Quantitative 4- Data Analysis Flashcards
As part of the continuing professional development, what is encouraged?
Important in healthcare to be up to date with current research, always read current articles and journals.
What did Greenhalgh (1997a, 1997b) provide in his two seminal papers?
- provides two seminal papers on how to interpret statistics in RCT’s, including how researchers cheat to produce results that appear significant.
What are the methods used in analysing data?
o Relative risk o Absolute risk o Odds ratios AND o Number needed to treat (NNT) to describe the results of a trial.
What are the differences in the two styles of analysis?
is one establishes statistical significance and other measures clinical effectiveness.
Why do we need statistics?
- Statistics is about dealing with uncertainty and variation and the aim of statistical analysis is to establish that an intervention or medication will work for most people, most of the time.
- This declares that an intervention should demonstrate a real effect at least 95 times out of 100 to be beneficial.
- We know that humans differ in their response to an illness, with some suffering more than others.
- Doctors diagnose patients probabilistically- there is a strong probability for this patient to have this disease.
What is essential when collecting data, that needs to be analysed?
-It is essential that the data are collectable in a form appropriate for the study and that they measure what they are supposed to measure otherwise the whole research will be misleading.
What are the two elements of discrete outcomes?
- Risk: the probability of something going wrong, and the negative consequences if it does.
- Rate: the actual number of events that occur.
What is the other type of outcome measure?
Continuous
Average, Mean, Median and Mode
- It is important to use the right measure to compare the control and experimental group.
- Often done by comparing the average or mean of the groups.
How can you calculate measures of effect?
- The purpose of a randomised control trial is to determine whether one treatment or intervention is better than the other.
- Can be done by constructing a 2x2 table
- Using this table we can enter the results of trial and calculate relative and absolute risk ratios, odd ratios and NNH.
What does risk identify?
-Determine what risk the control group has and what risk the experimental group has of getting the disease.
What is experimental event rate? -ALSO REFERS TO CONTROL EVENT RATE (CER)
is a measure of how often a particular statistical event (such as response to a drug, adverse event or death) occurs within the experimental group.
a- being the experimental group that caught the event
b-being the experimental group that did not get the event
EER= a/(a+b).
What is relative risk?
is the ratio of the risk of the event in one regimen compared with the risk of an event in another regimen.
The number one is an important number in understanding relative risk.
If the RR=1 the risk of the event is equal in each group
- therefore there is no difference in the risk of event in both treatment.
If the RR= <1 the risk of the event is lower in the experimental group
If the RR= >1 the risk of the event is higher in the experimental group
EER/CER
What is odds ratio?
measure off treatment effectiveness.
If the OR= <1 the odds of the event are lower and the treatment effect is better in the experimental group.
If the OR= >1 the odds of the event is higher which means that the control group has the better treatment
(a/b)/(c/d)
How can you represent the OR visually?
a forest plot
