Evidence-based Practice Residential Workshop for Librarians – Day 2

Day 2 of the EBP Residential Workshop for Librarians started with breakfast in the dining hall and discussion with attendees (by this time, we were getting more comfortable and knowledgable about each other). As with the first day, the program was split into large group lectures and small groups for activities. Day 2 focused on critical appraisal of randomised control trials (RCTs) and systematic reviews (SRs).  The most fun thing we did was split into groups during the lecture period and play EBM Bingo  (after lunch I think) where the answers were on sheets and questions were read out. We placed Minties on the answers we had. The group I was in needed one more  – then Bingo was shouted out. After dinner, we played EBM Jeopardy (none of us Australians knew how to play this game and fewer knew that it was a game – I had heard of it but had no idea of the rules) which turned out to be the reverse of Bingo. I really liked this game because it made you think! Here are the points I took down for Day 2:

  • When doing critical appraisal, the sections with the information needed is usually in the Methods (randomisation, allocation, statistics, the PICO (or introduction/abstract)), and the Results (patient follow-up, flow charts, baseline characteristics). Cons of RCTs – generalisability and context
  • Critical appraisal hints – peer review (but consider the  BMJ test of peer reviewers – so many didn’t pick up deliberate mistakes!), does the study match the question, currency and conflicts of interest, and the risk of bias (results skewed away from the truth)
  • Critical appraisal checklists: RAMMBO (we use this at our intern journal clubs), FRISBE and CASP
  • Blocked randomisation – occurs in permuted blocks
  • Maintenance – groups treated equally throughout the study
  • Follow-up – tracking patients through the study
  • People can change groups (treatment to control) if adverse reactions to drugs occur (eg)
  • Objectivity – results measured by a machine
  • Outcomes –  dichotomous (binary – yes/no) or continous (scale)
  • Relative risk – risk of outcome in intervention group compared to control group whether intervention increased or decreased likihood of outcome. RR -= 1 (no difference), RR = <1 [less than 1] reduces risk of event, RR = >1 [more than 1] increases risk of event
  • Absolute Risk Reduction – Absolute differenct of outcome in 2 groups (always a minus calculation eg ARR risk in control – risk in treatment). If a is 2% and b is 1%, the AAR is 1%
  • Relative Risk Reduction – risk of event in the treatment group relative to the risk in control (a / calculation). Using the AAR above (1%), the RRR is 50%
  • Number Needed to Treat (NNT) – no. needed to treat in order to prevent 1 outcome
  • Assessing chance due to intervention or accident is called a p (probability due to chance) value. The lower it is, the less likely it is due to chance. <P0.05 is the accepted mid point. The lower it is means statistical significance
  • Confidence interval (CI) – range of values likely to include the real value 95% of the time ie there is is 95% chance that the true value is included. CI should not overlap null value.
  • The more subjects in the study = the narrower the CI
  • If there is not enough to satisfy quality of life (clinical significance), then statistical significance is not clinically important
  • Randomisation means that known and unknown factors are equally distributed and there is an equal chance of allocation to each group
  • Intention to Treat – keeping people in the groups to maintain randomisation
  • Follow-up – accounting for everyone at the end of the study. There should be no less than 20% lost to follow-up
  • Systematic reviews condense research evidence from a number of primary studies and considers the clinical question indepth
  • Critical appraisal hint – PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)
  • If you can draw a straight line through the vertical lines in a forrest plot, the studies are similar enough

There are 4 questions for librarians to ask themselves when looking at a systematic review – was there a sensible question (did the review have a focused question which is not too broad), was the search comprehensive and exhaustive, were the studies selected for inclusion of high methodological quality, and is the review reproducable.  The questions clincians should ask are to do with results and whether the information can assist in patient care (we ask these questions at intern journal clubs).

A storm front cooled things down considerably and I had to change back into jeans and a thick t (I had put a cotton dress on) – and the air con couldn’t be turned off or down (as far as I could tell). I played some pool with some librarians after EBM Jeopardy and I went pretty well (I sunk 2 balls), considering that I hadn’t played since I was about 8 or 9 (and I had forgotten the rules that my Dad taught me). Reading a novel and texting my husband ended the day.

Want to add your thoughts? Fire away!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s