Prolegomena to Prononymity: What's the Worst that Can Happen?

America needs electronic medical records (EMR). There are plenty of reasons why we are so far behind other nations in consolidating medical data: lack of strong central leadership on the issue, unwarranted faith in markets to produce solutions, and overwhelmed medical professionals who have little if any slack time to put a new system into place. Even as President Obama pushes for investment in EMR, privacy concerns are also slowing down progress:

Lawmakers, caught in a crossfire of lobbying by the health care industry and consumer groups, have been unable to agree on privacy safeguards that would allow patients to control the use of their medical records. . . . The data in medical records has great potential commercial value. Several companies, for example, buy and sell huge amounts of data on the prescribing habits of doctors, and the information has proved invaluable to pharmaceutical sales representatives.

“Health I.T. without privacy is an excellent way for companies to establish a gold mine of information that can be used to increase profits, promote expensive drugs, cherry-pick patients who are cheaper to insure and market directly to consumers,” said Dr. Deborah C. Peel, coordinator of the Coalition for Patient Privacy, which includes the American Civil Liberties Union among its members.

Health IT turns out to be one many areas where a drive for prononymity--that is, the de-anonymizing of records of on- and off-line life--is running up against a wall of wary citizens and consumers. In the health field, I think that resistance is only going to end if we have a robust "backstop" of health care in place so that citizens don't have to worry about losing all coverage if a digital dossier presents them as a bad risk. (Medicaid as presently constituted does not count.) Far from overwhelming the health care system with pent-up demand, universal health coverage may be a prerequisite for generating support for the type of EMR that will provide us all with far better care.

A trend to prononymity in general should be matched with greater commitment to assuring that it won't result in particularly harsh results. For example, people should not be denied a job for being identifiable as a Democrat in a blog post, whatever Monica Goodling thinks. Nor should doctor's notes about a patient's dark thoughts come back to haunt the patient when she or he applies for medical insurance. And if they do, there should be a genuine insurer of last resort available--not the patchwork of Medicaid and charity care that presently leave so many uninsured people falling through the cracks.

That's one reason why I advocate the development of a Fair Reputation Reporting Act, which would allow individuals to know the documentary basis of certain key adverse decisions. I summarize the proposal here:

Reputation regulation has become essential because traditional restrictions on data flows inadequately constrain decisionmakers and important intermediaries (including search engines and bulletin boards). . . . Persistent and searchable databases now feed unprecedented amounts of poorly vetted information into vital decisions about employment, credit, and insurance. Rumors about a person's sexual orientation (or experiences), health status, incompetence, or nastiness can percolate in blogs and message boards.

Even if the First Amendment and anonymity protect the authors of such rumors, affected individuals deserve to know whether certain important decisionmakers rely on them. In limited cases, the intermediary source of the information should also provide the target of a derogatory posting with the opportunity to annotate it. A Fair Reputation Reporting Act would empower individuals to know the basis of adverse employment, credit, and insurance decisions—and to go to their source (and the source of their salience) to demand some relief from digital scarlet letters.

In summary, privacy concerns are only likely to die down if individuals know either 1) that the consequences of a privacy breach are not likely to be severe or 2) that they can find out instances of the improper use of data. In the health care context in the US, neither qualifier holds: the individual insurance market routinely denies care to individuals on the basis of pre-existing conditions, and individuals have little sense of exactly how such determinations are made. Prononymity needs to work both ways: if our health conditions are to be the subject of increasing availability, so too must the decision-making processes that could use that data to our detriment become more transparent.

PS: Market mavens may promote a "Google Health Search" as the optimal solution here. If this 800 pound gorilla can get all the publishers in line to settle their copyright claims, perhaps it has some chance at bringing the medical industry to heel; however, the political power of doctors and insurers dwarfs that of publishers. The concentration of that much data in one company should also provoke some worries.