Interventional Studies

Question 1

Give the common examples of an interventional study

Answer 1

Randomised control trials (RCTs)

Uncontrolled trials (everyone gets treatment)

Question 2

Describe intervention studies

Answer 2

• An intervention is administered to evaluate its efficacy and safety. Aim is to try and establish direct causation, by seeing if adding/removing a certain factor has a direct effect on developing/preventing a disease
• There is usually a ‘control’ group (standard therapy/no treatment/placebo) to compare to and an ‘intervention’ group (receives the intervention).
• Sometimes there is double blinding which everyone avoids knowledge of treatment influencing whether to enter patient into a trial (selection bias) or observers assessment of patients response (assessor bias). Also involves radomisation.

Question 3

Advantages of experimental studies

Answer 3

• Can help establish a very strong causation factor
• Very useful in determining possible cures/treats - used in marketing of new medications
• Can compare new treatments with current/conventional treatments to determine which is better

Question 4

Describe randomisation

Answer 4

This is allocating individuals to each study group randomly. It is where each individual has equal chance of being allocated to either control or interventional group. The only systemic difference between people in each group (treated and control) is the treatment they’re receiving. The other determinants of health outcome differ randomly between each group so effects are likely to cancel out.
This helps remove any selection/allocation bias.

Question 5

Disadvantages of experimental studies

Answer 5

• Can be time consuming and expensive
• Requires direct contact with the patient
• Ensure disease definition is consistent
• Need to take appropriate measures to avoid bias (double blind, randomisation, standardisation)
• May not always be feasible e.g. unethical to intentionally expose an individual to a risk factor
• Some patients get better due to placebo effect not treatment
• Need to ensure results observed are statistically significant (P<0.05) with suitable confidence intervals
• Assessor bias if double blind not possible

Question 6

What is a placebo?

Answer 6

Chemically inert and looks like an active drug. Placebo’s enable study to be double blind, enables placebo effect to be quantified

Question 7

Do you take into account non compliers in RCT?

Answer 7

Yes

If you do not include them, it may introduce selection bias as the subjects will no longer be randomly chosen.

Question 8

2 ways to interpret RCT results

Answer 8

Intention to treat analysis of results - compare all randomised to treatment with all randomised to control, it is important in real life as not everyone will comply with treatment 100%

On treatment analysis of results - compare those who only received/complied with treatment with those who received control. However, this may introduce selection bias.

Question 9

Two types of groups in RCT

Answer 9

Parallel

Crossover

Question 10

Describe parallel groups

Answer 10

Half allocated to control half allocated to treatment.

• When the individuals in the control group and the intervention group remain in those groups through the study
• This is used when health outcome is irreversible e.g. cure, remission or death. It is used for things like stroke, heart attack,

Question 11

Describe crossover groups

Answer 11

Halfway through the study, the individuals in the control group are switched with the intervention group.

• This is used for reversible health outcomes (where treatment can be reversed) e.g. analgesics for chronic pain, drug to lower serum cholesterol.
• It is good for continuous data outcomes e.g. blood pressure

Question 12

How do you get maximum compliance?

Answer 12

Select patients well and have a run in period where all get the treatment (identifies the patients who can’t tolerate and are unlikely to comply.