Sackett and Goldsmith developed a grading scheme for evaluating the performance of a clinical sign or symptom for diagnosis of a condition.
(1) type of comparison
(2) study population
(3) criterion for comparison
Level of Evidence
independent and blind
large number of consecutive patients suspected of having the target condition
small number of consecutive patients suspected of having the target condition
nonconsecutive patients suspected of having the target condition
sample of patients who obviously have target condition with or without normal individuals
other than gold standard (uncertain validity)
• Not all combinations of the parameters are provided for in the table.
• A gold standard is one with demonstrated validity for identifying the condition.
• The population in level 4 are of known diagnosis (preselected).
Grade of Evidence
3, 4 or 5
To read more or access our algorithms and calculators, please log in or register.