadest view of the evidence on your patient’s condition and help you devise an evidence-based treatment plan.

Pyramid of Evidence

(Image from CFCF / CC BY-SA)

However, sometimes the very high levels of evidence aren’t available. As you seek higher levels of evidence, they become less common, as shown in the pyramid above. With an especially narrow topic, or one about a new phenomenon, a single case study may be the best you can find. Additionally, for some types of questions—etiology or harm, for instance—it may not be ethically possible for a randomized controlled trial to exist.


Resources:

chat loading...