Big Data Analytics in Health Care and Drug Therapy Management—Opportunities and Considerations

J Clin Pathways. 2022;8(3):26-28. doi: 10.25270/jcp.2022.4.1

Modern technology allows for the mass collection of digital health data, in turn requiring novel analytical methods to make sense of it all.¹ Artificial Intelligence (AI)—including rule-based expert systems, machine learning, natural language processing, and so on—have become a vital tool for making sense of the huge volume of health data being collected daily through the use of big data analytics (BDA).² The application of BDA in health care has been extraordinary; clinically, methods have been used to improve patient-based care, detect disease spread earlier, and monitor the quality of health care.³ For pharmaceutical players that integrate BDA, there is an opportunity to excel above their comparators. BDA can improve treatment methods, increase knowledge on disease mechanisms, and monitor prescription choices and behaviors—necessary aspects for optimizing effectiveness.³

Administratively, BDA are used to gain insights into aspects of health care operations, improving offered services.⁴ For all decision makers: payers, providers, and policymakers, the expansive reach of BDA in health care is evident through its renewal of the health care landscape to achieve better patient outcomes, improved patient access, lower costs, enhance provider support, and improve health equity.¹ This article discusses key areas where BDA are being leveraged in health care, the implications to payer and pharma stakeholders, and key considerations for success.

Big Data Analytics in Health Care 

Precision medicine

Data  analytics  are  being  used  to  transform  treatment  recommendations,  enabling providers to more effectively customize patients’ therapy regimens to increase the likelihood of success from various therapy choices, improving patient outcomes while reducing unwarranted adverse events. Providers are using clinical decision tools that analyze massive amounts of data and integrate them with the patient’s own medical profile to best predict outcomes for individual patients.⁷ Providers and payers are also leveraging data analytics to reveal associations and data insights that otherwise would not be seen, ranging from responses to medications, to hospital readmission rates, to determining the most cost-effective treatments for certain breast cancers like the lanes of care created at Hackensack Meridian with its analytics partner COTA.^8,9 Leveraging their electronic medical record (EMR) data analytics, Meridian and COTA developed the breast cancer bundles in the lanes of care, which removed unwarranted variation from provider practice and reduced their total spend while improving patient outcomes.⁹ 

Also, as providers use predictive analytics and algorithms on a population and individual level through clinical decision support tools in the EMR, they are able to identify at-risk patients, make more accurate diagnoses, and tailor their treatment regimens, while at the same time, addressing patients’ social determinants of health to further improve patient outcomes, care, equity, and quality of life.

Population Health

BDA continue to upgrade the entirety of healthcare by increasingly considering aspects of  population health.  Historically, these approaches focused solely on the individual level, but now the shift to improve the health of defined populations or communities is increasing.⁵ The identification of high-risk populations for therapies is possible through curated algorithms. Long-term wellness for communities is achievable with the ability to monitor chronic conditions longitudinally.  The securement of a community’s health is a success shared by the community’s partners: payers, accountable care organizations, and health care delivery systems.⁶ US Medicare, Medicaid, and commercial insurers have all encouraged providers to adopt value-based delivery models, offering reimbursement for the improved health outcomes of total populations in addition to individual patients.⁶  Lower utilization translates into larger shared-savings for providers and payers.⁶

Additionally, BDA allow for the appropriate inclusion of the social determinants of health into a broad array of health and medical applications.⁶ Increasingly, a more complete picture of patients is being made available to improve all aspects of the health care experience, from understanding a patient’s social needs, to knowing their preferences, to addressing therapy choices leveraging their genomic data. There is also the opportunity to engage in segmentation and clustering. Within the volumes of electronic health data that now exist, analysts can separate patients into groups with similar characteristics. The ability to complete this can be used for predictive analyses, like forecasting what type of patient will benefit most from a treatment, or for descriptive analysis, such as outlining the defining factors of a community and any patterns that exist within that population. While the advances towards population health are welcomed, it is important to reinforce the integrity of data utilized due to potential limitations caused by variation in data standards, data.

Health Equity

While the benefits of BDA have the potential to help everyone, in reality, progress is not experienced by all. Marginalized and minoritized groups continue to experience health disparities even with these methods. Stakeholders can directly improve health equity by proactively and consciously using big data analytics to do so. With the vast amount of data offered by health records, medication lists, social media, wearables, and so on, it should be standard practice to incorporate patient demographics and social determinants as important factors.^10,11 

A common concern is that AI will have the opposite effect on health equity, increasing the bias experienced by patients via algorithms that perpetuate disparities. Bias can occur anywhere along the process, from data collection to model design and interpretation.¹¹ However, the effects of bias can be mitigated through improvements in the use of biased language, development of approaches to identify bias in results, using representative datasets, and validating algorithms across multiple health systems and datasets.¹¹

It has been established that AI utilizing BDA is able to correct for human biases that may arise while treating patients. Findings from a 2021 article regarding osteoarthritis highlighted that when providers used an algorithm developed through machine learning on existing radiological data from diverse people, they had the potential to improve the diagnosis and management of knee pain. By challenging human biases that may occur during a patient’s assessment, they reduced racial disparities.¹²  

