Clinical Trials in Respiratory Disease

Question 1

What does PICOT stand for, and what does it describe?

Answer 1

Describes the clinical (i.e. patient) setting

Population

Intervention

Comparator/Control

Outcome

Timing

Question 2

What is internal validity?

Answer 2

• extent to which the results of a study are valid (accurate, robust, etc.) for the sample of pt being studied
  
  • how well did the study answer the question it set out to?
• dependent on appropriate study design, data collection, and analyses
• from an epidemiological standpoint, seeks to ask whtether or not the study adequately identified and minimised:
  
  • confounding
  • information bias
  • selection bias

Question 3

How is a trial assessed for internal validity?

Answer 3

• randomization of subjects
  
  • reduce confounding by evenly distributing potential confounders, making treatment groups identical in all other aspects other than the intervention
  • reduces selection bias (e.g assigning a pt to a particular arm)
• blinding/masking of subjects, investigators, outcome assessors
  
  • reduces information bias - prejudice about the intervention influencing the outcome or its ascertainment
• objective outcome ascertainment by centralized, blinded committee
  
  • reduces information bias - predetermined strict, standardized, objective criteria to reduce subjectivity as to whether an outcome has occurred
• intention to treat analysis
  
  • reduces selection bias - those who drop out are almost always systematically different than those who stay in; tf assume no movement

Question 4

What is a p value?

Answer 4

• probability that the observed result arose from chance
  
  • i.e. there is truly no difference between the groups compared, and the observed difference was just a chance finding
  • null hypothesis = no differencce between the groups

Question 5

What is a confidence interval?

Answer 5

• interval within which there is 95% confidence that the true value lies
  
  • if null value excluded = statistically significant
  • if null value (i.e. no difference between compared groups) included = non-significant
    
    • e.g. 1.0 for ratios (HR, RR, OR) and 0 for differences (absolute risk differences)
    • above this value the result suggests the intervention is actually causing the outcome it intends to prevent
• width indicates precision
  
  • narrower = more precise and vv
  • larger sample size = narrower CI (more representative of true pop’n)

Question 6

What is type I/alpha error?

Answer 6

Determining a tx effect when there is no true tx effect

Question 7

What is type II/beta error?

Answer 7

Determining there is no tx effect when in truth there is a tx effect

Question 8

What is power?

Answer 8

• ability of a study to determine a tx effect when there is one
• able to detect a specifed difference between two groups in terms of the primary outcome
• determined by sample size

Question 9

What is number needed to treat?

Answer 9

NNT = 1 / (absolute risk or rate reduction)

of people needed to undergo intervention in order to prevent the outcome in one (marker of intervention efficacy)

**report with a time reference!**

Question 10

How is a trial assessed for external validity/generalizability?

Answer 10

• trial must first be internally valid
• PICOT table
  
  • assess concordance between the trial results and the clinical setting
• selection criteria
  
  • inclusion and exclusion criteria
• success of randomization (Table 1 or 2)
  
  • profile of study participants –> compare to your clinical setting/patient

Question 11

What is external validity?

Answer 11

• generalizability or applicability of trial results to the situation you want to apply them to e.g. clinical/patient, general pop’n