A lab worker analyzing DNA. Photo by National Cancer Institute on Unsplash

Can Epidemiology Provide Benefits to Healthcare?

Healthcare when benefit through the sharing of data across the world; however, data privacy would need to be kept as a top priority.

If you would like to read this on Medium click here.

Can Epidemiology Provide Benefits to Healthcare?

        Healthcare in the United States, US, is a complex topic. It would be ignorant to have a one-cut solution to health care, like a chaos equation, any input into the system will result in an unexpected outcome. Using effective research methods effective treatments can be given to patients.

        Epidemiology yields promising results in providing such effective medical care to patients. The Institute of Medicine (US) Committee on Health Research and the Privacy of Health Information defines epidemiology as the practice of conducting health research such as analyzing past and future causes of diseases. Recalde et al, a team of joint researchers from various Spanish universities,  claims obtaining quality health records is required for epidemiology research. The National Cancer Institute has set up a series of “information grids” allowing for information transfer locations. Some countries acquire data from regulations mandated by the government. Denmark’s Hip Arthroplasty Registry (DHR) requires data input. According to Gundtoft et al, researchers from different Department of Orthopedics in Denmark, the DHR has 92% to 98% completion since 1995, including around 140,000 cases of hip arthroplasty. With an immense number of cases, researchers were able to identify unknown risk factors. Having a sizable database can lead to developments in healthcare.

The Institute of Medicine (US) Committee on Health Research and the Privacy of Health Information believe big data is a better way to research than experimental studies. Using a comprehensive dataset, trends in health appear, because when conducting experimental studies with a small population lurking variables disturb the variable you are studying as defined by Aaron Osgood-Zimmerman, a Ph.D. student at the department of statistics at the University of Washington. Maximizing the amount of health information becomes beneficial for diagnosing and treating rare diseases. One example is the Kawasaki disease, the inflammation of arteries leading to the heart (Mayo Clinic). Moore et al, professors in the Department of Primary Health Care Sciences at Oxford, report  “GPs” or general practitioners would be better equipped in diagnosing niche diseases if presented with relevant data. In the example of Kawasaki, the patients had multiple symptoms including but not limited to, rash, lymphadenopathy, and conjunctivitis. Only 23% of GPs considered Kawasaki as a “differential diagnoses” and 53% did not consider Kawasaki. An algorithm-based system could decrease the percentage of GPs who do not consider rare diseases as the algorithm can report a score of the likeliness of diagnosis.

        There have been developments of a “propensity score” based on health data. Cook and Collins suggest propensity scores alongside modeling the propensity are becoming prevalent ways of diagnosis. Understanding the full mathematics behind the propensity score is outside the scope of the paper, however, it is possible to discuss the general aspects of the propensity score. Peter Austin, a professor at the Institute of Health Policy, Management and Evaluation in Toronto, describes a propensity score to estimate the effectiveness of treatment through a comparison of observational data. A key component in performing experiments are randomized control trials, RCTs. RCTs minimize the effect of confounding variables. Propensity scores mimic RCTs through all four different ways to calculate a propensity score. Oritz et al, Ortiz himself is a surgical oncologist, were able to utilize the propensity score to determine if abdominal perineal excision (APE) led to higher postoperative complications and mortality rates than extralevator APE (ELAPE). Through the use of the propensity score, Ortiz et al. were able to determine there was no difference between mortality rates. Some might claim the study provided no addition to the medical world, however with the confirmation of both procedures having the same outcomes it allows patients and surgeons to choose the operation they personally prefer. A similar outcome happened when Karran et al, a researcher fellow at the University of Wales, performed a study on if surgery or definitive chemoradiotherapy was more effective in treating oesophageal cancer, cancer affecting the food pipe (NHS). In England Holt et al, researchers in the Department of Outcomes Research at the St George's Vascular Institute, performed a propensity score on the outcome of ruptured abdominal aortic aneurysm, when the main vessel to your heart starts swelling,  and the role of endovascular repair (NHS). Using the score Holt et al were able to determine when an endovascular repair was implemented mortality rates lowered. A propensity score is a valuable tool in providing effective healthcare to patients by analyzing wide-scale data from a variety of sources and experiments. In addition to analyzing symptoms and operations, another sector of science could improve the health of citizens.

