The following post is the second in a two-part series on storytelling, healthcare, and data from SXSW Interactive. The first part can be read here.
Human health is the next battleground for advances in Big Data. From understanding the human genome to leveraging electronic medical records, the opportunities to improve our healthcare with data management and visualization is one of the most discussed topics this year at SXSW. Healthtech at SXSW Interactive has taken many forms, from start-ups like 23andme, which will map your genotype for only $99, to businesses at scale like GE Healthymagination, or venture accelerator RockHealth (who seems to be driving a lot of innovation in the space).
The healthcare provider community, including doctors, nurses and the organizations, like hospitals and clinics, that support them are in the middle of a massive overhaul of the methods and mechanism with which they interact with data. We would like to suggest an approach to includes the patient perspective, their raw data, and the public ecosystem.
Yesterday, Jon Freach described a methodology by which patients can construct coherent self-narratives of the history and state of their health. This methodology has the potential to reduce the emotional stress experienced during the treatment process and result in the best possible treatment outcomes. By giving the patient the ability to construct their own health narrative, this tool supports the current shift in the healthcare relationship from a provider/payer-centric model, focused on procedures and costs, to a patient-centric model, focused on health. The words “patient-centric” and “self-constructed” are likely to give many doctors (and a few hospital administrators) fits, as the mere concept of enabling a patient to own their diagnosis and treatment threatens to disrupt the established healthcare machine.
But do clinicians have a choice in the matter? In a connected world, the patient is armed with more information than ever before. And patients seeing multiple doctors for different aspects of the same condition often quickly realize that they possess more knowledge than any one of their individual doctors. I saw this growing up as my mother, who suffers from a chronic nerve disease, often would be seeing up to eight different doctors at the same time. In her case, she was the only one who could tie the story of her condition together. So, if the move to a patient-centric universe is a likely outcome, how can we give clinicians the appropriate tools to maximize healthy results and maintain a level of acceptable control?
Start with the Patient Narrative
You can imagine these patient narratives may be short and straightforward, in the case of a one-time event like a broken bone or cut requiring stitches. You can also envision that the narratives may be long and winding, in the case of a chronically ill patient dealing with multiple ailments over an extended period of time. My mother’s large Rubbermaid tub on wheels, containing all of her medical records that she lugged to every doctor’s appointment, attested to the likely complexity of many patient narratives.
Complexity is all the more reason to embrace the patient narrative right off the bat, as the care community could glean more information from a well-constructed narrative, than from a 1-minute glance through a chart and a 5-minute line of questioning. By building off of a pre-constructed, and hopefully validated, testimonial, both patient and physician can more easily ensure that all important pieces of information included. It is likely that a well-visualized and accessible narrative can surface the appropriate level of detail to the care provider, saving time and reducing opportunities for misunderstanding. This narrative, visualized at a workstation or printed on a piece of paper, has the potential to improve the common “differential diagnosis” approach, where a physician proceeds with a line of questioning to rule out a checklist of diagnostic options in his head.
Dr. Daniel Kraft (Executive Director, FutureMed) discussed many of these topics at his panel discussion, mHealth: Take Two Apps & Call Me in the Morning. Kraft raised the idea of a personal health record that would supplement an Electronic Medical Record. He also called out companies like ginger.io, that can help put patient-generated data into this personal record in a graphically digestible form for clinicians.
Supplement with the Electronic Medical Record
The second set of information at the provider¹s fingertips should be the patient’s underlying electronic medical record (EMR). While the patient may be best equipped to tell their overarching narrative, it is the minute details in the medical record, that the patient may consider mundane or inconsequential, which can lead a physician to a new discovery. In the case of my mother, it was data about her teenage years that first tipped her endocrinologist to a pituitary disorder that went undiagnosed until she was in her 50s.
SXSW panelists discussed the difficulty healthcare providers have in accessing EMRs, which often come in different formats and on isolated platforms. In fact, it seems the rise of EMRs have actually reduced efficiency in provider organizations. Using the patient narrative as the primary diagnostic tool and the EMR to support it could eliminate inefficiencies encountered when the process starts with wading through the jumbled EMR network.
Ideally, electronic medical records could be cross-referenced and compared to one another to improve care based on similar demographic information or medical histories. However, since privacy regulations restrict the use of records stored on the Health Information Exchange (HIE), care networks will need to be diligent in capturing and utilizing the information they can capture on their end, such as the doctors a patient has seen or the treatment a patient has received within their specific facility. For providers with significant records, like hospital systems or private practices with a long history of patients, software could be used to a cross-reference a database of shared patient narratives to reveal patterns between seemingly different patients and suggest opportunities for clinical trials based upon previous successes.
Leverage the Data Commons
Staying current on cutting edge findings and treatments is a task for all doctors. I have two doctor friends who rigorously stay current on relevant research in their domains, but this behavior is unlikely to be the norm across all of their peers. Combine that with the limited time that doctors can spend with each patient and a large need emerges to put the best publically available data at a doctor’s fingertips. Diagnostic supplements, like Isabel are available and used by some of the world’s most accomplished doctors. What is needed is the next level of supplementary tools and services that can automatically surface the relevant public data based on the patient narrative and underlying medical records.
To date, the fault in providing this information is not the lack of data. The FDA maintains clinicaltrials.gov<http://clinicaltrials.gov/> to make information from all registered clinical trials accessible to the public. All records from the Human Genome project and the Cancer Genome Atlas are in the public domain and theoretically available for use. PubMed is a widely used platform to find data on clinical questions. The problem is not the availability of data, but the tools to make sense of it. Technology is needed that can use the patient information and then scour the data commons to provide a handful of studies that are relevant at that moment. Eliminating the need for fruitless searching and hours of online research will drastically improve the diagnostic process from the patient perspective and reduce doctor-patient interaction time to keep costs down for care networks.
Dr. Kraft cited the emergence of data commons tools, embodied in IBM’s current partnership with Sloan Kettering, where IBM will employ Watson intelligence to expose Sloan Kettering’s professionals to the most relevant cancer research to develop an “individualized cancer diagnostic and treatment and recommendations for their patients based on current evidence.”
Supplement the Clinical Team
As providers look to maintain service levels, but reduce costs, we see a growth in Nurse Practitioner time and a reduction in physician responsibilities.This is largely a good thing, as patients can quickly access help and that help comes in the form of an NP who is specifically trained to interact with and listen to patients. But the responsibility for digesting an increasingly large amount of patient data still falls on the physician. My mom recounts a story of a doctor she saw who specialized in complex diseases. Upon seeing her rolling plastic tub of data he asked, “you don’t expect me to read all of that do you?” So, how will the exponential growth of health data be managed within the system?
As the processors should be doing the heavy lifting, the need is likely not for more doctors or nurse practitioners, but for more data specialists, working in-house with health care providers. Data specialists are common in research settings, where they are well-versed in applying statistical models to medical data to help researchers identify statistically significant events. In the care organization, a data specialist would help structure and provide the relevant data streams for each domain or area of expertise, as well as providing for the integration for patient generated material, like a health narrative or other documents brought in from other facilities. With data specialists designing and improving the clinician’s dashboard, the care community would be able to more accurately recognize patterns in their patients and optimize care.