Lecture 9: Quantitative Research- Non-randomised designs

Question 1

Equipoise

Answer 1

Doctors have a sworn responsibility to do the best for their patients.
Doctors need to be able to justify all treatments in and RCT in order to expose patients to the risk of getting allocated into any one of the groups.

Question 2

Other reasons we can’t Use RCT’s

Answer 2

1. Reduced access to new drugs- in order to get enough participants for a new experiment, funding has to be reduced to new, better and more promising drugs. This is so that taking part in the study is the only real chance of getting the drug.
2. Ethics - Cannot do tests on vulnerable groups, some experiments are too risky
3. Time to report - all the people that aren’t able to access the drug in the mean time :(
   a. Opportunity Cost: RCT’s are expensive - other trials that aren’t being funded
   b. Orphan Disease: Rare diseases that take a log time to gather enough study participants for- not Plausible to do an RCT

Question 3

Complexity of Clinical Practise

Answer 3

Controlling the vast variation in medicine e.g. surgical techniques is difficult. Sometimes we cannot directly control how treatment differs for each arm

Question 4

Exploratory Studies

Answer 4

Preliminary studies- the aim is to identify potential relationships between variables that might be worth studying. Used to study/look at data in a field where you may not know as much
Develops a series of research questions that can be studied using explanatory methods

Question 5

Observational Designs

Answer 5

Look at populations as a whole and identify and compare groups within the population

• Deductive: Start with a theory
• No Randomisation, Bias is probably introduced

Question 6

Cross-Sectional Study

Answer 6

• Snapshot
• Observe differences in characteristics (IV) between groups and compare the outcomes (DV)

Advantages: No follow-up
Cheap and quick

Disadvantages: only a correlation
People may end up not representing the underlying variables that were captured in the study because there is no follow-up e.g. link between being breast-fed as a child and adult obesity

Question 7

Cohort

Answer 7

A group of people are followed up over time

Advantages: Can look at historic events- therefore there are more chances to imply causation b/w variables as you cant detect the variation between the IV and DV over a longer period of time- more accurate

Disadvantages: subject to drop-out- therefore sample size may end up being not as representative of the POI
Bias due to demographic change- societal changes could cause confounding.

e.g. Growing up in NZ

Question 8

Case-Control

Answer 8

Start with the DV
Find similar people (based on the control variables)
look at the values of the IV to see if there is a difference between people with the disease and without.

Advantages: Cheap
Good for rare outcomes and where RCTs may be deemed unethical

Disadvantages: Simple- subject to recall bias
Wastes a lot of data about the control population