Academic station Flashcards
Impact factor
Measure of frequency that an article in a journal has been cited in a particular year - surrogate indicator for importance of a journal
Calculated - total number of articles (over 2yrs) divided by sum of citations (over 2 yrs
Levels of evidence
Oxford Centre of Evidence Based Medicine classification
1a - Meta-analysis for RCT
1b- RCT
2a SR - Well designed controlled study (no randomisation) (Cohort)
2b- At least 1 well designed quasi-experimental study
2c - Epidemiological studies
3-a-SR of case-control studies
3b-well designed non experimental descriptive study (comparative, correlation or case)
4- Case Series
5-Committee /expert opinion
Grades of recommendation
A- Consistent Level 1 studies
B - Consistent level 2 or 3 studies or extrapolations for level 1
c- Level 4 studies or extrapolation for level 2-3
D Level 5 evidence or troubling inconsistent or inconclusion studies
Phases of RCT
0- Testing low dose of treatment - no harm
1- Finding best dose of treatment and side effects
2- Checking best dose of treatment type and how well it works
3- Comparing New Tx to standard Tx or placebo
4- Long term benefits
Power of a study
The power of a study is the probability that the study will detect a difference between the 2 groups given a set P value and sample size
Related to type 2 error -
Randomisation
The main types of randomisation are Simple, block and stratified
Simple- similar to tossing a coin - does not take into account confounders
Blocked - way of ensuring comparable sample sizes in each group (usually set at 1:1). Can lease to deduction of treatment allocation
Stratified - used to match certain characteristics which ,ay affect the results.
Type 1 and 2 Error
Type 1 - Rejecting the null hypothesis when there is no difference (Alpha is probability of making type 1 error)
Type 2 - Failing to reject the null hypothesis when a difference exists (false negative) - Beta is probs of making a type 2 error
ROC curve
Plots sensitivity and 1-specificity
Used for a single test or individual test
Measure of Hetrogeneity
Q - measures a specific question of whether the differences between the effect sizes of individual studies are Statistically significant
I2 - quantifies the degree of difference between effect sizes of individual studies
Propensity Score matched comparison
Propensity score matching is a statistical analysis to match groups according to certain variables
Tried to limit confounding factor when assessing statistical differences
Evaluating Cohort studies quality
Assess using the STROBE guidelines from the EQUATOR network website
Regression Analysis
A statistical analysis to estimate relationships between a dependent and an independent variable.
Nested case control study
Group of patients with the case of interest are controlled to patients without the disease and risk factors for case examined.
Nested is where the control group is pre-defined and can help in reducing selection bias and facilitating appropriate analysis and avoiding confounding factors
95% Confidence interval
Probability that the true value of the odds ration (or outcome) lies between the upper and lower values given.
P valvue
The P value is the probability that for a given statistical test the results are at least as extreme as the observed result.
aka
The likelihood of there being no difference between intervention is low (<5%) but typically we used it to mean that their is a difference
