Clinical trials

Question 1

What is a clinical trial

Answer 1

planned experiment on humans

Question 2

What is the role of randomisation

Answer 2

ensured balance and eliminates bias

Question 3

How is bias in treatment allocation reduced?

Answer 3

1. Block randomisation: assign A or B randomly
2. Stratification: done by centre - can be divided by important patient characteristics e.g male/female
3. Minimisation: adaptive stratification - calculates imbalance and allocates to maintain balance

Question 4

How is measurement bias avoided?

Answer 4

1. Blinding: single, double, triple
2. Endpoint selection
   
   • Objective/subjective
   • Consistent and repeatable
   • Primary/secondary/tertiary end points
3. Loss to follow up
   
   • Missing data
   • Different between groups
   • Intention to treat analysis

Question 5

What is the chance of false positive?

Answer 5

chance of obtaining a difference in trial if no real different between groups

alpha=0.05

Question 6

What is the chance of false negative?

Answer 6

Chance of obtaining no difference in trial if there was a real different between groups

beta=0.1-0.2

Question 7

What are factors that have to be taken into consideration for clinical trials?

Answer 7

1. Patient selection
2. Controls
3. Study size
4. Unbiased data collection
5. Specific design
6. Ethics - if appropriate to do study
   - Approved by independent research ethics committee
   - Adhere to international and governemtnal guidelines
7. Analysis

Question 8

What is the role of the Independent data monitoring committee:

Answer 8

• Research check progress during trial
• Unblind results to see any major difference inoutcome
• If large difference may suggest stopping trial

Question 9

What are studies of association?

Answer 9

1. Clinical observation: 
		○ Case
		○ Series 
2. Vital statistics: 
		○ Trends 
		○ Ecological l 
3. Animal experimention 

4. X sectional
   ○ Dont know what happened before but collecting information
   ○ Gives quick answer as to if there is something worth looking at

5. Cross over study: 
		○ Big number 
		○ Long duration
6. Case control 
7. Cohort 
8. Trials

Question 10

What does the study type depend on?

Answer 10

how much time you have

• Less time means more imprecise results
• During trials you always make assumptions of causality
• Eligibility criteria and inclusion/exclusion criteria

Question 11

What should studies comparing and assessing effectiveness of more than two treatments investigate?

Answer 11

• Is it better than nothing

- Is it better than the old drug

Question 12

What must the objectives of the study define?

Answer 12

Objectives: specific, determines size and follow up rate

• Define entry criteria, disease, measures of response, end points
• Target population, study population, study sample size
• Determines how generalisable data is

Question 13

When doing observational study 4 possible outcomes of why data is statistically significant

Answer 13

1. Chance - reduce by large sample size
2. Bias
3. Confounding - when variable related to both study variable and outcome so effect of study variable on outcome is distorted
4. True Association

Question 14

What is bias?

Answer 14

systemiatic error in design, conduct or analysis of study which produces a mistaken estimate of an exposure on risk of diseas

Question 15

What are possible sources of bias?

Answer 15

• Issues with sponsorship
  
  • Study patient selection/allocation
  • Prejudice of patient
  • Prejudice of observer
  • Faulty method
  • Fualty analayses
  • Faulty interpretation

Question 16

What is the role of sponsorship in an RCT?

Answer 16

1. Critical requirement
   
   2. Data ownership
   3. Design issues
   4. Independent analyses
   5. Independent reporting

• Want to be able to design study appropriately with academic peers
• Can’t have that sponsors block data publication if don’t like results

Question 17

What is the role of patient selection in an RCT?

Answer 17

• Validates statistics
• Excludes biased allocation
• Equally distributed prognostic factors (known/unknown)
• Need stratification?
• Restricted/block randomisation

At baseline: similiar groups of people at the beginning and similiar physiologcial data between the two groups

Question 18

What is the role of controls in an RCT?

Answer 18

basis of comparison, without don’t know if there is observed effect

• Also realize side effect
• Placebo
• Regression to mean: any intervention at bad times will be followed by improvement
• Acclimatisation
• Seasonal effect

Question 19

What is the role of sample size in an RCT?

Answer 19

• Difference to be detected (minimum important)
• Significsnce of observed difference (alpha p)
• Confidence by which a negative result is genuience (beta 1 power)
• Variation of data (SD)

Bias in analysis

Question 20

Why must withdrawn subjects be considered?

Answer 20

adhereance to trial important

Question 21

What is the control event rate

Answer 21

incidence in control group

percentage

Question 22

What is the experimental event rate

Answer 22

incidence in experimental group

percentage

Question 23

How is absolute risk reduction calculated

Answer 23

CER-RER

Question 24

How is relative risk reduction calculated

Answer 24

(CER-RER)/CER

Question 25

What is number of patients needed to treat and how is it calculated?

Answer 25

number of patients that needs to be treated for one to get the benefit
1/ARR

Question 26

Why does randomisation occur

Answer 26

to prevent allocation bias

Question 27

Why double blind?

Answer 27

prevent measurment bias

Question 28

What is the difference between efficacy and effectiveness

Answer 28

efficacy: the true biological effect of a treatment effectiveness: effect of a treatment when actually used in “normal” practice

Question 29

How is relative risk calculated?

Answer 29

EER/CER

Question 30

What is consort

Answer 30

reporting guideline