EBM FINAL Flashcards
guidelines vs. protocols
o “guidelines are systematically developed statements to assist practitioner decisions about appropriate health care specific clinical circumstances”
o guidelines: allow room for professional judgment
o protocols: are explicit instructions used in specific, predictable circumstances and restrict judgment
takes away your professional judgment
what are some drawbacks of guidelines?
- reflect expert opinion, UNSOUND PRACTICE
- average rather than best practice
- inhibit INNOVATION and prevent individuality
- may not reflect LOCAL needs
- may not recognize PRIMARY care setting
- undesirable shifts in balance of power
- out of date guidelines might hold back implementation of NEW research
what are 3 ways to implement guidelines
- develop guidelines based on board medicine
- disseminate through educational intervention
- provide patient specific reminders at time of visit (Ex. yearly eye exams)
from book: internal development, specific educational intervention, and patient specific reminder
types of economical analyses
cost minimization
cost effectiveness
cost utility
cost benefit ratio
minimization: no outcome measure, compare brand name with generic (cheaper)
effectivness: compare two looking at ONE variable
cost utility: compare two looking at MULTIPLE variables
made up ratio!!
cost benefit ratio: MONEY:: buying new equipment ($$$)
how do we measure costs?
- questionaires (most common!)
2.Health state preference value: value which a healthy person places on a particular deterioration in health in a hypothetical situation - QALY: calculate: health state preference value x time patient is likely to be in that health state
.4. Cost consequences analysis: expresses outcomes in terms of natural units (e.g months of survival, legs amputated or babies going home
-allows individuals to assign their own values to differing interventions
-allows for health state preference values to change over time
continuous variable vs. discrete
continous: infinite variable
discrebe: whole
types of scales:
- ratio/interval
- ordinal
- nominal
- provides greatest amount of information (ratio:: weight)
- ordered
- gender, race, etc
distribution of sample pattern:
normal
skewed
normal:symmetrical, bell shaped distribution of values where the mean represent the highest point of the curve and whose spread is defined by the standard deviation
• Skewed: asymmetrical distribution of values
- transforming skewer values into a normal distribution is sometimes possible by taking the log of the original values
- mean, median, mode more helpful here!
types of statistics:
quantitative
qualitative
• Quantitative
o parametric test: underlying assumption is the sample population(s) has a normal distribution ; based on the MEANS of the samples
o non-parametric test: make the assumption the sample population(s) is skewed ; basis of MEDIANS
• Qualitative: used to make a relative comparison
o ex. Ranking from high to low or likert scale
regression
•Regression: a mathematical equation that allows the TARGET variable to be predicted from the INDEPENDENT variable. These relationships must be linear
o overall, their function is to describe the relationship between variables
o linear regression
relative strength of relationship
random chance
association
correlation
CAUSATION: GOLD STANDARD
basic assumptions for PEARSON’S R
- normal distribute data
- 2 variable are independent of each other
- single pair of measurements on successive subjects
- every r should be accompanied by p==due to chance
reasons not to reject null (you WANT to reject it!)
o 1. there is no different between the groups in reality
o 2. there were too few subject to demonstrate such a difference existed
remember: Power, sample size is important
o 3. Logical flaw is the assumption for the cut-offs values as significant or not arbitrary when the reality is that values fall on a continuum.
Confidence intervals
the smaller=_____
the more precise, wider confidence intervals due to size and standard deviation
quality of pre-appraisal reports (low to HIGH)
studies: original journal articles, NOT PRE-APPRAISED
synthesis: systemic reviews for specialized services
synposis
summaries: textbook like guidelines
systems: computerized decision support;; ex compulink
set null at ___ if dealing with ratios
set null at ___ if dealing with comparions
1; 0
8 poster principles
• 1. uncluttered delivery of your message within 5-10 minutes
• 2. use graphical elements to grab readers
o no centerfolds please- Dr. F with the jokes!
• 3. be aware of requirements for design by meeting
o submissions are generally online in order to get approved
• 4. If not otherwise specified follow the IMRAD or abstract format with references, acknowledgments and disclosures
o disclosure such as industry funded ex. Alcon
o remember the idea of transparency
• 5. should be able to read the title at 25ft
o use 72 point font in times new roman or arial (use the same font throughout setting off the headings)
• 6. body of the text should be visible from 4-6 feet
o use 20-25 point font, same style as the title
• 7. no more than two fonts and simple color scheme
• 8. Recommended fonts: times, Arial, Century Schoolbook
power point, effective elements
- use design templates
- standardize position, colors, and styles
- include only necessary information
- limit it to the essential information
- content should be self-evident→ should not require a lot of interpretation
- use contrasting colors
- be consistent with effects, transitions, and animations
- too many slides can lose an audience
text guidelines on powerpoint
- no more than 6 words per line
- no more than 6 lines per slide
- avoid long sentences
- key points with larger fonts
- font size 18-48pt (Dr.F doesn’t like to go below 24)
- should contrast with the background and not fade in
- avoid fancy fonts
- words in all uppercase are hard to read
- Avoid acronyms and abbrev.
- Limit punctuation marks!?$
costs
direct: treatment, indirect: transportation
intangible
what are some purposes of guidelines?
- to make EBM standards explicit AND accesible
- to make decision making more objective
- to provide a yardstick
- deliniate the division of labor
- educate about current best practice
- serve as a tool for external control
Sensitivity vs specificity
include good information vs. exclude unnecessary
ways to organize EBM information
- quality of original studies (high to low )
interventional: RCT/NRT
observational
descriptive- case studies - quality of pre-appraised studies
name different study designs
RCT: prospective, cohort(one disease): prospective; case-control:retrospective (association), cross-sectional case
