Thanks to the creative faculty of Middlesex Health System, Inc. for permission to publish this web site.
Welcome
to the
Middlesex
Health System
Center
for Evidence Based Practice
Thanks to the creative faculty of Middlesex Health System, Inc. for permission to publish this web site.
Welcome
to the
Middlesex
Health System
Center
for Evidence Based Practice
(Developed by Katherine Schneider, MD, MPhil, MHS Czarina of EBP, kschneider@midhosp.org)
(Maintained by Elisabeth Righter, MD, FAAFP) Last updated 5/10/01.
What is Evidence Based Practice? (Also known as Evidence Based Medicine, to non-interdisciplinary thinkers)
EBP or EBM is a process which aims to improve the ability of clinicians to make well informed clinical decisions for and with their patients:
1. Identifying knowledge needs in clinical encounters
2. Translating needs into answerable questions
3. Searching for quantitative answers
4. Appraising the evidence
5. Integrating the evidence into the clinical situation
Even seasoned clinicians generate about 5 knowledge needs for every inpatient encounter and 2 needs for every 3 outpatients. The purpose of this web site is to bring the EBM process closer to the point of care by collating the most useful introductory information and searching tools into one up-to-date location that will be easily accessible during patient care and teaching encounters.
Any underlined blue phrases may be clicked on to link to other sites of interest. The colored buttons at the bottom of each page can help you jump around the site.
A
well written overview of Evidence Based MedicineA
variety of definitions of Evidence Based Medicine
Glossary
of Evidence Based Medicine terms
Once you have gotten past the hurdle of identifying your knowledge need, the biggest challenge is learning to format your question into a structure that gives you the best chance of finding the quantitative answer you need. This process is for "foreground" questions, such as "In patients with nonvalvular atrial fibrillation, does treatment with warfarin compared to aspirin improve mortality from all causes". This approach should not be used for "background questions", such as "What is hypertrophic pulmonary osteoarthropathy?" which is best answered through traditional means, e.g. a textbook or quick reference manual (also available now on many internet sites such as Dynamed or MDConsult).
The exact structure of the clinical question will vary somewhat depending on whether it is about a therapy, diagnostic test, etiology, prognosis, cost effectiveness, clinical exam finding, etc. However...
...the well built clinical question can be broken down into four parts:
1.
Patient population
2.
Intervention (e.g. treatment or test)
3.
Comparison (e.g. placebo or gold standard)
4.
Outcome
The handy mnemonic "PICO" might help at first, especially helping learners to rephrase questions.
An example:
Focusing
Clinical Questions
The following is a list, by no means exhaustive, of some of the most useful web based resources that I have found for finding answers to questions. Try the secondary sources first, as they are more likely to give you a quick, pre-digested answer that you can use at the point of care, as opposed to Medline which is better when you have more time to pick through abstracts and evaluate which are valid and relevant.
Click directly on any underlined phrase to get to that site.
Background Knowledge
Searchable
Databases
Secondary Sources (advantage: already synthesized for you)
If you are not always able to connect to the internet and would like portability of a wide collection of secondary sources, consider purchasing Inforetriever or Best Evidence (CD-ROM) which can be loaded directly onto your desktop or laptop computer. Inforetriever is also available for your pocketpc handheld computer. The combination of these contains all of the above, and is well worth the money you'll spend to update them.
Medline
(If you are good at structuring your questions but bad at Medine searching like me, you'll love this one because you can just type in your sentence and the program figures out how to translate it into Medline-ese)
Other
InterpretOnce you have found your "answer", the next step is to understand its validity and relevance to your patient. If it scores low on validity (i.e. it's a lousy study) or relevance (doesn't apply well to your situation), you should probably not waste time going further. If you did your search within secondary sources (e.g. Cochrane, POEMs), you can have some confidence that the validity has already been assessed.
A major goal of the "EBM Process" is to provide quantitative answers so that you and your patients can make better informed decisions about diagnostic tests, treatments, etc., rather than giving you a black and white or cookbook answer. To this end, you should familiarize yourself with some of the basic tools of EBM. You probably already are familiar with sensitivity and specificity, but how about "numbers needed to treat" (NNT), "numbers needed to harm" (NNH), likelihood ratio (LR), and pre-test probability. These were designed to be intuitively understood measures that can summarize the bottom line effect of a treatment or test.
Use of these measures can be mastered with a little practice. The following sites can help you with examples and explanations.
Teaching Resources forCritical Appraisal