ABSTRACT
Control event rate (CER) d. Georgium is better than placebo at achieving a 50% reduction in
pain
6. Which of the following are commonly used statistical tests? a. The x2 test b. Student’s t-test c. The Mann-Whitney test d. Fisher’s exact test
7. The following are the results of a trial using Jasarch, a newly discovered protein in blood, that may be useful in the diagnosis of sarcoid
Sarcoid
Jasarch Present Absent
Positive 70 130
Negative 30 120
a. The sensitivity of Jasarch in the diagnosis of sarcoid ¼ 70/ 200 ¼ 35%
b. The specificity of Jasarch in the diagnosis of sarcoid ¼ 120/ 150 ¼ 80%
c. The positive predictive value of Jasarch in the diagnosis of sarcoid is 70/100 ¼ 70%
d. The negative predictive value of Jasarch in the diagnosis of sarcoid is 120/250 ¼ 48%
9 . S ta tistic
s: Q u e stio
n s
8. Likelihood a. A likelihood ratio describes the ratio of positive to negative tests b. A positive likelihood ratio of 1 indicates that the test is of no
clinical utility c. Likelihood ratios can be used to convert a pre-test probability to a
post-test probability d. Likelihood ratios are a clinical application of Bayes’ theorem
9. Regarding confidence intervals when analysing data from a clinical trial a. A 95% confidence interval gives a range around the trial data
within which we can be 95% sure that the true value lies b. The narrower the confidence interval the more likely that the trial
result is true c. A 95% confidence interval is calculated as 2 standard deviations
either side of the trial result d. Confidence intervals assume the use of parametric data
10. Regarding potential errors in a study a. The probability of a type 1 error is the same as the significance
level b. If the sample size stays the same, reducing the chance of a type 1
error will increase the chance of a type 2 error c. Assuming that all other conditions stay the same, reducing the
sample size will always increase the chance of both type 1 and type 2 errors
d. Type 1 error (a) is the rejection of the null hypothesis when it is true
9 . S ta tistic
s: Q u e stio
n s
1a. T 1b. F 1c. F 1d. T The results of diagnosis trials are usually presented in the form of a 22 table and described using sensitivity, specificity, positive and negative predictive values:
Target disorder
Present Absent
FRB-positive a = 80 b = 20 FRB-negative c = 120 d = 180
Sensitivity ¼ a/(aþc) ¼ 80/200 ¼ 40% Specificity ¼ d/(bþd) ¼ 180/200 ¼ 90% Positive predictive value ¼ a/(aþb) ¼ 80/100 ¼ 80% Negative predictive value ¼ d/(cþd) ¼ 180/300 ¼ 60%.
