In: Statistics and Probability
A new test has been developed for diagnosing Rheumatoid Arthritis. The test was evaluated in four communities with the following results Table-1:
Table 1: Results of a new screening test for Rheumatoid Arthritis in Four Communities
Community A |
Rheumatoid Arthritis |
||
Present |
Absent |
Total |
|
Test Result + |
887 |
888 |
1775 |
Test Result -- |
99 |
7989 |
8088 |
986 |
8877 |
9863 |
Community B |
Rheumatoid Arthritis |
||
Present |
Absent |
Total |
|
Test Result + |
1485 |
385 |
1870 |
Test Result -- |
165 |
3465 |
3630 |
1650 |
3850 |
5500 |
Community C |
Rheumatoid Arthritis |
||
Present |
Absent |
Total |
|
Test Result + |
3634 |
269 |
3903 |
Test Result -- |
404 |
2423 |
2827 |
4038 |
2692 |
6730 |
Community D |
Rheumatoid Arthritis |
||
Present |
Absent |
Total |
|
Test Result + |
2866 |
56 |
2922 |
Test Result -- |
318 |
506 |
824 |
3184 |
562 |
3746 |
2.1. Calculate the prevalence of Rheumatoid Arthritis in each of these four communities.
(4 points)
2.2. Calculate the sensitivity of the test for each community. (4 points)
2.3. Graph the sensitivity percentages against the prevalence for the communities. No need to use a graph paper! (1 point)
2.4. Calculate the specificity of the test for each community. (4 points)
2.5. Graph the specificity percentages against the prevalence for the communities. No need to use a graph paper! (1 point)
2.6. Calculate the positive predictive value of the test for each community. (4 points)
2.7. Graph the positive predictive values against the prevalence for the communities. No need to use a graph paper! (1 point)
2.8. What problems might one have in applying the positive predictive value if, as is usually done, a test is first evaluated in a hospital population and then used in the community? (2 points)
Data:
community A | community B | community C | community D | |||||
Present | Absent | Present | Absent | Present | Absent | Present | Absent | |
887 | 888 | 1485 | 385 | 3634 | 269 | 2866 | 56 | |
99 | 7989 | 165 | 3465 | 404 | 2423 | 318 | 506 | |
Total | 986 | 8877 | 1650 | 3850 | 4038 | 2692 | 3184 | 562 |
2.1
Prevalence = no. of cases / population size
Population size = 25839
Community A : 9863/25839 =38%
Community B : 5500/ 25839 = 21%
Community C : 6730/ 25839 =26%
Community D : 3746/ 25839 =14%
2.2
Sensitivity
Community A : 986/9863 =10%
Community B :1650/ 5500 = 30%
Community C : 408/ 6730 =60%
Community D : 3184/3746 =85%
2.3
Community | prevalence | sensitivity |
A | 38% | 10% |
B | 21% | 30% |
C | 26% | 60% |
D | 14% | 85% |
2.4
Specificity
Community A : 8877/25839 =34%
Community B :3850/ 25839 = 15%
Community C : 2692/ 25839 =10%
Community D : 562/25839 =2%
2.5
Community | prevalence | specificity |
A | 38% | 34% |
B | 21% | 15% |
C | 26% | 10% |
D | 14% | 2% |
2.6
Community | prevalence | sensitivity | specificity | Positive predicated value |
A | 38% | 10% | 34% | 9% |
B | 21% | 30% | 15% | 9% |
C | 26% | 60% | 10% | 19% |
D | 14% | 85% | 2% | 13% |
2.7