The Rotherham NHS Foundation Trust Library & Knowledge Service
  • About Us
    • What we do
    • Can I join?
    • How to find us
    • A-Z of services
    • Book Group
    • Publications Repository
    • Contact us
    • Privacy Policy
    • Accessibility
  • My Speciality
  • Resources
  • Training
  • Request Search
  • Keep up to date
  • Knowledge Mobilisation

Critical Appraisal Glossary

General glossaries and other useful information:

Bandolier - Glossary
CEBM - Glossary
Bandolier - Describing results of trials and reviews
CMAJ - Tips for Learners of evidence-based medicine
BMJ Evidence-Based Medicine - Down with odds ratios
CEBM - Likelihood ratios
CEBM - CATMaker (computer-assisted critical appraisal tool) and EBM calculators

Text book - Dictionary of Statistics and Methodology

Glossary of specific terms

Bias
Bias is systematic error that may “distort the design, execution, analysis and interpretation of research”. There are many types of bias which can occur at any stage of the research design and execution. 
Sackett, David L. Bias in analytic research. Journal of Chronic Diseases 1979. 32(1-2):51-63. Doi: 10.1016/0021-9681(79)90012-2
CEBM - Catalogue of Bias

Power calculation

A power calculation is used to calculate the minimum sample size (e.g. of participants) required in order to be reasonably sure of detecting an effect of a given size. It is used to prove that the sample size in a particular study or trial is justified.
See:
Bandolier - statistical power
Effect Size FAQs - How do I calculate statistical power?

P-value
The p-value is the probability that any particular outcome would have arisen by chance. Standard scientific practice usually deems a p-value of less than 1 in 20 (i.e. p<0.05) as statistically significant and a p-value of less than 1 in 100 (p<0.01) as statistically highly significant.
See:
Bandolier - p value
Jerry Dallal's Little Book of Statistical Practice - p value

Confidence interval
A confidence interval provides a way of assessing the effects of chance. It aids the interpretation of clinical trial data by showing how small or how large the true size of effect might be.
See:
Bandolier - confidence interval
Useable Stats - confidence interval tutorials

Experimental event rate (EER)
The experimental event rate is the rate at which an event (such as a response to a drug or other intervention) occurs in an experimental group, for example, how many times an event occurs in the experimental or intervention group. It is calculated by dividing the number of people in whom the event occurred by the total number of people in the experimental group, then multiplied by 100 to get a percentage.
The calculation of absolute risk reduction, relative risk reduction and number needed are all reliant on the EER.
See also:
Bandolier - experimental event rate
Wikipedia - experimental event rate

Control event rate (CER)
The control event rate is the rate at which an event occurs in a control group, for example, how many times an event occurs in the group with no intervention, or using a placebo. It is calculated by dividing the number of people in whom the event occurred by the total number of people in the control group, then multiplied by 100 to get a percentage.
See also:
Bandolier - control event rate
Wikipedia - control event rate

Absolute risk reduction (ARR)
Absolute risk reduction is the difference between the event rate in the experimental group and the event rate in the control group. i.e. the difference between the experimental event rate (EER) and the control event rate (CER). It is calculated by subtracting the experimental event rate from the control event rate.
See also:
Bandolier - absolute risk reduction
CMAJ - Tips for Learners of evidence-based medicine

Relative risk reduction
This is the percentage reduction in events in the event rate in the treatment group (EER) compared to the event rate in the control group (CER). It is calculated by subtracting the EER from the CER, then dividing by the CER. 
See:
Bandolier - relative risk reduction
CMAJ - Tips for Learners of evidence-based medicine

Number needed to treat (NNT)
This is the average number of patients that needs to be treated in order for one to benefit from the treatment. So the higher the number needed to treat, the less effective the treatment. It is calculated by dividing 100 by the absolute risk reduction (ARR).
See also:
Bandolier - number needed to treat
CEBM - number needed to treat
Bandolier - ACP article - Using Numerical Results from Systematic Reviews in Clinical Practice
CMAJ - Tips for Learners of evidence-based medicine

Back to Critical Appraisal Page
Contact us
Find us
A-Z of services



Privacy Policy 
Accessibility statement 
How can we improve?
Books
Journals
Health promotion resources



Request a search
Search the evidence
Speciality gateways
Request training
Help with Athens
Bulletins and alerts
All content © 2023 The Rotherham NHS Foundation Trust Library & Knowledge Service unless otherwise stated.