Electronic health records (EHR) and patient portals, which epitomise the digitisation of medical care, are, ironically, major roadblocks for better health care. Much has been written about the shortcomings of EHRs, yet the unmet needs are broader and include not only the objective of control and ownership but also the capacity to search and share records by patients. The announcement earlier this year that Apple has launched a personal health record feature on its Health app that aggregates existing patient-generated data with a user's electronic medical record is a step in that direction.
Dozens of companies offer electronic medical records software — many provide portal websites that give patients access to some of their health information. But in all truth, EHRs are built for business, not for people. While institutions are incentivised to structure medical data in the way that best serves their practices and reimbursement, the needs of individuals are radically different.
Think how easily people interact with tools such as Google and Facebook. Patient-oriented EHR applications should build on consumer-centric technologies that support portability as well as search and share operations. This approach could allow medical services to transfer the files to a personal handle (computer identifier, address), or patients could elect to drag documents or medical images and “drop” them into a folder. Today many people use the cloud, which allows for redundancy, synchronisation, backup, remote access, and the use of mobile devices like phones with full portability. Having access to all data is only a first step—the documents should be easily, intuitively searched and retrieved so that patients could query for any term and find the related health documents or images in milliseconds. The question “when was I last vaccinated against tetanus” or “when was my last chest x-ray taken” becomes a simple query.
In this model, the patient's EHR becomes a searchable collection of non-structured materials.
Social networks are used to post information and interact with friends, family, and colleagues. Features such as the organisation of materials and access to content, creation of groups, and news feed could be models for the management of personal medical data across health-care workers, selected family members, and other concerned people or institutions. At least in the USA, the task of sharing a hospitalisation dataset with one's family physician is arduous or dependent on the hospital contacting the primary care practice.
The individual may have additional incentives to share the data: for research, or perhaps for its economic value. Many patients do not know that their health information—de-identified—is routinely sold to third parties for a profit. These third-party organisations receive all of the benefit from aggregating and distributing others' data in a suboptimal way that prevents deriving the greatest value from them. What if, instead, de-identified medical data could be shared or sold by the owner. There are multiple companies that provide financial compensation for medical and genomics data. Many of the companies promote the use of blockchain technology to ensure data provenance and protection. Think also about Uber—where drivers monetise on their car and time. A person might choose to “rent” her or his MRI to a company that needs 1 million MRIs for research, or might opt to provide information on drug response to a manufacturer. In countries with health insurance, a person might trade health data with the insurance company in exchange for a lower premium.
Medicine is abuzz with the promise of big data and artificial intelligence. However, this promise fails without all the data for input—deep learning needs deep data. National projects and research consortia will progressively solve many of the bottlenecks of amassing consented data. However, giving a powerful incentive to the individual, whether to altruistically share data for research, to join advocacy and patient groups, or to obtain an economic benefit, is a direct path to generate massive data. It will also prepare the field to face the paradox of personalised medicine: your data is less meaningful when analysed in isolation than when analysed in the context of everybody else's data. Let users manage their medical records using successful models of information retrieval and sharing.
Source : http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30538-5/abstract