        Genomics is another avenue to conduct a wave of well-informed medical decisions. De la Torre Díez, a professor in the Department of Signal Theory and Communications and Telematic Engineering at the University of Valladolid,  et al, claim, “[g]enomics is one of the main fields in which Big Data is playing an increasingly larger role.” The World Health Organization describes genomics as  “the study of genes and their functions, and related techniques.” The reason why genomics is now a viable method of research is due to the infrastructure set in place. Kris Wetterstrand, the Scientific Liaison to the Director for Extramural Activities at the National Human Genome Research Institute, reports the beginning of the 21st century the Human Genome Project thirteen years and three billion dollars to be completed. The extensive cost and time came as a necessity for the human genome is sequenced to a precision of 99.99%. The cost of sequencing a genome varies but the average for non-commercial sectors is $1,000. According to Phillips et al, a professor of Health Services Research and Health Economics at the University of San Francisco, through genomic advancements, the Centers for Medicare and Medicaid Services passed a policy allowing “advanced cancer patients” to have their genome sequenced as part of their clinical trials. Phillips et al believe there is still work the policy needs but is on a road to promise. Obtaining patient data creates a foreboding cloud.

        The use of big data in specific data with sensitive patient information puts a target on the industry for criminals to gain access to it. Verizon Wireless, a telecommunications company, released a report in 2015 about data security.  In Verizon’s report, they claim medical data is compromised due to miscellaneous errors, insider and privilege misuse, physical theft and loss.  Miscellaneous errors are harmless in intention but are due to human error such as sending classified information to the wrong recipient. On the other hand, 40% of insider and privilege abuse happen due to financial gain. Finally, theft and loss can be better prevented through encryption. A bar graph present by Matej Mikulic, an office administrator at the Institute of Public Health Osijek-Baranja County, displays patient’s data being leaked from 2009-2019. There were 2 abnormally high years of data breaches, the higher of the 2 was in 2015 with 80 million patients’ data leaked. In the Congressional House Hearing Honorable Will Hurd, Chairman of the IT Subcommittee, announces the company Anthem, a provider of health insurance, leaked over 80 million records. The US has provisions in place to reduce the number of leaked data. According to the Centers for Disease Control and Prevention, CDC, in 1996 the Health Insurance Portability and Accountability Act was enacted mandating, “the creation of national standards to protect sensitive patient health information from being disclosed without the patient’s consent or knowledge.” With the understanding of the importance of patient data privacy, the 21st century is endearing for those who seek epidemiology and its subsidiary fields of research.

        Using the data around the world the healthcare system can make better and cheaper decisions leading to the efficacy of healthcare. It is paramount data privacy always remains persistent when creating and sharing registries. Once the registries are established, researchers should use propensity scores to provide effective algorithms creating personalized treatments.

Works Cited

“2015 DATA BREACH INVESTIGATIONS REPORT.” Verizon Enterprise, Verizon, 2015, higherlogicdownload.s3.amazonaws.com/GOVERNANCEPROFESSIONALS/a8892c7c-6297-4149-b9fc-378577d0b150/UploadedImages/Landing%20Page%20Documents/DBIR%20Executive%20Summary%20vv%204-10-15.pdf.

“Abdominal Aortic Aneurysm.” NHS Choices, NHS, 4 July 2017, www.nhs.uk/conditions/abdominal-aortic-aneurysm/.

Austin, Peter C. “An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies.” Multivariate behavioral research vol. 46,3 (2011): 399-424. doi:10.1080/00273171.2011.568786

Cook, J.A. and Collins, G.S. (2015), The rise of big clinical databases. Br J Surg, 102: e93-e101. doi:10.1002/bjs.9723

De la Torre Díez, Isabel & Cosgaya, Héctor & Zapirain, Begoña & Lopez-Coronado, Miguel. (2016). Big Data in Health: a Literature Review from the Year 2005. Journal of Medical Systems. 40. 10.1007/s10916-016-0565-7.

