The purpose is to evaluate the utility of diagnostic screens in informing patients about the presence or absence of diseases/conditions based on analysis of the performance data of the tests. Specifically, you will analyze the performance of two screens for colorectal cancer and interpret the results of the screen in the context of the incidence of colorectal cancer. Based on your analysis of colorectal cancer screening, you will evaluate the ethical considerations of communicating the diagnostic value of prenatal screening for genetic defects.
Details
Colorectal cancer is the second most common cause of cancer death in the United States. The overall incidence rate is 38.7/100,000 population, and the mortality rate is 13.9/100,000 population. The figure1 below shows the increase in the incidence of colorectal cancer as a function of age. The inset shows an expansion of the data for 50–59-year-old individuals.
Despite the supporting evidence, recommendations, and availability of several screening tests, a substantial proportion of people are untested. A simple, noninvasive test with high sensitivity for both colorectal cancer and advanced precancerous lesions could increase adoption and adherence, which could improve clinical outcomes.
Multi-targeted stool DNA test (MT-sDNA) is a simple, non-invasive test for colon cancer that is an alternative to colonoscopy in individuals at average risk for colon cancer. The test is based on DNA mutation in a specific gene, DNA methylation changes, and a hemoglobin immunoassay. A cross-sectional study at 90 sites in North America with 10,000 asymptomatic persons between the ages of 50 and 84 years who were considered to be at average risk for colorectal cancer and who were scheduled to undergo screening colonoscopy, was designed to measure the ability of MT-sDNA to detect colorectal cancer3. Fecal Immunochemical Test (FIT) for hemoglobin, which is non-invasive and used in clinical practice, was included as a comparative test.
The results of the study demonstrated that 65 patients had colorectal cancer, and 60 of these patients had a positive MT-sDNA test. By comparison, 48 patients had positive FIT results. The study also showed that 9167 patients did not have colorectal cancer, but 1231 of these patients had a positive MT-sDNA test and 472 had positive FIT results.
The ROC curves and performance measures of MT-sDNA test and FIT are shown below. The solid curve shows the performance of FIT, and the dashed curve shows the performance of MT-sDNA.
The sensitivity for MT-sDNA and FIT is 92% and 74% respectively, and the specificity for MT-sDNA and FIT is 87% and 95% respectively.
The recommendation is that patients with positive MT-sDNA or FIT results should have a diagnostic colonoscopy, and that patients with negative results should continue to be screened at a regular interval4,5.
Answer the following questions:
- A 57-year-old individual who receives positive MT-sDNA results does not have a 92% chance of having colorectal cancer. Explain why that is true if the sensitivity of the test is 92%.
- What is the probability that a 57-year-old with a positive test result has colorectal cancer? Use Bayes’ Theorem.
- Another 57-year-old individual who receives negative MT-sDNA results does not have an 87% chance of being colorectal cancer-free. Explain why that is true if the specificity of the test is 87%.
- What is the probability of a 57-year-old with a negative test result not having cancer? Use Bayes’ Theorem.
- What is the probability of colorectal cancer in an 80-year-old with positive MT-sDNA results?
- Is the MT-sDNA test more useful in screening for the presence of cancer or the absence of cancer? Explain your answer.
- Compare the performance attributes of MT-sDNA and FIT tests. Which test would you recommend for the 57-year-old and why?
- Cologuard is a MT-sDNA test marketed in the US. It is reimbursed by insurance so there is no cost to the patient, and the test is highly sensitive. There is a public health proposal to make the test more broadly available. You have been asked to evaluate lowering the age for eligibility from the current 45 to 35. Would you recommend lowering the eligibility age? Explain your answer in details.
- Read this article in the New York Times (https://www.nytimes.com/2022/01/01/upshot/pregnancy-birth-genetic-testing.html). Based on what you learned from analyzing the MT-sDNA test, explain how a test can be both 90% accurate and 90% wrong in predicting prenatal defects. Discuss in elaborated and thorough detail the ethical implications of the use of data and analytics and communicate the value of the test and of the insights from the results of the test.
