I have very little experience in ststistics so my question is very basic I am testing a new assay for detecting B12 deficiency, active b12 compared to the current test in use total B12. I Am analyzing the two tests on the same patients and have 357 patients in total. I am using spss and excel so what would the most appropriate test to use.
Very generally, you can compare the area under the curve (AUC) of the reciever-operator characteristics (ROC). It might also be (more?) interesting to compare the positive and negative predictive values (PPV, NPV) for given classification rules.
Hello Aileen. See Chapter 12 (Screening and Diagnostic Tests) of Robert Newcombe's book:
- https://www.crcpress.com/Confidence-Intervals-for-Proportions-and-Related-Measures-of-Effect-Size/Newcombe/p/book/9781439812785
Note too that on the Additional Resources tab, you can download an Excel file (inside a zip file) that implements the methods Newcombe describes. HTH.
I am just adding this, not exactly on diagnostic test, but if you are comparing a new method with an older method, there are the Bland Altman plots, which looks at the results of one test over a range of values. It is called a Method Comparison Test, (more relevant for comparing two methods over a range of values.
Bland and Altman recommend plotting the difference (T-S) between the test (T) and comparative or reference method (S) against the average (T+S)/2 of the results obtained from the two methods. In the ideal case, there should not be any correlation between the difference and the average whether there is a bias or not.
Dear Aileen,
You may also want to check the NICE guidance for the Diagnostic Assessment Program.
https://www.nice.org.uk/about/what-we-do/our-programmes/nice-guidance/nice-diagnostics-guidance
Please check the Methods section of the programme manual (link to pdf at the bottom of the page).
Practical applications are published in HTA journal:
https://www.journalslibrary.nihr.ac.uk/search/#/?search=%22Diagnostic%20Assessment%20Process%22
Good luck!
Isaac
sensitivity, specificity,PPv,NPV ,ROC curves to compare and AUC to evaluate diagnostic tests.
Hello
I'm sorry, I don't have enough information in the field, it is better to ask my brother Mojtaba Mousavi Bozaz.
Why not use R . There are method comparison technics that look at comparing two methods.
Since you have errors on both y and x when plot one against the other. You can use a passing bablock regression, there are also others.
Use MEDCALC comparison of ROC curves by De Long et al, 1988
- Badges
- Science topic