We as a healthcare industry have tested the hypothesis that patients want to manage their own healthcare data multiple times over the past decade or so. As of November 20, 2019, Microsoft will be retiring its HealthVault platform. Although other industry groups suggest more patients will be empowered to manage their own health information (e.g., Blue Button, OpenNotes, CARIN Alliance, NATE, CommonWell Health Alliance, Health Record Banking Alliance), patients seem to be reluctant to manage their own health records. The most recent publicized development in health information sharing seems to be Apple‘s move to use tokens to perform specific actions within every health system patient portal the user may have accessed over their lifetime. The company provides the perception of aggregating health information without actually receiving, maintaining, or transmitting protected health information.
Most states have no statute conferring specific ownership or property right to the medical record. Twenty-one states have laws decreeing that the hospital and/or physician own the medical record. Only New Hampshire explicitly states that the patient owns information in the medical record. Even if patients can overcome the health system barriers that challenge them to obtain electronic copies of their health records, there are additional reasons for an individual not to want to maintain their own health records. First, medical records do not lend themselves to easy indexing or methods to reduce duplication. Our existing rules for physician service reimbursement promote additional documentation without value. Maintaining a succinct and accurate medical history for patients with multiple comorbidities or complex medical problems can be challenging. Second, some health events (e.g., exposure to sexually transmitted disease, psychotic break, illicit drug overdose) may not be phenomena that the individual would want to share with their family, friends or other members of their care team. Some of these encounters may threaten an individual’s insurability or reputation. The closest analogy for a third-party storing detailed information about individuals in our society is credit agencies, but the September 2017 Equifax data breach argues against this model moving forward.
What features might a patient (or caregiver) want to see in a system to track personal health information? The system should have the following abilities:
- Capability to both import data from other platforms as well as export data to another platform if the user decides to try another solution,
- Aggregate different data elements together around symptoms, syndromes, and disease clusters (i.e., an ontology) with tools to help increase one’s confidence in the accuracy of a particular diagnosis or the efficacy of a treatment plan,
- Isolate medical problems within free-text documents (e.g., discharge summary, ambulatory encounter) to capture details not available within discrete concepts like diagnosis codes or laboratory test results
- Separate features for initial diagnosis (high-ambiguity) and long-term management of known conditions (low-ambiguity),
- Method to connect patient-generated information to specific symptoms, syndromes an disease clusters (e.g., weight, blood sugar, depression questionnaires),
- Capacity to share relevant health information with a care team to manage a particular medical condition or all records in a medical emergency requiring a more comprehensive view,
- System to track patient preferences around diagnostic uncertainty, tolerance for side effects and other trade-offs as well as willingness to pay to avoid (or guarantee) particular health states, and
- Allow for data anonymization to support data sharing in social networks that may enable behavior change or support medical research without fear of re-identification
Would a patient or caregiver pay for a platform supplying these services? Employers and payers (including state and federal government) may be very interested in funding such a platform if it reduces healthcare costs. From an individual’s perspective, it may be fine if their employer pays for the platform, but if the information includes anything that might threaten their employment or raise their health insurance premiums, user adoption may be low. A union, disease-specific support group or association may be in a better position to offer the platform as a membership benefit like disability insurance, but the data ownership should be kept by the individual. States and federal groups could endorse specific features to be deployed across a political domain as a condition for providing healthcare coverage (e.g., state exchange, Medicare Advantage).
From the platform provider’s perspective, the competitive advantage would need to be focused around how the data is used rather than the data itself. Particular revenue-enhancing functions might include
- Data views that are meaningful to patients (address knowledge gaps, increase self-efficacy, sustain health behavior change), providers (determine the next best action) and payers (identify cost-effective interventions that are likely to be completed by the individual),
- Tools that help users identify, initiate and sustain actions that lead to improved health states (gamification, social sites, nudges to promote new behaviors), and
- Connections to recent scientific findings that might impact a particular disease trajectory.
Healthcare decisions may never be as simple as buying a car or choosing a family vacation destination, regardless of how much data we put in front of patients. Even with this reality, but our existing approaches to health information data sharing have not supported our patients in ways that help them save money or improve their health. We in health information technology may be better served thinking more deeply about our customers’ needs that can be met with new approaches rather than waiting for the next medical (e.g., “-omic”) or technology breakthrough (e.g., blockchain). A fresh look at our healthcare data sharing paradigm might help us on this journey.