Animal and laboratory studies form the lowest level of evidence for informing clinical decisions
Clinical Research
Quantitative Data Types and Tests
Quantitative data is that which can be expressed numerically and is associated with a measurement scale; not all numbers constitute quantitative data (e.g. tax file number!)
Logistic regression analysis
Logistic regression analysis is used to predict a dependent binary outcome (yes/no, dead/alive) based on one or more predictor variables
Randomisation
Randomisation is the process of assigning clinical trial participants to treatment groups such that each participation has a known (usually equal) chance of being assigned to any of the groups
Early Stopping of Trials
Reviewed and revised 26 August 2015 OVERVIEW Clinical trials are planned to stop when one of the following is reached sample size event count length of follow up goal scheduled date of closure Unplanned early stopping of trials may occur for a number of reasons (e.g. lack of funding, failure to recruit participants), which may […]
Case-Control Studies
Case-Control Studies: a type of retrospective observational study; control patients are ‘matched’ using some criteria (age, gender), typically should have no history of the disease of interest and should be representative of the general population
Risk and Numbers Needed to Treat
Risk is a nebulous term in evidence-based medicine, as it may refer to either relative or absolute risks. Overall, absolute risks are more meaningful and can be converted in ‘numbers need to treat’ (NNT), which are useful in the clinical setting
Receiver Operating Characteristic Curve
Receiver Operating Characteristic (ROC) curves plot sensitivity versus false positive rate for several values of a diagnostic test
LITFL posts by