Statistics Flashcards
What is sensitivity and how is it calculated?
- How good is a test for the condition of interest.
- True positive rate.
- Negative result in high sensitivity test useful for rulng condition out.
- Sens = T+/(T+ + F-)
What is specificity and how is it calculated?
- How good is a test for the condition of interest?
- True negative rate.
- Positive result in highly specific test useful for ruling in disease.
- Spec = T-/(T- + F+)
What is a positive predictive value and how is it calculated?
- How good is the test?
- If the test is positive what is the chance of the patient having the disease?
- PPV = T+/(T+ + F+)
What is a negative predictive value
- How good it the test?
- If the test is negative what is the chance the patient does not have the disease?
- NPV = F-/(F- + T-)
What is a likelihood ratio and how is it calculated?
- Assess the value of performing a diagnostic test.
- Use sensitivity and specificity to determine whether test result usefully changes the probability that a condition exists.
- +LHR = sensitivity/(100% - specificity)
- -LHR = (100% - sensitivity)/specificity
What is incidence?
- Incidence is a rate –> cases per 5 years.
What is prevalence?
- Prevalence is a cross section in time.
What is relative risk and how is it calculated?
- Ratio of an outcome between 2 exposures.
- Probability of an event occurring in an exposed group compared to an unexposed group.
- RR = (a/a+b)/(c/c+d)
What is odds ratio and how is it calculated?
- Odds like the TAB.
- Event occurring:event not occurring.
- Does not take in to account total events.
- Useful when population risk unknown.
- With low prevalence OR ~ RR.
- OR = ad/cb
What is the absolute risk and how is it calculated?
- Real difference in absolute terms between 2 groups.
- AR = (c/c+d) - (a/a+b)
What is a type I error?
- a error
- False positive.
- Reject the null hypothesis when it is actually true.
- p-value quantifies risk of rejecting the false hypothesis.
What is a type II error?
- b error
- False negative.
- Fail to reject the null hypothesis when it should have been rejected.
- Greatest risk of making type II error is underpowering.
What determines the power of a test?
- Population size.
- Size of effect.
- Variance within populations.
What is phase 0 of a drug trial?
- PK and PD in humans.
- Small sub-therapeutic doses given.
What is phase 1 of a drug trial?
- Safety.
- Small number.
- Safe dosing and adverse affects.
What is phase 2 of a drug trial?
- Safety and efficacy.
- 2a dosing requirements.
- 2b efficacy.
- Also test toxicity.
What is phase 3 of a drug trial?
- Confirmation of safety and efficacy.
- Large numbers.
- Vs. placebo RCTs - often >1.
- Input from licensing committees.
What is phase 4 of a drug trial?
- Contiuned pharmacovigilance.
- New uses/populations.
- Paediatrics.
- Interactions.
What are case control studies used for?
- Used to estimate the odds (OR) of developing the studied condition/disease.
- Can determine if there is an associational relationship between condition and risk factor.
- Cannot establish causality.
What are randomised controlled trials used for?
- “Gold standard” in medical research.
- Best for answering questions about effectiveness of different therapies or interventions.
- Widely considered the gold standard for establishing causality.
What are cohort studies used for?
- Establish causation of a disease or evaluate the outcome/impact of treatment.
- Used when randomised controlled clinical trials are not possible.
