Statistics Flashcards
(28 cards)
Define Absolute Risk
The likelihood of an event occurring in a specific group of people
The number of events in a group, divided by the number of people in that group
What is the absolute risk control (ARC)?
Absolute risk in the control group (i.e. number of events in control arm divided by total numbers in control arm)
What is the absolute risk treatment (ART)?
Absolute risk in the treatment group (i.e. number of events in treatment arm divided by total numbers in treatment arm)
Absolute risk reduction
ARC - ART
Number needed to treat
1 / Absolute risk reduction
Relative Risk
Relative risk is a ratio of the probability of an event occurring in the exposed group versus the probability of the event occurring in the non-exposed group
ART / ARC
Relative Risk Reduction
(ARC - ART) / ARC or 1 - RR
Sensitivity
TP / (TP + FN)
If someone has a disease,
how often will the test be
positive
Specificity
TN / (TN + FP)
If someone doesn’t have a
disease, how often will the
test be negative
Positive Predictive Value
TP / (TP + FP)
Probability that if the test result
is positive, the individual does
actually have the disease
Negative Predictive Value
TN / (FN + TN)
Probability that if the test result
is negative, the individual
doesn’t actually have the
disease
What is Prevalence?
The proportion of a population found to have a condition. e.g. 1 in 10,000.
What is Incidence?
The probability of occurrence of a given medical condition in a population within a specified period of time. E.g. 1 in 10,000 per year.
It can be thought of as the rate of proliferation in the population.
Odds Ratio
Odds of the event : non-event
OR = 1 is no difference
What are the main problems associated with meta-analysis?
Publication bias
Language bias
Replication bias
Requires some homogeneity between studies
What is type 1 (α) error
Incorrect rejection of the null hypothesis (that no significant difference exists) - aka finding a difference when none exists
What is a type 2 error?
Incorrect acceptance of the null hypothesis - aka finding no difference when one exists
What are observational studies?
The investigatordoes not assign exposures, but there can be comparison groups
With comparison groups =>analytical study
- Cohort study
- Case-control study
- Cross-sectional study
Without comparison groups =>descriptive study
- Case report
- Case series
What is the hierarchy of medical evidence?
1a) Systematic review of RCTs
1b) Individual well designed RCT
1c) All or none study
2a) Systematic review of Cohort studies
2b) Individual cohort study
3a) Systematic review of case-control studies
3b) Individual case-control study
4) Case-series
5) Expert opinion, bench research, first principles
What is the grading of recommendation for scientific literature?
A - Consistent level 1 studies
B - Consistent level 2 or 3 studies, or extrapolations from level 1
C - Level 4 studies, or extrapolations from level 2 or 3
D - Level 5 evidence, or troubling/inconsistent studies at any level
Describe the process required for drug approval
Pre-clinical studies
Phase 0
- First in human
- Micro dosing
- Preliminary data on pharmacokinetics
Phase 1
- Healthy volunteers
- Dose-ranging
Phase 2
- Larger groups
- Efficacy testing
Phase 3 => at least 2 required to market
- RCTs to assess efficacy
Phase 4
- Post-marketing surveillance
What was the OPTIMISE (2014) trial investigating?
In high-risk patients undergoing gastrointestinal surgery, does a cardiac output-guided haemodynamic therapy algorithm compared to clinician-guided standard therapy reduce death and major morbidity within 30 days?
What was the intervention protocol in OPTIMISE?
Colloid and dopexamine administration from induction of anaesthesia to 6 hours after end of surgery guided by LiDCO monitor.
250ml bolus colloid over 5 minutes – positive response if rise in stroke volume of at least 10% sustained for 20 minutes.
What was the outcome from OPTIMISE?
There was a reduction in 30 day composite adverse major outcomes, but not statistically significant (p = 0.07).
No difference in other end points
