Once a taboo, conversations about mental illness are increasingly taking a center stage in the public’s eye. I myself read a number of reports on the subject and some statistics caught my attention. The National Alliance on Mental Illness estimates that in the US between 25% and 40% of the 62 million Americans suffering from mental illness will be incarcerated at least once in their lifetime. Among the current jail population approximately 35% of inmates suffer from mental illness. In average it takes between 6-8 weeks for an inmate that enters the system to be identified and receive treatment. During this gap, these individuals are lost in the system. In many cases they stop taking medication and end up committing aggravating offenses. Both the human and the monetary cost in the present system are staggering. Last year alone the cost of housing and care for mentally ill prisoners topped $15 billion.
The challenge in improving the system lies not only in the profound shortage of experienced clinicians, but also in the volume of individuals needing care. Fortunately, technology advances in data processing and analysis can identify problems and solutions, helping to mitigate this crisis. For example, if historic clinical data about patients were easily accessible in a digital format, healthcare providers could mine patient data and use statistically-based predictions to prioritize care to determine which individuals are at the highest risk and require an escalated level of care. This “predictive intervention” would give providers the ability to determine whether an individual needs incarceration or inpatient care; hospitalization or outpatient treatment.
To achieve the desired result requires a careful selection of the right technology. There is a growing demand for services that simplify the management of mental health data, effectively tracks individuals, coordinates care across networks, and provides frontline decision support. The technology solution must be able to aggregate large data volumes from different sources as well as consume and analyze broad data, which includes text, images and more.
To accomplish its mission of providing “preventative intervention” to health organizations Stella, a HarrisLogic technology, has chosen SAP HANA. They have partnered with EV Technologies for the implementation and have joined the SAP HANA Idea Incubator program. This project uses SAP HANA to deliver analysis and decision making based on big data to the next level.
SAP HANA played a critical role in enabling end-users to consume data from virtually any source with interactive access to predictive models, which run against the Stella data universe. SAP HANA in-memory database allows for the analysis of data in real-time. Its predictive engine not only allows for the creation and maintenance of models, but also sends proactive alerts to care givers with the information required to provide the right level of care to its patients.
This means any provider who uses Stella, from a doctor’s office to a prison psychiatrist, can analyze a patient’s profile and accurately create and fine-tune a personalized care plan, live and in-person. With SAP HANA’s predictive capabilities , Stella’s end users can make informed data driven decisions that are guided by sound predictive models. The predictive care service allows providers to intervene before an individual experiences a crisis, thereby lowering costs, improving the quality of care and literally saving lives.
In this POC, the Stella Predictive Intervention service was focused on recidivism, or the likelihood that an individual will return to jail. Here, Stella took mental health history from several different sources and paired it with criminal data to predict who would return to jail within 6 months. The ultimate goal of this project, Stella 3.0, powered by SAP HANA predictive analytics is to help achieve a substantive and long lasting improvement of mental health services in America. Stella plans to shift its current architecture and create an in-memory-powered application on SAP HANA, reducing existing latency due to growth and the evolution of history data.
Today more than a decade worth of case notes exist and are still sitting untapped, mainly consisting of unstructured data. The text analysis capabilities in SAP HANA can be invaluable for the future consumption and stratification of this data, enabling users to categorically score and better service millions of patients.
SAP HANA Idea Incubator program is recently restructured to accelerate the innovation PoC projects built on the SAP HANA platform with optional technology training sessions, sample codes, small financial sponsorship, or additional marketing opportunities. This new structured program is running in the pilot phase. If you have a project and you would like to participate, please contact firstname.lastname@example.org. We will launch the official SAP Idea Incubator website soon when it is ready.