Healthy Predictions? Questions for Data Analytics in Health Care

• Author: Janine S. Hiller
• Published date: 01 June 2016
• DOI: http://doi.org/10.1111/ablj.12078
• Date: 01 June 2016

Healthy Predictions? Questions for

Data Analytics in Health Care

Janine S. Hiller*

INTRODUCTION

The Patient Protection and Affordable Care Act (“Affordable Care Act” or

ACA),

1

health information technology (HIT) adoption, and increasing

implementation of electronic medical records, are all propelling healt h

care into the world of big data.

2

Big data, analytics, and predictive algo-

rithms are poised to play a large part in the transformation of health-care

delivery in the United States, determining who will benefit and, unfortu-

nately, who may suffer from its insights. Health-care reform depends on

cost savings derived from the application of sophisticated data analytics to

the ever-expanding mass of data collected from and about individual

patients.Healthdataanalyticscanleadtoimprovedcare,newscientific

discoveries, and better medical treatment. Encouraging healthy behaviors,

eliminating health disparities, and addressing the underlying determinants

of health in society are important national goals.

3

It is unclear, however, whether massive data collection about personal

health and individual social status, both within the health-care system

and outside of it, will serve the goal of addressing historical discrimina-

tion in health care, or whether data analytics will lead to the loss of

*R.E. Sorensen Professor in Finance, Professor of Business Law, Virginia Tech. The author

thanks the Center for Business Intelligence Analytics, Pamplin College of Business, Virginia

Tech, for its support ofthis research. The author thanks theparticipants at the 2014 Law and

PredictiveAnalytics Conference at Virginia Tech for theirhelpful comments.

1

Pub. L. No. 111-148, 124 Stat. 119 (2010).

2

See infra text accompanying notes 91–92 for a description of big data.

3

See U.S. DEP’THEALTH &HUMAN SERVS., HEALTHY PEOPLE 2020 FRAMEWORK 1, https://www.

healthypeople.gov/sites/default/files/HP2020Framework.pdf.

V

C2016 The Author

American Business Law Journal V

C2016 Academy of Legal Studies in Business

251

American Business Law Journal

Volume 53, Issue 2, 251–314, Summer 2016

bs_bs_banner

individual privacy, unequal treatment of individuals, and the perpetuation

of health inequality. Data amassed from electronic health records (EHRs),

private sector health website visits, personal health devices, mobile health

applications, and social networks, are being linked together in a big data

environment. Secondary use of health data by employers, insurers, mar-

keters, and others heightens concerns. The collection and use of massive

amounts of data about individuals, fed into a fragmented health analytics

framework, may impose personal and societal costs if not carefully con-

structed. Furthermore, a predictive analytics

4

environment in health care

may affect some groups differently than others, not decreasing health dis-

parities but segmenting populations and resulting in differential care. As

one author succinctly summarized, “An era of ‘big data’ promises exhila-

rating and frightening opportunities to cure and exploit human

vulnerabilities.”

5

Health-care providers and policy makers should ask

hard questions about how harms to personal privacy can be avoided, stig-

mas prevented, and threats of unbridled commercialization ameliorated.

This article proceeds in five parts. Part I provides a policy overview of

health care and data. In order to examine the evolving, complex issues sur-

rounding health data, analytics, and relationships to determinants of health,

equality, and privacy, Part II broadly outlines use of health-care data from a

policy perspective, including the ramifications of health-care reform for the

collection of personal health data. Because the success of reform measures

depends uniquely on the use of health data, Part III reviews a variety of

methods for data collection from public and private sources, mobile devices,

and social media, and then considers its use in health analytics. Part IV

examines the laws and regulations that aim to both protect health data and

patient privacy and to prohibit discrimination in health care.

Part V discusses the dynamic interplay between three aspects of

today’s personal health data environment that create strong pressure on

personal privacy and health priorities in ways that have not been crit-

ically acknowledged. First, policy goals, on the one hand, depend heav-

ily on the use of big data to solve financial costs and population health

problems and, on the other hand, promise to deliver equality in health

care and to protect personal privacy, yet they fail to incorporate specific

4

See infra text accompanying note 142 for a definition of predictive analytics.

5

Frank Pasquale, Grand Bargains for Big Data: The Emerging Law of Health Information,72

MD.L.REV. 682, 684 (2013).

252 Vol. 53 / American Business Law Journal

policies that will do both. Data take priority. Second, due to the empha-

sis on data driven health care, participation of commercial entities out-

side of the traditional health-care industry is mushrooming, as entities

as diverse as data brokers and consulting firms collect and manage

health data from within and outside the health system. Responding to

the call for health analytics can both assist in promoting cures and

deliver efficiencies, but the aggregation and manipulation of individual

health data is occurring in ways that make it impossible for individuals

to control its reach, and laws are inadequate to provide for robust pri-

vacy and antidiscrimination protection.

Lastly, the combination of policy that treats data as a solution and

increasing data fusion threatens to turn health equality into data-driven

discrimination. These thorny problems are not subject to easy solutions,

but recognizing the problems can lead to intentional actions to avoid

the most harmful consequences. Potential avenues for avoiding the

unintended consequences of emphasizing data for health-care decisions,

and for protecting individual privacy, are suggested.

I. HEALTH CARE AND DATA:POLICY OVERVIEW

The U.S. health-care system is known as one of the most expensive and

yet least efficient systems in the developed world, characterized before

recent reform as a disparate have or have-not system without universal

health care for all.

6

Decreasing the disparity in access to medical care

between populations is probably the most commonly discussed aspect of

health-care reform, and it is addressed significantly by extending the

health-care mandate.

7

However, individual health in the United States

also depends to a large degree on social constructs that lurk behind the

health-care system, known as the social determinants of health; these

6

See Lenny Bernstein, Once Again, U.S. Has Most Expensive, Least Effective Health Care System

in Survey,W

ASH.POST:TOYOUR HEALTH (June 16, 2014), http://www.washingtonpost.com/

news/to-your-health/wp/2014/06/16/once-again-u-s-has-most-expensive-least-effective-health

-care-system-in-survey/ (presenting survey data before the health-care reforms).

7

This aspect of the ACA, while critically important, is not the focus of this article. It has

received considerable attention with regards to expanding access to underserved popula-

tions. See, e.g., Rene Bowser, The Affordable Care Act and Beyond: Opportunities for Advancing

Health Equity and Social Justice,10H

ASTINGS RACE &POVERTY L.J. 69, 79–95 (2013) (describ-

ing the opportunities and challenges for expanded care).

2016 / Healthy Predictions? 253