Economists are beginning to lose faith in technological progress. As one wag puts it: instead of cancer cures, “Captain Kirk & the USS Enterprise, we got the Priceline Negotiator and a cheap flight to Cabo.” Even formidable companies like Google have fled the health field, daunted by the complex legal environment. Some have called for radical deregulation as a solution. But a more viable approach is to turn to the work of some of the smart, committed, and impartial legal scholars who are pioneering the field of cyberhealth law. Particularly instructive is Sharona Hoffman & Andy Podgurski’s article, Improving Health Care Outcomes through Personalized Comparisons of Treatment Effectiveness Based on Electronic Health Records.
In an information economy, even cheesecake can be optimized using data-driven methodology. Unfortunately, leading health care providers often resist such methods of improvement. Pharmaceutical firms have sometimes continued to market drugs even after reports emerge that undermine the rationale for taking the drug, let alone paying for it.That troubling method of attaining short term profits at the cost of long term sustainable business models needs to be countered by sophisticated methods of analyzing (and disseminating) data on the real effect of medical interventions. Hoffman and Podgurski help develop a legal and technical framework for assuring that happens.
The President’s Commission Advising on Science and Technology has endorsed aggressive use of health data to ensure new research opportunities. The PCAST authors conclude that many clinical research studies today are “out of date before they are even finished,” “burdensome and costly,” and too narrowly focused. They endorse health information technology that is enabled for “syndromic surveillance,” “public health monitoring,” and “adverse event monitoring” by aggregating observational data.
The free flows of data elevated to constitutional status in the case of Sorrellv. IMS Health Inc. may eventually improve pharmacovigilance, including efforts to understand the effectiveness of drugs on a population-wide level, beyond clinical research. But it will take a great deal of computing power for them to do so. Cyberlawyers will need to rethink how privacy, IP, and health law interact in order to help researchers and physicians make the most of the oncoming data deluge.
Hoffman and Podgurski have detailed how advanced programs of observational research on effectiveness could work. They explain the benefits of personalized comparisons of treatment effectiveness (PCTEs), a form of personalized medicine, that uses information obtained through a large database search to “find a cohort for a patient needing treatment.” Their proposal for new forms of personalized medicine takes to the individual level what has often been envisioned for population-wide analysis:
We propose the development of a broadly accessible framework to enable physicians to rapidly perform, through a computerized service, medically sound personalized comparisons of the effectiveness of possible treatments for patients’ conditions. A personalized comparison of treatment effectiveness . . . for a given patient (the subject patient) would be based on data from EHRs of a cohort of patients who are similar to the subject patient (clinically, demographically, genetically), who received the treatments previously and whose outcomes were recorded. (P. 425.)
As they explain, such a database query could identify “for a given patient, an appropriate reference group (cohort) of similar, previously treated patients whose EHRs would be analyzed to choose the optimal treatment for the patient at issue.” Their proposal is a logical extension of an idea promoted in an Institute of Medicine report known as the “Wilensky Proposal,” which called for more targeted comparative effectiveness research. Research has already demonstrated that pharmacogenetic algorithms can sometimes outperform algorithms that consider only clinical factors.
From Transparency to Intelligibility
Of course, there are challenges to this type of research. Systems must move beyond mere transparency to data entry standards that allow for the intelligibility required by personalized medicine. As Hoffman and Podgurski recognize, “the need to code all presenting comorbidities” and to identify “patients who have the specific condition to be studied” is crucial to data quality. There is a tension between untrammeled innovation by vendors at any given time and later, predictable needs of patients, doctors, insurers, and hospitals to compare their records and to transport information from one filing system to another.
For example, one system may be able to understand “C,” “cgh,” or “koff” as “cough,” and may well code it in any way it chooses. But to integrate and to port data, all systems need to be able to translate symptoms, diagnoses, interventions, and outcomes into commonly recognized coding. Competition also depends on data portability: health care providers can only credibly threaten to move their business away from an unsatisfactory vendor if they can transport those records. Patients want their providers to seamlessly integrate records. Hoffman and Podgurski show the necessity of Stage II of meaningful use rulemaking to promote a common language of medical recordkeeping. As they recommended in 2008:
[I]t is necessary for all vendors to support what we will call a “common exchange representation” (“CER”) for EHRs. A CER is an artificial language for representing the information in EHRs, which has well defined syntax and semantics and is capable of unambiguously representing the information in any EHR from a typical EHR system. EHRs using the CER should be readily transmittable between EHR systems of different vendors. The CER should make it easy for vendors of EHR systems to implement a mechanism for translating accurately and efficiently between the CER and the system’s internal EHR format.
There are also important opportunities for standardization in the security field. The discussion can quickly become technical, but the underlying purpose is clear: to develop some standard forms of interacting in a realm where “spontaneous order” is unlikely to arise and network effects (as well as what David Grewal describes as network power) could lead to the lock-in of suboptimal patterns of data storage and transfer.
Better health information technology infrastructures in the United States can enable forms of surveillance that are more rigorous, comprehensive, and actionable in the world of policy, and more user-friendly for patients. Rather than getting between doctor and patient, advanced EHR stands poised to silently monitor and improve their relationship. The same record systems that are designed to digitize health diagnoses and interventions can also generate outcome data if they are configured appropriately. Such data would help ensure patients and authorities are truly informed about the risks and benefits of drugs.
Hoffman and Podgurski are among the first legal academics to convincingly merge literatures of health system transformation and cyberlaw. They suggest the practical feasibility of productivity gains in the health sector that we usually associate only with Silicon Valley. Just as U.S. Department of Homeland Security (“DHS”) and National Security Agency (“NSA”) have advanced domestic intelligence capabilities by querying distributed databases from diverse public and private sector partners, we can now apply such technology toward improving population health. Hoffman and Padgurski demonstrate a “proof of concept” for reallocating more of these technologies from the diminishing marginal returns of seeking an “enemy within,” to fighting the truly pervasive menace of disease.