July 7, 2009

Last week, we diagnosed healthcare-related data exposure as a serious condition although we refrained from offering a cure. Given the complexity of the case, it is wise to avoid raising hope prematurely. Besides, non compliance is guaranteed unless one consults with all the actors involved. Prudence does not mean letting nature follow its course. Lest the condition become fatal, clear principles should guide the treatment.

The first principle is that the data exchange which must take place during a healthcare related transaction must be held in confidence by each party. As far as physician patient encounters are concerned, this is a long held tradition and indeed the only way to keep the necessary trust. Honest bankers too are bound by their fiduciary duty because possession is not the same as ownership. Data is like money.

Many interpret this principle as freezing the value contained in the data revealed by the encounter. And since progress has made it possible to share data efficiently and hence to release this value, they consider the principle to be so much economic nonsense. Others on the contrary will applaud the principle but agree with their opponents that it closes the discussion. Both positions are erroneous though and much remains to be explored.

To convey what is at stake, think of the European Dark Ages. In the IXth century, whether through a raid by Norsemen and Magyars or a razzia by Saracens, the norm of international markets was plunder. Setting aside the human toll, economists will tell you it proved a welcome stimulus as accumulated treasures were thus put back in circulation. With respect to personal data, the situation today is no different. Yet between hoarding and pillage, haven't later times proven there is a better way?

The second principle is that each party owns one's profile and in particular one's history. This is the first key to unlock economic value for, in opposition with the other parties, held to their duty of confidence, the owner is free to share what he or she owns. If a physician prescribes an anxiolytic to a patient, the patient does not need any agreement to tell someone else of having disabling anxiety nor the physician of having prescribed this anxiolytic one more time.

Together these two principles give an immediate resolution to the vexing issue of physician profiling. A pharmacy does own its sales data but should sell physician prescribing histories no more than it may patient medication histories. Even if for the wrong reasons, New Hampshire is correct to forbid it (1). The same clarity carries over to data robberies. Banks make account holders whole in case of a holdup. Victim of a security breach, a medical benefits management company ought to acknowledge its responsibility towards those whose data it held in trust and may have been stolen. Recent federal law (1) is indeed pointing in this direction.

The third principle is that each party is both entitled to and responsible for, his or her own opinions. We saw last week how online ratings can poison patient physician relations. Unfortunately no solution currently proposed satisfies our third principle.

The Medical Justice "mutual privacy agreement" is problematic because patient consent may well be coerced if not explicitly, at least covertly. "Patients would sign the privacy forms without reading them", Peter Vieth (*) was told by Rod Smolla, "a First Amendment scholar", and patients are at risk to lose their rights. Still, as pointed out by "civil liberties advoca[te]" John Whitehead, the First Amendment may well invalidate any such agreement within "publicly funded health care", leaving those patients free to publish irresponsible slanders.

We are no more sanguine with Rod Smolla's holding more postings by "other patients" as an efficient redress mechanism against unfair criticism. Public controversies sell newspapers and boost rating site readership, rarely do they establish truth. As I suggested last week, physician rating sites should at least carry a twin service to rate patients.

Similarly we deem Byzantine if not frankly hypocritical Angie Hicks' decision to forbid on her rating site "anonymous postings, although the poster's identity is not publicly available". For starters it reduces Rod Smolla's redress to joke status. On a site which fails to disclose contributors' identities backed by due diligence, all positive comments are suspect to the alert reader. Though they look democratic, modern trappings can be as deceptive as an Iranian election. Responsibility is a matter of identity.

Responsible behavior has always come at a price. No doubt the restrictions we suggest would have a chilling effect on public rating sites. But we should not lose the forest for the tree. The demand is for a sound physician recommendation mechanism. The objective of public rating sites is quite the opposite, to let consumers vent in the most vivid discommendations. Between the risks of bottling up and slandering, is there then a better way?

Why not base a solution on our rule of three? Enable patients in search of a physician to list the recommenders whose advice they are ready to follow. Enable grateful patients and independent opinion makers to allow the physicians they like to mention their names when asked. Last enable the patient to whom a physician presents an acceptable recommendation to verify it with the recommender. Carry this triangular exchange online in total privacy, for example with ePrio's technology. For good measure, prevent recommenders from being paid by those they recommend

Jeffrey Segal, the founder of Medical Justice, would hardly find this a threat to physicians. Consumer advocates might object that this approach does not allow discommendations but this is precisely the niche occupied by so-called rating sites. And in the rare cases which warrant it, the consumer can either sue if personally wronged or blow the whistle on the physician to the proper organizations set up to accept and investigate such confidential reports, even if anonymous. And if a cover up is suspected, what are local newspapers for?

While an important topic in itself, the physician rating issue helps us to show how principles can lead to practicle solutions. We are far from done though. Beyond personal opinions, there is much more data to share. Think of the study of bone-growth proteins, whose efficiency is being questioned according to Barry Meier (**). "The research team based its report on records collected through a database known as the Nationwide Inpatient Sample, which gathers information from 20 percent of the nation's hospitals."

If such sharing can be done for the sake of society without harming the parties concerned, Facebook may well be interested too. David Gelles' analysis makes clear how vital it is for the health of the company to "give businesses a new and valuable way to reach [its] users" (***).

For more guidance then, I recommend you come back next week for another fillip.

Philippe Coueignoux

• (*) ..... Doctors try to stop patients from rating them online, by Peter Vieth (Massachusetts Medical Law Report) - Summer 2009
• (**) ... Bone-Growth Proteins Do Poorly in New Study, by Barry Meier (New York Times) - July 1, 2009
• (***) . What friends are for, by David Gelles (Financial Times) - July 3, 2009
• (1) for details, check out the texts listed in Handling of Medical Records from our lectures on Liabilities and Vulnerabilities in the Information Age.