Project 1 - learning EBM by doing an evidence-based edit

Interesting idea:

“A doctor who accesses the world wide web … to seek the answer to a question but does not find it there and has the facility to place material on the web … must place that answer on the web where the next one to ask the same question can find it.” (refererence)

• Example of a student's edit from last year: http://en.wikipedia.org/wiki/Heparin-induced_thrombocytopenia (see paragraph starting "According to systematic review...")
• Instructions on editing a website such as Wikipedia
• Template that will suggest wording for your edit

Goals

• Make two succinct summaries and clinical recommendations - one for diagnosis and one for treatment.
• In the process of doing this you will have to use what you have learned about searching, assessing articles, and numeracy.

Getting this done

• Work with student who has the most similar topic.
• Find a 'representative article' for your treatment question and a representative article for your diagnostic question
  • Use UpToDate, PIER, or Clinical Evidence. Then check for major missing article with SUMSearch or PubMed clinical queries.
  • For treatment, look for the most recent meta-analysis and any RCTs published more recently. Preferably an article that gives your the event rates in both the control and treatment groups.
  • For diagnosis, look for the most recent meta-analysis and any studies of sn/sp published more recently. Preferably an article that gives your the sensitivity, specificity (or LR+ and LR-), and prevelence.
• Use the template to help construct your edits.

Writing hints

• Write formally, eg no abbreviations unless you first define them, for example, "congestive heart failuire (CHF)".
• Put summaries before details - example (Upper GI Bleeding)
• Hyperlink to Medline references and to other web pages that provide additional detail
• Be conservative and just report facts. Avoid synthesing hypotheses unless you believe you are expert on the topic and can subtantiate (with references) your hypothesis .
• No plagerism or copyright violations