Person-centered Care - overview
Containing costs while improving outcomes from coordinated social and health care services has been identified as a societal grand challenge for the 21st Century . Multiple studies demonstrate the importance of a more holistic approach, in particular for the most vulnerable individuals [2, 3].
Supporting the most vulnerable individuals requires a view of individuals across multiple stakeholder systems; integrating individual needs, key aspects of their environment, their family (or household) situation and communities where they belong (4). This integrated person-centric care model is key to providing analytical insights to reduce the burden of care coordination for high-cost-high need patients.
To improve clinical and financial outcomes, health must be defined beyond clinical and financial terms , across key aspects of cognitive, clinical , behavioural, social, functional, and even demographics and environments dimensions. We should not under-estimate the challenges posed in an ecosystem of complex legacy systems, involving numerous agencies and a range of proprietary data formats and technology platforms. Despite some progress on standardization efforts, such as SNOMED-CT, interoperability still poses a huge challenge.
Consider the example of an individual social worker who may be responsible for thousands of people . In this environment, providing a care delivery team with timely, concise, relevant, multi-dimensional and fit-for-purpose information about the context, strengths, needs and vulnerabilities of a person, family or community is a critical step in delivering person-centric care.
The Global AgeWatch Index , which measures the social and economic welfare of those over 60, predicts that by 2050, 21% of the global population will be over 60. But what if you could tap the power of cognitive computing to improve health services for this most vulnerable — and costly — patient population?... In here we focus specifically on the elderly and senior communities, and propose two main contributions to a solution to support aging populations by supporting care workers with care planning and delivery, always having the right patient information at the right time:
- Bussiness Case 1: Multi-dimensional care planning to help improve outcomes and reduce costs by taking into consideration multiple variables such as patient experience, accessibility, cost, etc.
We are exploiting open data and customer data to build a Safety Net and data models for individualized better care planning. Linked-Data technologies are used for: (1) making this complex cross-domain information accessible, (2) understand the resources available for each population (needs) and the connections between them (Safety Net of providers, community programs, etc..), and (3) facilitating analytics to obtain the care plans with the most efficacy. Semantic techniques can help make sense of large amount of open data, published by governments and agencies, about providers and facilities in a wide-range of domains (e.g., health, housing, employment, transport). From these datasets that are silo-ed, noisy, heterogeneous, and do not subscribe to any fix schema, entities are automatically extracted to understand the type of organization (hospital, shelter, etc.), the services it offers and other attributes – such as locations, contact details, etc. Well-known and widely used models and vocabularies (211 Taxonomy of human needs, Dbpedia, schema.org, etc) are used to annotate, catalog and link the data, therefore mapping each entity into a linked Safety Net graph.
Baed on the individual and family needs, experts make assessments on what kind of services are needed by the household. From this initial selection, the Safety Net is used to present the relevant providers (private or community-based). Experts can then select and constrain the relevant criteria for each provider, like estimated cost, distance or hospital ratings for given medical specialties. These are used as input to analytics and optimization algorithms to calculate the optimal services and plans, according to the combination of the various criteria and constrains specified by the care team. As output, a ranked list of optimal plans are obtained (a subset of the very large space of possible plans), and the care team can explore, compare and choose the plan with the best combination of providers. The overall purpose is to enable an optimized and well-informed decision making process in the care planning phase
For example, the Safety Net supports the care workers selecting and filtering the relevant relations and attributes across providers to care for Teresa's household, such as cardiology and dementia (neurology) checks, dialysis and home delivered meals for Teresa. Furthermore, family-based community services are also needed for Teresa's daughter, Teresa's main carer, as well as after-school programs for Teresa's grandchild, while Teresa's daughter is at work. For each of these needs, the relevant facets, dynamically selected for the appropriate providers, are presented to the care worker to filter. For example, in the case of cardiology check, these includes different attributes for hospitals, combined across sources, such as: estimated cost, performance score and other quality measures like mortality rates by heart failure. After the facet filtering, the optimization algorithm can calculate and compare different combination of providers and care plans. As shown in the above figure, the plans are sorted based on costs, travel time, travel assistance (that is, if a family member needs to drive Teresa to hospital), performance score, and all other prioritized issues and configurable factors.
- Bussiness Case 2 :A mobile client for the care delivery team to support and reduce barriers for vulnerable seniors to live independently and engage in their own communities for as long as possible
We present a prototype for improving care delivery using mobile and wearable technologies, which aims at delivering timely, customized, and contextual information to support decision making and coordinated on-line response for care workers on the field.
The demo-video showcase an example scenario, for which the system exploits a back-end knowledge-graph to deliver contextually relevant information and to adapt to unplanned situations. A care worker receives the following relevant information about her patient Teresa in her mobile, before an scheduled patient appointment: her tasks on the day before Teresa's FMRI, Teresa's profiles, details about major events in her timeline history, how to get access to her home (through a trusted neighbor), and information about the FMRI center. At the hospital, the FRMI examination is running late, and so the system (after detecting that the care workers mobile phone is still at the GPS location of the hospital), suggests that she may want to cancel the meal delivery for that day.
M. Rigby, P. Hill, S. Koch, and D. Keeling. Social care informatics as an essential part of holistic health care: a call for action. Int. Journal of Medical Informatics, 80(8): 544-554, 2011.
D. Peikes, A. Chen, J. Schore, and R. Brown. Effects of care coordination on hospitalization, quality of care, and health care expenditure among medicate beneficiaries. JAMA: Journal of the American Medical Association, 301(6): 603-618, 2011
- Marmot, M., Wilkinson, R. Social determinants of health. Oxford University Press, 2009
Kodner, D., Spreeuwenberg, C.: Integrated care: meaning, logic, application and implications-a discussion paper. International Journal of integrated care, 2, 2002
- Onie, R., Farmer, P., Behforouz, H.: Realigning health with care. Standford Social Innovation Review, 10:28-32, 2012
The Global AgeWatch Index 2014: http://www.helpage.org/global-agewatch/
IDC Health Insight: http://www.idc.com/getdoc.jsp?containerId=HI251774