This template is a guide'

1
This template is a guide.

• Notes for your preparation
• Please work on all slides before we meet.
• The most difficult slides are the evidence table
  at the end.
• You may add slides not included, or delete some
  not needed.
• Replace italicized text with your content and
  delete italicized notes.
• There are some additional notes to guide you on
  some slides.

2
General points

• Avoid blocks of text here (and on every slide)
• Remove unnecessary words
• Try to keep font at least 24
• I can help resize images

3
Patient presentation

• This is bad example
• A 31-year-old man presented to the emergency
  department with pain in the left shoulder. He had
  tripped over the shoulder strap of his backpack
  earlier in the day and noted immediate severe
  pain around his left shoulder, without
  paresthesias or neck pain. Physical examination
  revealed bony point tenderness over the humeral
  head and limited range of motion due to pain
  crepitus was present.

4
Patient presentation

• This is a good example
• A 31-year-old man c/o left shoulder pain after
  tripping earlier in the day
• Immediate, severe pain around his shoulder
• Denies paresthesias or neck pain
• Physical examination
• Bony point tenderness over the humeral head
• Crepitus was present
• Limited range of motion due to pain

5
Clinical question
On this slide, phrase your question as it is
asked not as addressed by your article which
may be different.

• Among patients with septic shock, how well do
  corticosteroids, as compared to placebo, reduce
  mortality
• P
• patients with septic shock
• I
• corticosteroids
• If your question is more specific, please add the
  following
• C
• placebo
• O
• mortality

6
Background

• Giving background beyond would is in the next
  slide is optional. If you do add more background,
  please limit to one slide of salient points.

7
What is the current prevailing opinion?

• UpToDate
• UpToDate says
• PIER
• PIER says
• Optional other sources such as the most recent
  relevant practice guideline or a review article
  from a high impact journal such as NEJM, JAMA,
  Lancet, Annals.

8
Finding an answer

• Search terms
• steroids and sepsis
• Website(s) used
• Consider using references from UpToDate, PIER,
  PubMed, OVID, or http//sumsearch.uthscsa.edu,
  etc. Note that SUMSearch will simultaneously help
  you find recent reviews or guidelines to guide
  you to the current dogma, while sumsearch will
  also help you find an article for your review.

9
Todays article

• First author . Title. Journal name. Year. (PMID
  )
• Overview of methods
• Please do not use narrative review articles.
• What type of article is this (FOR HELP, SEE
  LEVELS OF EVIDENCE AND PICO TABLE ON THE BLUE
  CARD)?
• Funding

10
Todays article
On this slide, phrase your question as it is
addressed by your article. Note this might be
different than the way you originally phrased if
you could not find a perfect article.

• P What is the patients/population/problem being
  studied?
• If your article is a practice guideline, you may
  only be able to specify the P.
• I What is the intervention or test or causal
  factor being studied?
• C What is the intervention/test being compared
  to?
• O What is the outcome of interest in the study?

11
Methods patients

• Add anything you want about the patients beyond
  what was in the prior slide.

12
Methods intervention

• Add anything you want about the intervention
  beyond what was in the prior slide.
• If this was a study of a dx text, change the
  title of this slide to test

13
Methods comparison

• Add anything you want about the comparison.
• If your study is not about an intervention, you
  may delete this slide.

14
Methods outcome

• Add anything you want about how the outcome in
  the study
• How was the outcome measured
• How did the authors define outcomes
• When was it measured

15
Methods analysis
If you are reviewing an meta-analysis, use this
page for guidance.

• Heterogeneity
• Did the authors measure, report and deal with
  heterogeneity of results? Or did they pool apples
  and oranges
• Publication bias
• Did the authors mention a funnel plot or file
  drawer analysis of or measure of publication
  bias?
• Method of combining studies
• Conventional meta-analysis
• Individual patient meta-analysis

16
Methods analysis
If you are reviewing an RCT, use this page for
guidance.

• Sample size calculation
• Did the authors provide an estimate of the
  projected study size and a rationale for the
  size?
• Missing data
• Did the authors state how missing data is
  handled? If your study is an RCT, the authors
  should mention how an Intention to treat
  analysis and how they do it
• Interim analyses
• Did the authors plan interim analyses in order to
  stop the trial early if needed?
• See http//en.citizendium.org/wiki/RctInterim_ana
  lysis_-_stopping_trials_early for more
  information.

17
Methods analysis
If you are reviewing an dx test, use this page
for guidance.

• Sample size calculation
• Did the authors provide an estimate of the
  projected study size and a rationale for the
  size?
• Missing data
• Did the authors state how missing data is
  handled?
• Impact
• Did the authors simply report measures of
  accuracy, or also report impact?

18
Results patient flow

• Here copy/paste the Consort diagram for the
  study. If one is not present, please provide the
  number of patients recruited and how many
  completed the study.

19
Results primary outcome

• See calculator at http//clinical.uthscsa.edu/calc
  
• If this is a treatment question
• please put NNT if available, else show what ever
  you can find.
• Also try to show the event rate from the control
  group here.
• OR if this is a diagnosis question, show
• sensitivities, specificities
• pretest prevalences

20
Results secondary analyses

• This slide is optional.
• Sometimes you should show secondary outcomes
• Sometimes you should show subgroup analyses.
• However, be careful about over-interpreting these.

21
Results adverse effects

• This slide is optional.
• Are there important adverse effects that should
  be mentioned?
• Can you calculate a number needed to harm? See
  calculator at http//clinical.uthscsa.edu/calc
• Are these patients squeeky clean and less likely
  to have adverse effects than patients in the
  community?

22
Todays article strengths and limitations

• Use our critical appraisal cheat sheet (the blue
  card with help for assessing the article) The
  card is at http//medinformatics.uthscsa.edu/journ
  alclub/appraising.shtml
• Strengths
• Put strengths here
• Limitations
• Put limitations here

Use the blue card cheat sheet to help you with
this (http//medinformatics.uthscsa.edu/journalclu
b/appraising.shtml).
23
Other studies of this topic
You already have chosen the one study that best
supports your topic, now pick the one or
sometimes two studies that best refute your idea.
Comparing the studies in an evidence table will
help illuminate problems in the studies. Your
column headings will vary using various pieces of
PICO. Click here for very good examples.
24
Contribution

• Put here what, if anything, this article adds
  over the prevailing dogma as represented in
  whatever is the best current content from a
  recent high impact review article, practice
  guideline, UpToDate, or PIER.
• So, what did this article teach you clinically
  beyond what you read in UpToDate?
• Any unusual EBM point learned from this?