Gundtoft, Per Hviid et al. “The Danish Hip Arthroplasty Register.” Clinical epidemiology vol. 8 509-514. 25 Oct. 2016, doi:10.2147/CLEP.S99498

“Health Insurance Portability and Accountability Act of 1996 (HIPAA).” Centers for Disease Control and Prevention, Centers for Disease Control and Prevention, 14 Sept. 2018, www.cdc.gov/phlp/publications/topic/hipaa.html.

Holt, P.J.E., Karthikesalingam, A., Poloniecki, J.D., Hinchliffe, R.J., Loftus, I.M. and Thompson, M.M. (2010), Propensity scored analysis of outcomes after ruptured abdominal aortic aneurysm. Br J Surg, 97: 496-503. doi:10.1002/bjs.6911

Institute of Medicine (US) Committee on Health Research and the Privacy of Health Information: The HIPAA Privacy Rule; Nass SJ, Levit LA, Gostin LO, editors. Beyond the HIPAA Privacy Rule: Enhancing Privacy, Improving Health Through Research. Washington (DC): National Academies Press (US); 2009. 3, The Value, Importance, and Oversight of Health Research. Available from: https://www.ncbi.nlm.nih.gov/books/NBK9571/

Karran, A., Blake, P., Chan, D., Reid, T.D., Davies, I.L., Kelly, M., Roberts, S.A., Crosby, T. and Lewis, W.G. (2014), Propensity score analysis of oesophageal cancer treatment with surgery or definitive chemoradiotherapy. Br J Surg, 101: 502-510. doi:10.1002/bjs.9437

“Kawasaki Disease.” Mayo Clinic, Mayo Foundation for Medical Education and Research, 31 Oct. 2019, www.mayoclinic.org/diseases-conditions/kawasaki-disease/symptoms-causes/syc-20354598.

Mikulic, Matej. “Number of People Affected by Health Data Breaches U.S. 2014-2019.” Statista, 13 Jan. 2020, www.statista.com/statistics/798564/number-of-us-residents-affected-by-data-breaches/.

Moore, Abigail, et al. “Recognising Kawasaki Disease in UK Primary Care: a Descriptive Study Using the Clinical Practice Research Datalink.” British Journal of General Practice, vol. 64, no. 625, 2014, doi:10.3399/bjgp14x680953.

Ortiz, H., Ciga, M.A., Armendariz, P., Kreisler, E., Codina‐Cazador, A., Gomez‐Barbadillo, J., Garcia‐Granero, E., Roig, J.V., Biondo, S. and (2014), Multicentre propensity score‐matched analysis of conventional versus extended abdominoperineal excision for low rectal cancer. Br J Surg, 101: 874-882. doi:10.1002/bjs.9522

“Oesophageal Cancer.” NHS Choices, NHS, 3 July 2019, www.nhs.uk/conditions/oesophageal-cancer/.

Phillips, Kathryn A et al. “Insurance coverage for genomic tests.” Science (New York, N.Y.) vol. 360,6386 (2018): 278-279. doi:10.1126/science.aas9268

Recalde, Martina et al. “Validation Of Cancer Diagnoses In Electronic Health Records: Results From The Information System For Research In Primary Care (SIDIAP) In Northeast Spain.” Clinical epidemiology vol. 11 1015-1024. 3 Dec. 2019, doi:10.2147/CLEP.S225568

Wetterstrand, Kris A. “The Cost of Sequencing a Human Genome.” Genome.gov, 30 Oct. 2019, www.genome.gov/about-genomics/fact-sheets/Sequencing-Human-Genome-cost.

“WHO Definitions of Genetics and Genomics.” World Health Organization, World Health Organization, 22 Apr. 2016, www.who.int/genomics/geneticsVSgenomics/en/.

Zimmerman, Aaron. Chapter 5: Experiments, Good and Bad. Summer 2014, https://www.stat.washington.edu/~azimmer/Lect7_Ch5.pdf. PowerPoint Presentation

Share :