References
Sung, H., Ferlay, J., Siegel, R. L., Laversanne, M., Soerjomataram, I., Jemal, A., & Bray, F. (2021). Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clinicians, 71(3), 209–249. https://doi.org/10.3322/caac.21660
Siegel, R. L., Miller, K. D., Goding Sauer, A., Fedewa, S. A., Butterly, L. F., Anderson, J. C., Cercek, A., Smith, R. A., & Jemal, A. (2020). Colorectal cancer statistics, 2020. CA: A Cancer Journal for Clinicians, 70(3), 145–164. https://doi.org/10.3322/caac.21601
Imperiale, T. F., Ransohoff, D. F., Itzkowitz, S. H., Levin, T. R., Lavin, P., Lidgard, G. P., Ahlquist, D. A., & Berger, B. M. (2014). Multitarget Stool DNA Testing for Colorectal-Cancer Screening. New England Journal of Medicine, 370(14), 1287–1297. https://doi.org/10.1056/nejmoa1311194Links to an external site.
U.S. Preventive Services Task Force. (2021, May 18). Final Recommendation Statement: Colorectal Cancer: Screening. United States Preventive Services Taskforce. https://www.uspreventiveservicestaskforce.org/uspstf/document/RecommendationStatementFinal/colorectal-cancer-screening#bootstrap-panel–4
Selby, K., Jensen, C. D., Zhao, W. K., Lee, J. K., Slam, A., Schottinger, J. E., Bacchetti, P., Levin, T. R., & Corley, D. A. (2019). Strategies to Improve Follow-up After Positive Fecal Immunochemical Tests in a Community-Based Setting: A Mixed-Methods Study .Clinical and Translational Gastroenterology, 10(2), e00010. https://doi.org/10.14309/ctg.0000000000000010Links to an external site.
Additional Details
The first two articles below about how to communicate scientific information may be useful to you as you answer the questions and develop strategies for communicating insights to persuade health workers and patients to change their behavior. Wilber’s article is helpful for answering these questions as well.
- How the information deficit model helps create unidirectional and paternalistic mode of healthcare communicationLinks to an external site.. (2018). Journal of Communication in Healthcare, 11(4), 239-240, https://doi.org/10.1080/17538068.2018.1567026 [2 pages]
- Simis, M. J., Madden, H., Cacciatore, M. A., & Yeo, S. K. (2016). The lure of rationality: Why does the deficit model persist in science communication?. Public understanding of science, 25(4), 400-414, https://doi.org/10.1177/0963662516629749 [15 pages]
- Wilber, J. (2022, March 2022). Precision and Recall. MLU-Explain. https://mlu-explain.github.io/precision-recall/?utm_medium=email&utm_source=topic+optin&utm_campaign=awareness&utm_content=20220411+data+ai+nl&mkt_tok=MTA3LUZNUy0wNzAAAAGDuMJDcUJkhvrCGqjULcDfx-GwcbuAwkw-J94UGUdd6ZGWBQHKCD7LQhIfC3CURcRVlv4Vq1fznGk7wpPh9t0porrk2DWp5J5hcnWn_v9HZGOlkq
NOTE:
- The article by Wilber, J. (2022, March 2022). Precision and Recall. MLU-Explain (https://mlu-explain.github.io/precision-recall/?utm_medium=email&utm_source=topic+optin&utm_campaign=awareness&utm_content=20220411+data+ai+nl&mkt_tok=MTA3LUZNUy0wNzAAAAGDuMJDcUJkhvrCGqjULcDfx-GwcbuAwkw-J94UGUdd6ZGWBQHKCD7LQhIfC3CURcRVlv4Vq1fznGk7wpPh9t0porrk2DWp5J5hcnWn_v9HZGOlkqo) is very useful in helping you understand how ROC curves can be used to understand the performance of machine learning-based medical screening. Read before starting.
AssessmentRubric is attached.
Rubric is attached.
