How can big data and analytics help to improve Population Health? By Dr. Elizabeth Davidson

How can big data and analytics help to improve Population Health?  By Liz Davidson, Ph.D.

The population health movement has been driven, in large part, by the shift from fee-for-service to value-based payment for healthcare providers. The movement has been fueled by the digitization of clinical health data in electronic health records systems (EHRs), advances in computer processing and in analytics and visualization software, and development of health informatics knowledge. However, whether population health will result in improvements in the health of populations depends on several key factors beyond these information technology (IT) advances.

First is the uptake of evidence-based medicine. In the past, evidence on the efficacy of clinical protocols depended on small-scale randomized clinical trials or longitudinal panel survey data. Now, with the “big data” captured by EHRs and administrative IT systems, healthcare providers can rely on data that are more timely and representative of the population of patients they actually serve. These “big data” can also be analyzed with greater specificity of patient characteristics and health conditions. To improve population health, this knowledge must be translated into clinical practice guidelines and applied in day-to-day practices. Embedding guidelines into EHRs can be an effective method to translate knowledge and affect clinical decision-making.

The second factor is the infusion of an evidence-based approach throughout the organization. Buying data analytics tools, hiring consultants and setting up a small pool of “data scientists” will produce few results unless the majority of clinicians and administrators are fully on board with the program. For decades, businesses have tried to reap the benefits of data analytics, business intelligence, and now data science, often with little to show from these investments. The most effective firms have fully integrated an evidence-based analytics approach into the organizational culture from top management down. This will be true for healthcare organizations and population health programs as well.

Third, the health of a population cannot be improved without the engagement of the members of that population, that is, patients. Patients interact with a broad spectrum of healthcare providers, third-party payers, health behavioral coaches and even mobile health apps. Each setting offers opportunities to “nudge” patients towards healthier self-care practices and also generates data on patients’ health engagement and status. To realize the full potential of “big data” and analytics requires data interoperability and aggregation challenges be met as well as effective collaborative partnerships across the care spectrum. Accountable Care Organizations (ACOs) are tasked with these dual challenges to Population Health.

Finally, linking Population Health with providers’ financial reimbursements and incentives must be done with care. Financial incentives tied closely to performance metrics drive self-interested and short-sighted behaviors as well as desired behaviors. In the 1990s, some Health Maintenance Organizations (HMO) focused on cutting costs at the expense of quality and patient access. This led to patient and provider revolts against managed care practices. Today, sophisticated analytics tools are available to utilize healthcare’s “big data.” However, third- party payers must work cooperatively with health provider organizations and patient advocacy groups to develop balanced, meaningful metrics to ensure that population health programs indeed improve the health of the populations served.

Published in Population Health, Volume 5, Issue 8, Aug, 2018