Psychiatrynet.eu

an independent selection of sites for psychiatrists, by psychiatrists

Evidence Based Medicine

Colophon


'Searching for the Evidence'top


General information and coursestop

  • Examples Number Needed to Treat

    Hidden in the www.cebm.utoronto.ca site (via the glossary) several examples of Number Needed to Treat can be seen. CRITERIA FOR INCLUSION OF EXAMPLES are: (i) random allocation of participants to comparison groups (ii) ≥ 80% follow-up of people enrolled (iii)outcome of clinical significance

  • Centre for Evidence Based Mental Health (Oxford, UK)

    Offers information about how to use EBM in in psychiatry. Also EBM activities and links to useful tools.

  • McMaster University Evidence-based Practice Center

    The Canadian counterpart to the British Centre for Evidence-Based Medicine (see above).

  • The Cochrane Collaboration

    The objective of the Cochrane Collaboration is to summarise all of the available evidence for the effectiveness of medical interventions by means of systematic literature reviews and to make these available for everyone by publication in the Cochrane library. There is also a Dutch language equivalent.

  • Bandolier

    A Webzine with lots of information about new developments in the field of EBM and reviews of important clinical articles (general medicine).


Journals critically appraising the evidencetop

  • Evidence Based Menthal Health

    The most important, clinically relevant, research in mental health. This quarterly digest was begun in 2001

  • Clinical Evidence

    BMJ Clinical Evidence books are regularly updated and give relevant EBM-based information about treatment decisions.

  • Health Evidence Bulletin Wales

    Act as signposts to the best evidence. Includes Mental Health National Service Framework (Wales) edition


Databases with systematic reviewstop


Databases with abstracts or free full text articlestop

  • Open access journals

    This service covers free, full text, quality controlled scientific and scholarly journals.

  • PubMed

    Medline-database from the National Library of Medicine (NLM).

  • PubMed: Clinical queries

    Deep link to the clinical queries in the Medline-database. Clinical queries are useful to categorize the search in diagnostic, therapeutic, prognostic or etiological categories. Search can be done sensitive and specific (default is specific).


Step 1: Formulate a questiontop

The first step in the process of evidence-based medicine is to formulate a clinical question.  It is easy to formulate many questions, but finding an answer is often much more difficult.  There are many subjects in daily practice which lead to a search question.  The most frequently asked questions relate to therapy, diagnosis, or prognosis.

In view of the fact that the formulated question is the guiding line for researching the literature, it is important to formulate a good, answerable question.  As a tool for formulating questions, the EBM group at Oxford developed the PICO system: Patient -- Intervention -- Comparison -- Outcome.  All of these elements together form the basis of an answerable clinical question.  An example of such a question is:

Does a treatment with risperidon reduce the number of relapses in patients with schizophrenia, as compared with haldol?

This question consist several components:

P: patients with schizophrenia;
I: risperidon;
C: haldol;
O: number of relapses,


Step 2: Searching for Evidencetop

The second step in searching for evidence is to translate the clinical question into a search strategy.
It is of course possible to find everything there is to know about a subject, but often there is not enough time to work one's way through the enormous quantity of information found. In daily practice it is often more useful to apply a more specific search strategy. You can devise a specific search strategy with the help of the following principles:
• Break down the clinical question into PICO concepts;
• Translate each concept into search terms (one or more terms per concept);
• Group the search terms for each concept with "OR";
• Then combine the concepts with "AND".
There are a great many types of databases one can search. It is useful to begin the search for systematic literature reviews or an evidence-based guideline. If this search comes up short, one can search further in databases of abstracts from peer-reviewed journals. Finally one can search directly for abstracts of medical articles in for example the Cochrane Library, Pubmed (Medline), EMBASE, Current Contents, and Psyclit.
The search strategy in these last four databases sometimes yields a great quantity of irrelevant articles. In that case it is useful to combine this search strategy with methodological filters from the problem statement (therapy, diagnosis, or prognosis).
There are also search assistants which step through the various databases in the sequence requested.


Sites for Step 2 - Search aidstop

  • Detailed analysis of PubMed Search results

    PubReMiner will query pubmed with your specified searchquery, get all abstracts and generate frequency tables. The first table will show you journals in which your query is published the most. The second table will show you the authors which are most active in the field of your query.The third table will show you words that have been used most in the title and abstract of the articles.

  • ClinicalTrials.gov

    The American National Institutes of Health (NIH) developed this site to inform patients, family members, and other interested parties about studies currently underway.  Also explains clearly what clinical trials involve.

  • SUM Search

    A useful tool to perform a systematic literature search of review articles about a particular subject in several databases simultaneously.  Also searches databases with clinical guidelines.


Step 3: Evaluating the evidencetop

The third step in evidence-based medicine is the evaluation of the literature found for validity and applicability.  Checklists have been developed in England as well as in the Netherlands to help the clinician evaluate the articles found.  Under the topic Sites for step 3 you will find links to English-language forms.


Sites for Step 3top

  • Overview

    Appraisal form developed by the Centre for Evidence Based Mental Health in Oxford for evaluating the medical literature(PDF).

  • Therapy

    Appraisal form developed by the Centre for Evidence Based Mental Health in Oxford for evaluating therapy studies (PDF).

  • Diagnosis

    Appraisal form developed by the Centre for Evidence Based Mental Health in Oxford for evaluating a study of diagnosis (PDF).

  • Prognosis

    Appraisal form developed by the Centre for Evidence Based Mental Health in Oxford for evaluating a study of prognosis (PDF).


Step 4: Applying the Evidencetop

The last step in the evidence-based medicine process is to apply the answer found to the clinical problem.  In this interpretation phase, clinical knowledge is indispensable, and it is important to weigh the advantages of a a particular solution against the disadvantages.
Once articles which may answer the question have been found, another question arises: "Do results of these research articles also apply to the patient whose situation engendered the question?" For example: What chance for recovery from depression within six weeks does the use of medication X offer a borderline personality disorder patient? Two studies suggest that the chance of recovery within  six weeks is 60%.  However one of the exclusion criteria in these studies was the presence a personality disorder. Thus the results are not directly applicable.  No studies seem to have specifically included patients with both borderline personality disorder and depression.  In such a case the articles found represent the best available evidence and the clinician’s decision is partially based on this.
The next question is then whether the chance of recovery in this patient outweighs the chance of side effects and the costs.  Sometimes it is useful to make a list of the pros and cons for the patient.  In general, an evaluation about the proposed treatment has to be made whether the proposed treatment offers more advantages than disadvantages for the patient in your consultation room.  In this, the experience and preferences of the physician play an important role.

Furthermore the question that is more relevant is why the results found are not applicable to the patient whose situation engendered the query. Are there important reasons not to generalize/extrapolate the results from the studies to your patient?

In conclusion, the most important factors in applying the evidence are:

  • the best available evidence;
  • the patient's situation (age, comorbidity, compliance, etc.) and prognosis (the natural development of the illness in this patient);
  • the preferences of the patient and doctor.

     

For more informationtop

Books

  • Sackett DL, Straus S, Richardson S, Rosenberg W, Haynes RB. Evidence-Based Medicine: How to Practice and Teach EBM (2nd Edition): London, Churchill Livingstone. 2000.
  • Ian K, Crombie. The pocket guide to critical appraisal: ISBN 072791099X., ISBN 072791099X.
  • Clinical Evidence. Book published twice a year with overviews of the newest evidence for various disorders.
  • Trish Greenhalgh, How to Read a Paper: The Basics of Evidence Based Medicine. 2nd ed. London: BMJ, 2001.