Evaluation of a novel serum biomarker for the rapid diagnosis of sepsis is performed in a sample of 100 patients with fever. The biomarker is compared with positive culture results as the gold standard and yields the following information:
Sepsis present {culture positive) 
Sepsis absent {culture negative) 

Biomarker positive 
30 
10 

Biomarker negative 
30 
30 

n 
60 
40 
With reference to these results, define the following and give the values for the performance of the test:
a) Sensitivity. (20% marks)
b) Specificity. (20% marks)
c) Positive predictive value. (20% marks)
d) Negative predictive value. (20% marks)
e) Accuracy . (20% marks)
Sepsis present (culture positive) 
Sepsis absent (culture negative) 

Biomarker positive 
30 (a) 
10 (b) 
(a + b) 
Biomarker negative 
30 (c) 
30 (d) 
(c + d) 
n 
60 (a + c) 
40 (b + d) 
(a+b+c+d) 
a) Ability of test to identify true positives
Or the probability the test will be positive in individuals who do have the disease
Sensitivity a/(a+c) 30/60 50%
b) Ability of test to identify true negatives
Or the probability the test will be negative in individuals who do not have the disease.
Specificity d/(b+d) 30/40 75%
c) Likelihood of positive test meaning patient has sepsis
PPV a/(a+b) 30/40 75%
d) Likelihood of negative test meaning patient does not have sepsis
NPV d/(c+d) 30/60 50%
e) The ability to differentiate patient and healthy cases correctly.
Accuracy (a+d)/(a+b+c+d) 60/100 60%
Additional Examiners' Comments:
The question clearly stated that a definition was required. Many candidates either could not define the terms or just missed this part of the question and therefore missed out on marks. This question has come up a number of times in past exams and these are basic statistical concepts that some candidates clearly do not understand.
This question closely resembles all other previous questions about the measures of diagnostic test accuracy:
After being absent from the papers for over five years, one might have been forgiven for thinking that such calculatorintense statistics questions were demoted to the level of primary exam material (as most recent statistics questions in the Fellowship Exam have been more about interpretation of metaanalysis data and other such ultraclever "fellow level" uses of EBM). The main difference in 2016 was the addition of accuracy as one of the examined parameters. This has never been examined previously, and is not a frequently mentioned measure (even though colloquially we might use the term nearconstantly). An excellent 2008 article was used to define it for the purposes of this model answer.
Clearly, at least one candidate remembered all the definitions, and got 10 marks.
a)
b)
c)
d)
e)
Šimundić, AnaMaria. "Measures of diagnostic accuracy: basic definitions." Med Biol Sci 22.4 (2008): 615.