December 31, 2014 katharina.janus

Care-Tank expert Peter Zweifel talks about big data and coordinated care

This post is part of a series on the Health Affairs blog related to the 4th European Forum on Health Policy and Management: Innovation & Implementation, to be held in Berlin, Germany on January 29 and 30, 2015. For more information or to request your personal invitation contact the Center for Healthcare Management.

What Does ‘Big Data’ Mean In The Context Of Coordinated Care?
By Peter Zweifel

In economic terms, coordinated care is about vertical integration in the quest of a competitive edge. Just as IBM subjects its computer chip suppliers to rigorous monitoring to ensure a high-quality, high-price product, so too do health insurance companies impose restrictions on participating providers designed to achieve a favorable ratio of patient utility to cost and with it, a competitive ratio of the utility of policy holders to premiums.

This endeavor calls for collating information from multiple sources, which is typical of big data: When does a particular health problem arise? Why? What is the appropriate intervention? Who should provide it? How should it be carried out? Where should it be provided? Answers to these questions are necessary for the implementation of quality assurance programs.

The Parties Affected by Big Data

The initiative to use big data usually comes from health insurers who face pressure from both politicians and their competitors to offer coverage of services as comprehensive as possible at a given premium. If insurers enjoy the freedom to contract with service providers of their choice (which is the exception outside the United States), they face the crucial challenge of matching a given group of patients with health care providers that are willing to treat them at comparatively low cost.
To make this match, health insurers need a great deal of information about their potential patients and about their potential contractual partners. For instance, it can be important to know whether Hispanic patient X has a preference for Hispanic physicians — or whether she prefers white physicians due to personal biases or perceived advantages. Yet, Ms. X may feel uncomfortable when asked to disclose her preferences.

To get around this problem, the insurer may buy information provided by a chain store revealing details about her purchasing habits and favorite brands, data that may be extrapolated to make predictions about her preferences. For insurers who are prepared to invest in this type of analysis, big data holds considerable promise.
Forpolicy holders, big data has benefits and costs. If insurers can use big data to create networks that better meet the needs and preferences of their policy holders, those policy holders are more likely to build a relationship with their providers, which in turn may lead to more effective care, treatment, and maybe even lower costs.

On the other hand, the health insurer might in turn be tempted to sell its own health and health care utilization data to an employer. At the very least, “informed consent” combined with stiff financial sanctions for failure to obtain it is required to render the benefit-cost ratio of big data favorable for consumers.
The most affected party, however, are the health care providers. Promoters of information technology in medical practice have been emphasizing the benefits of record linkage for a long time. Indeed, professional ethics should arguably motivate providers to share information with other clinicians seeing the patient about the patient’s characteristics, diagnoses, and treatments to the greatest extent possible.
Indirectly, big data of this type also can be of an economic interest to them. By avoiding duplicate procedures and unnecessary tests, providers may improve diagnoses, treatment, and cost savings further enhancing their reputation and career prospects.

The Costs of Big Data

However, the costs of big data weigh at least as heavily. In spite of vertical integration in coordinated care, physicians in particular have continued to enjoy a good deal of professional freedom, shielded from the direct interference of payers by the medical directors of clinical departments and group practices financier.
Much data on individual providers has been anonymized and shared only within limited group practices or hospital. With big data, however, the insurer can conceivably purchase data from a network that is willing to sell it and begin matching the characteristics of physicians from that network with similar physicians of its own for ever closer monitoring.
The question of how manageable big data might be when applied to coordinated care is primarily the concern of health insurers, who have a strong incentive to invest in big data. Once insurers and medical directors agree on measures of performance, big data can be condensed so it is more manageable for both of these parties.

However, it is doubtful whether the cost savings are sufficient to overcome the resistance of service providers who seek to retain their professional autonomy. Big data is a double-edged sword for consumers; improved coordination of care comes at the risk of a loss of privacy, which governments may be eager to exploit.