Decoding The Data: Contingency Tables鈥擯art 2
Scott Sherrill-Mix, Ph.D., continues his presentation on how to interpret contingency tables when evaluating diagnostic tests and associated measures of test accuracy using a real figure example.
Video Notes
Scott Sherrill-Mix, Ph.D., continues his discussion on contingency tables and covers the following topics:
- How to calculate PPV when taking a test from 1 population and applying it to another population.
- Common pitfalls that lead to misleading results when applied to low鈥憄revalence populations.
- How Bayes' rule explains these effects and allows adjustment of PPV/NPV across populations.
Key Takeaways
- Positive predictive values depend on the population you are testing.
- True positives vs false positives and how to calculate them.
- Be careful with calculating positive predictive values and negative predictive values, as the value calculated depends on the prevalence of the disease in the population. It will differ between a screening test and a targeted test.
- Before acting on a test result, always ask:
- What is the prevalence of disease in this patient population?
- Am I interpreting sensitivity/specificity or PPV/NPV?
- Has the test been validated in a population like mine?
Before watching the video above, please watch part 1 of this video series where Sherrill-Mix discusses how contingency tables are structured and how to calculate and interpret sensitivity, specificity, positive predictive value and negative predictive value.
Author Information
Scott Sherrill-Mix, Ph.D., Assistant Professor, Department of Microbiology, Genetics and Immunology, Michigan State University.
Stay up to date on all 黑料正能量Clinical Public Health Microbiology (CPHM) activities when you sign up for our monthly CPHM Newsletter. Be the first to know about free educational webinars, continuing education credit opportunities, advocacy, updated guidelines and more.
