Interventional studies

Question 1

Experimental intervention

Answer 1

-one group remains control (or standard) and the other receives some intervention

Question 2

Randomisation

Answer 2

-after a subject enters the study they are randomly allocated to either control arm or intervention arm through a process of randomisation

Question 3

Simple randomisation

Answer 3

-randomising each entrant separately using a computer generated list

Question 4

Block randomisation

Answer 4

• where randomisation occurs in a set of specified numbers, say blocks of 6, to ensure equal number distributed between two arms
• the larger the block size the lesser the ability to predict allocation

Question 5

Stratified randomisation

Answer 5

• baseline characteristics which may have an implication on final outcome are stratified so that equal distribution across groups is enforced
• does not eliminate the need for adjustment for differences in the baseline composition

Question 6

Cluster randomisation

Answer 6

• where a unit of randomisation and analysis is various centers or catchment zones, not individual patients
• all patients in that arm receive the same intervention
• this leads to a loss of statistical power in comparison with a patient randomised trial

Question 7

Effective sample size

Answer 7

• used in cluster RCTs in the place of actual sample size when undertaking a power calculation
• to calculate ESS the intracluster correlation is used which represents the degree to which the various individuals in a cluster resemble each other in the outcome measure

Question 8

Minimisation

Answer 8

• useful in 2 situations
  1. you want to do a stratified randomisation but your trial is too small
  2. you are conducting a cluster randomisation
• minimisation is used to achieve a balance between treatment groups
• in minimisation schemes, the next allocation depends on characteristics of those already allocated
• the allocation of each participant aims to ensure a balance of prognostic factors between groups
• not actually random and can be biased

Question 9

Quasi-randomisation

Answer 9

• refers to randomising using even/odd numbers of the date of birth, day of the week etc
• not reproducible
• sequences cannotensure equal distribution of variables
• must be avoided

Question 10

Use of randomisation

Answer 10

• permits use of probability theory in making inferences
• eliminates effects of bias
• facilitates blinding
• distributes baseline characters in an unpredictable fashion

Question 11

Cross over trial

Answer 11

• refers to an interchange of study and control groups after a washout period so that all subjects in a study receive both placebo and treatment but in a different order
• cannot be used for curable diseases

Question 12

Parallel RCT

Answer 12

-both control and intervention happens in a parallel with no cross-over

Question 13

N of 1 trial

Answer 13

• here a single subject is administered placebo and active intervention in tandem under double blind controlled conditions
• optimises treatment schedule for a patient but cannot be generalised to other patients

Question 14

Factorial trials

Answer 14

• most RCTs evaluate a single therapeutic factor of interest compared to a placebo
• 2x2 factorial design involves participants being allocated to one of the four possible combinations
• tests independent effect of a drug and combined effect in 2x2x2

Question 15

Patient preference trials

Answer 15

• patients are allocated the therapy they prefer
• only those who have no preference are randomised
• maximises recruitment
• complex consent

Question 16

Zelen’s modified RCTs

Answer 16

• generally informed consent is obstained before randomising a sample for RCT but patients withdraw if they don’t feel they are getting treated
• in Zelen’s approach the randomisation occurs before consent
• not felt to be ethical

Question 17

Non-inferiority trials

Answer 17

• most RCTs are conducted to show superiority of one intervention over another but sometimes drug companies only want to demonstrate the new intervention is not inferior
• also known as an equivalence trial or non-inferiority
• higher sample size is needed

Question 18

Uncontrolled trials

Answer 18

• no control group

- study inference is made using a historical control like the treated group before intervention

Question 19

Before and after trials

Answer 19

• the outcome is assessed before and after some intervention in one group of subjects
• the observed change may be attributable to a factor other than the intervention itself-makes the design inherently weak
• no randomisation occurs in this design

Question 20

Multi-arm trial

Answer 20

• more than one study arm is included
• easy to design and allows intervention-1 vs intervention 2 comparison
• requires a large sample size and power may be reduced as a consequence

Question 21

Efficacy

Answer 21

• how well a drug can work under optimal circumstances

- trial undertaken to meet regulatory approval

Question 22

Effectiveness

Answer 22

• how well it works under usual practice circumstances

- trial helps with formularies of usefulness of the drug

Question 23

Efficiency

Answer 23

-measures whether health care resources are being used to maximise value for money

Question 24

Pragmatic RCTs

Answer 24

• aka practical clinical trials
• looks at effectiveness
• CATIE

Question 25

Characteristics of pragmatic RCTs

Answer 25

• select clinically relevant alternative interventions to compare
  2. include diverse population
  3. recruit patients from heterogenous practice settings
  4. collect data on a broad range of health outcomes
  5. they do not tell how well a drug works (efficacy) they inform how effective the intervention is in the real world

Question 26

Problems with pragmatic designs

Answer 26

• substantial drop outs
• blindings can fail
• less rigorous methodology
• higher rates of non-specific treatment effects

Question 27

Key points for pragmatic RCTs

Answer 27

• reflect heterogeneity of patients in clinical practice
• minimise exclusion criteria
• focus on groups with a wide range of diagnoses
• define pateitns on presentation
• may not employ placebos
• may not be blinded
• must carefully conceal allocation during randomisation
• may evaluate real world, function based outcomes