unm fm journal club a new paradigm and example july 16, 2014
TRANSCRIPT
UNM FMJournal Club
A New Paradigm and ExampleJuly 16, 2014
Types of clubs
Our reality
Objectives of Journal Club Develop clinical question and identify
population, intervention, comparison, outcome
Search literature efficiently and effectively to find evidence to address question
Critically appraise available evidence and determine if it is valid and if it applies to your patient(s)/population
Gain experience presenting to group
Learn and practice effective teaching strategies
Journal Club
Intern
PGY2
Peanut Gallery
PGY3
PGY2 + PGY3
Each month, 1-2 PGY3’s and 1-2 PGY2’s will prepare journal club presentation which includes the following:
Develop question
Find article
Prepare presentation
Send materials to rest of group before day of presentation
Present and guide the group through critical appraisal
Well Built Clinical Question
P I C O One model for developing clinical
question
4 components P Patient/Population/Problem I Variable of Interest C Comparison/Control O Outcome
PICO + T: Type of Question
T=Type of Question
Different types of questions Intervention/Therapy Etiology/Harm Diagnosis Prognosis
Before finding an article that addresses your question, figure out what type of question it is.
Type of Study
Question Study Design
Intervention/Therapy
Systematic Review/Meta AnalysisRandomized (efficacy)Cohort study (effectiveness)
Etiology/Harm CohortCase ControlCase Series
Diagnostic Testing Cohort Cross-sectional
Prognosis CohortCase ControlCase Series
Different types of questions are addressed by different types of study designs
How to pick an article
How to find evidence to answer question
Databases of primary research PUBMED/MEDLINE
www.ncbi.nlm.nih.gov/pubmed/ Free
EMBASE Proprietary
Cochrane Library Data base of systematic reviews www.thecochranelibrary.com/view/0/
index.html
Other sources
ACP Journal Club annals.org/journalclub.aspx
Evidence Updates plus.mcmaster.ca/EvidenceUpdates/
Essential Evidence Plus www.essentialevidenceplus.com/
Searching
Use PICO to formulate search strategy
Start with Patient Problem and Variable of Interest. Enter into search field Example: Obesity AND diabetes AND
bariatric surgery
Limiting by Language Humans Study Design
For more details, see hand out
Preparing for the presentation Team: question, article
Consult with TP/SW
Send abstract and important tables/figures out to group at least 24 hours prior to Wednesday of Journal club
Presentation of question and article
Engaging and guiding the rest of the group in the critical appraisal
Resources Journal club schedule on WIKI Guide for presentation/appraisal Ideas for group activities
Presentation Question
Background-why is this issue important Epidemiology Impact on patient/population
Study Purpose Study Design Population Variable of interest Comparison Outcome Results
Critical Appraisal
Determine what the study is about
Examine research and judge trustworthiness, value and relevance
Determine if study is done in a way that make findings reliable
Make sense of the results, in context of decision making for individual patient
Form/guide to be made available
Challenge: Figuring out ways to engage the group
Example of how it will
work
Question
31 year old female presents with history of vaginal discharge that has fishy odor. Speculum exam done, sample of discharge collected.
Is it enough to find positive amine odor to make dx
P = non-pregnant women with vaginal discharge
I = amine odor as diagnostic test for BV
C = Gram Stain
O = Accurate diagnosis of BV
Type of Question = diagnostic
Type of Study to look for: Cohort, cross sectional
Background
Bacterial vaginosis most common vaginal infection among women
Prevalence is 9-37%; 22-50% in symptomatic women
In pregnant women, BV increases risk of: Preterm delivery PPROM Chorioamnionitis Postpartum endometritis
In all women, increases risk of PID
Diagnosis of BV
Current gold standard for diagnosis is gram stain.
Other reliable diagnostic: DNA probe of vaginal d/c 3 of 4 criteria (Amsel)
Presence of thin homogenous discharge
Vaginal pH >4.5 Positive “whiff” test or release of
amine odor with addition of base Clue cells on saline wet preparation
Is amine odor enough to make diagnosis
Article
Citation: Gutman et al.
Study purpose: Can diagnosis of BV be made accurately using only 2 of clinical criteria
Study design: prospective cohort study
Population: 269 women undergoing speculum exam for any reason Excluded if large amount of vaginal
bleeding on exam
Variable of interest
Odor of discharge collected Specimen collected
By 2-4th year OB-GYN residents, research nurse or PI. All trained
Did not know BV status of patient Describes specimen collection and
preparation-KOH Assessment of smell
Comparison
Gram stain for everyone Separate slide Sent to outside hospital lab Standardized 0-10 point score Nugent criteria, score of 7+ gold
standard
Outcome
Accurate diagnosis of BV by amine odor only vs. gold standard
Sensitivity, specificity
Results
Prevalence of BV in study population was 38.7%
Table 1: characteristics of women with an without BV
Table 2: Sensitivity/Specificity Sensitivity: 67% (CI: 57-76%) Specificity: 93% (CI: 88-97%)
Did not provide PPV, NPV or LR but I can compute them myself
+ -
+ 67%N=70
7%N=12
- 33%N=34
93%N=153
BV diagnosed by gram stain
Presence Amine odor
100%N=104
100%N=165
0.67*104 = 700.33*104 = 340.07*165= 120.93*165= 153
N + test = 70 + 12 = 82N - test = 34 + 153 = 187
PPV = 70/82 = 85%NPV = 153/187=82%
LR+ = 0.67/0.07 = 9.6LR- = 0.33/0.93 = 0.35
Interpretation of Results
Using amine odor alone vs. gram stain 85% of women who had BV were
positive for amine odor 93% of women who did not have BV
were negative for amine odor In other words, only 7% of women
without BV had amine odor and would have be incorrectly diagnosed with (and maybe treated for) BV (false positives)
For my patient, if she has an odor, she has an 85% risk of having BV; if she has no odor; she still has 16% of having BV (False negative)
Critical Appraisal:
Is the question studied by these investigators similar enough to my question?
Is the population studied similar to the population I am interested in (recall: non-pregnant women with vaginal discharge)
What are the flaws in the way the data were collected, analyzed? How could this affect the validity of the results. Blinding? Gold standard for everyone? Reliability of data collection?
Critical Appraisal: continued
Do I understand the results, do I interpret them the same as the authors?
Were the results compelling enough? Clinical significance demonstrated?
Are the results applicable to my patient?
Can I use this test in my practice?
Will I change my practice based on this evidence?
Questions? Thanks for your attention