Impact to Payers and Pharma

For many payers, the pandemic has brought a renewed focus on artificial intelligence, real-time data analytics, and predictive modeling to help successfully steer through this ever-changing health care landscape. Some areas in which payers are leveraging artificial intelligence and BDA include gaining information and insights into many aspects of operations and patient care, forecasting shifts in membership churn and mix, and monitoring prescription choices and behaviors.¹³  

In the medication area specifically, key payers such as CVS Health, SCAN Health plan, and many others are leveraging machine learning/AI models and BDA to see which therapy regimens are providing the best outcomes at the best costs for certain chronic conditions like cancer and diabetes, identifying at-risk patients who may have adherence challenges based on geography and other social factors, identifying patients with potential drug interactions or on inappropriate therapies, and much more. With these capabilities identified in real time, payers can work with providers to leverage more precise tools, guidelines, programs, and processes to optimize patient outcomes.¹⁴

Payers and providers are also using BDA to gain insights into patient outcomes and cost of care based on therapy regimens. Therefore, pharma will increasingly be expected to demonstrate product differentiation from multiple angles including clinical, economic, relevant real-world evidence, genomics, and other factors to hone in on the most appropriate patient populations for their therapies. When done successfully, this may be the competitive edge that allows pharma to succeed in the ever-evolving BDA health care landscape. 

Pharma is also leveraging BDA all along the development pathway. From increasing the speed of product development, which allows therapies to get to market faster, to predicting drug efficacy and side effects, BDA is improving the work of clinical development, and in a few cases finding new uses and indications for current therapies. 

Leveraging BDA in health care has the potential to revolutionize the ecosystem and positively change the trajectory of care. Organizations (payers, health systems, and pharma) who are not embracing and maximizing AI and BDA in their operations and care delivery may just see competitors who are move right past them and seize dominance and control of their market space.  

References

1. Johnson KB, Wei WQ, Weeraratne D, et al. Precision medicine, AI, and the future of personalized health care. Clin Transl Sci. 2021;14(1):86-93. doi:10.1111/cts.12884

2. Davenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future Healthc J. 2019;6(2):94-98. doi:10.7861/futurehosp.6-2-94

3. Noorbakhsh-Sabet N, Zand R, Zhang Y, Abedi V. Artificial intelligence transforms the future of health care. Am J Med. 2019;132(7):795-801. doi:10.1016/j.amjmed.2019.01.017

4. Dash S, Shakyawar SK, Sharma M, Kaushik S. Big data in healthcare: management, analysis and future prospects. J Big Data. 2019;6(1):54.

5. Gamache R, Kharrazi H, Weiner JP. Public and population health informatics: the bridging of big data to benefit communities. Yearb Med Inform. 2018;27(1):199-206. doi:10.1055/s-0038-1667081

6. King RJ, Garrett N, Kriseman J, et al. A community health record: improving health through multisector collaboration, information sharing, and technology. Prev Chronic Dis. 2016;13:E122. doi:10.5888/pcd13.160101

7. Raghupathi W, Raghupathi V. Big data analytics in healthcare: promise and potential. Health Inf Sci Syst. 2014;2:3-3. doi:10.1186/2047-2501-2-3

8. Harvey C. What Is data analytics? your guide to data analytics. 2021. eWeek. Published March 14, 2022. Accessed March 30, 2022. https://www.eweek.com/big-data-and-analytics/data-analytics/

9. Morse S. Physicians build bundled payments system for breast cancer. Healthcare IT News. Published January 22, 2018. Accessed March 30, 2022. https://www.healthcareitnews.com/news/physicians-build-bundled-payments-system-breast-cancer.

10. Zhang X, Pérez-Stable EJ, Bourne PE, et al. Big data science: opportunities and challenges to address minority health and health disparities in the 21st century. Ethn Dis. 2017;27(2):95-106. doi:10.18865/ed.27.2.95

11. Cutillo CM, Sharma KR, Foschini L, Kundu S, Mackintosh M, Mandl KD; MI in Healthcare Workshop Working Group. Machine intelligence in healthcare-perspectives on trustworthiness, explainability, usability, and transparency. NPJ Digit Med. 2020;3:47. doi:10.1038/s41746-020-0254-2

12. Ibrahim SA. Artificial intelligence for disparities in knee pain assessment. Nat Med. 2021;27(1):22-23. doi:10.1038/s41591-020-01196-3

13. Chaturvedi S. 7 ways predictive analytics can improve healthcare. Digital Health Today. Published November 8 2017. Accessed March 30, 2022. https://digitalhealthtoday.com/blog/predictive-analytics/.

14. Waddill K. How payer forecasting is shifting toward real-time data analytics. Published February 21, 2021. Accessed March 30, 2022. https://healthpayerintelligence.com/features/how-payer-forecasting-is-shifting-towards-real-time-data-analytics