Patient Controlled Healthcare

One of the most important aspects of the heath care reform package is in the efficiencies and cost savings from full implementation of interoperable electronic medical records. Up until now, there has been a division of medical record keeping between the HMO/PPO provider networks and the hospitals, with these systems for the most part not interoperable. The data sets that are currently available in the health care system include pharmacy prescription record data, health insurance reimbursement claims data, hospital stay medical records, and physician office medical records. In addition to these, there are government data sets for Medicaid and Medicare payers by fee for service and DRG.

All of these data sets combined provide a 360 degree view of a person's medical history. But these pieces of information are owned by parties with different vested interests. And presently there is no way to link them all together in one comprehensive longitudinal medical record for a particular identity record. This ambiguity benefits individuals by retaining some measure of privacy because if federalized then these records would become the property of the government rather than the individual. The question then becomes how do we achieve the efficiencies in health care that we desire through IT enablement without ceding our freedom and our privacy.

One model for electronic medical records is the integrated health system model, exemplified by Kaiser and the VA Medical systems. These institutions control and disease manage a population and integrate ambulatory and hospital care in a seamless system. In the case of the VA the control is derivative of military service. In the Kaiser model, it is a non-profit with voluntary participation. But since they have experience with the end-to-end data captured across the entire health care delivery, their systems are more advanced and yet not flexible enough to be used in a more hybrid system of care, where patients would have portable records under their control to take from one site of care to another.

The framework for a portable, tamperproof, encrypted, interoperable record could be based on an extension of some emerging models for personal health records (PHR's) that have been introduced by Google and Microsoft, as well as an open source platform developed at Children's Hospital in Boston and funded by Intel and others called Indivo. Given the urgency of unifying the electronic records to move forward, it may take too long for an open source solution to mature. So for the first generation of this re-engineered process, we should look to Google and Microsoft to harmonize their APIs and provide a choice so that there is a check and balance in the system.

The interfaces to these records could be a standard Application Progamming Interface usable by all health systems in the United States. They would need to add the flexibility for the patient to open selected parts of the record to each institution on a need to know basis. And the central servers for these would need to be put into some neutral framework not controlled by a corporation.

This approach would give us time to integrate the IT across the system while developing a more complete second generation solution to include the genomic data, which will need to have a higher level of security. Once our genomes are sequenced in the next decade it will be especially important to have strong controls on the disclosure by patients to third parties. Data can be grouped into categories based on their impact for medical treatment. Some genomic data will be useful prospectively to predict disease susceptibility and guide disease management programs. Other data such as susceptibility to drug response will need to be kept as current drug allergy data is currently maintained -- for example, so that an emergency hospital visit would trigger opening the data.

The current state of electronic medical records is such that most doctors do not use them because of the amount of time needed to be invested to integrate them into practice. In addition to the software needed to provide secure interoperable records, there will need to be some hardware components optimized for physician office workflow and for patients maintaining an electronic key to their records.

These measures are more complex than just letting the government take over the process. So in countries where socialized medicine is in place, the one size fits all model with no control by the patient has allowed greater efficiency in the short term. But there is no reason why we cannot have a hybrid system that provides a baseline of critical care combined with a broader spectrum of alternatives beyond this, determined by the individual patients' values, stage of life, and economic situation. Altruism aside, we need to plan for how the government can actually pay for the expanding the coverage to uninsured or temporarily unemployed individuals at reasonable rates, and avoid drafting regulations that seem to create greater efficiency but limit the choices in the system for all.