Transition of Care Actionable Data (ToCADA) with Deep Learning (DL) to predict real-time admissions and reconcile discharge plans in chronic care.
During outpatient/clinical discharge process patient’s to home, or subacute nursing homes, caregivers or relatives get a complex printed instructions (drug regimens are only followed at 40% at best, and are root-case of 70% of non-reconciled drugs regimens detected in re-hospitalized preventable visits.
These revisits are detected in 20% of Medicare readmissions within 30 days, for an estimated $15 billion in annual Medicare spending).
Another weak points of peak vulnerability with respect to patient health needs points of peak vulnerability with respect to patient health needs include lack of 48 discharged summaries availability, forgotten flag verbal recommendations when close to time of discharge, and lack of continuum of care indicators to be predicted in the community after hospital stay.
These drawbacks of transition of care not only are focused under the poor patient/caregiver to clinician (sometime nurse) personal communication, but also in the perfect timing to schedule primary care visit (addressing the correct time for first visit can lower rates of follow-up visits in some pathologies that had a higher risk of 30-day readmission) In fact, Medicare beneficiaries readmitted to the hospital within 30 days of discharge, only 50 percent had been evaluated by a physician in the ambulatory arena with a had a ten-fold- greater risk for readmission.
The lack of transition of care continuity of flow limit the possibility to anticipate possible clinical scenarios that may manifest over the weeks after discharge, along with recommendations for adjustments to the treatment plan that in many times are performed retrospectively. It is known that a drug low adherence, and misinterpretations in follow-up care instructional care plan can lead poor outcomes, high antibiotic resistance rates, and many preventable revisits with unnecessary costs especially in chronic diseases.
From an economic perspective, inadequate transitions in care can be quite costly. Under the Medicare Hospital Readmissions Reductions Program (HRRP) established in the Affordable Care Act, financial penalty will be imposed on hospitals with excess readmissions. The bottom line is that there are many assess patient risk readmission tools and risk stratification tools found highly effective to assess patient’s retrospectively over the fragmentation in the transition from hospital to home, but there are not a hand offs solution between hospital, ambulatory predictor of care to be focused on anticipating the likely points of peak vulnerability with respect to patient health needs upon discharge.
The size of transition care in the US, it is about 34.4 billion dollar. In US the healthcare spent is almost 18% of GDP which is one of the highest in the world and out of the entire healthcare spent, a good 12 percent is actually spent on post acute care or post hospitalisations. Within this 12% again a good 7 % is spent on transition care which is largely a combination of in-house rehabilitation facility and skilled nursing facilities, so market opportunity is quite large, since the transition care market size is a derivative of the overall healthcare span ($2,4 billion dollar).
The United States is estimated to have a population of 328,863,150 as of October 24, 2018 As of 2017, people are distributed by age 55-64 years: 12.91% and 65 years and over: 15.63%, for a total of 28%, which are the main of population that use Transition of Care. Among hospitalized patients 65 or older, 21 percent are discharged to a long term care or other institution.
Approximately 25 percent of Medicare skilled nursing facility (SNF) residents are readmitted to the hospital. Individuals with chronic conditions—a number expected to reach 125 million in the U.S. by 2020—may see up to 16 physicians in one year. Between 41.9 and 70 percent of Medicare patients admitted to the hospital for care in 2003 received services from an average of 10 or more physicians during their stay.
There are main three solutions in the market:
The Care Transitions Intervention (CTI), is primarily a transitions self management model that provides coaching, skills and tools to help patient’s and caregivers assert a more active role during transitions. This intervention is low-cost, low intensity and has been shown to produce a sustained effect reducing hospital readmissions significantly for five months following the one-month intervention. It is also expected to result in nearly $300,000 in savings for the care of 350 chronically ill adults over 12 months. (The program has been tested with patients that are 65 years or older with poor self-health ratings, multiple chronic conditions, and a history of recent hospitalizations. One randomized study of the program indicated that the annual total intervention cost was $115,856 ($982 per patient). The study also concluded that reductions in utilization of health services led to mean annual cost savings, over and above the costs of the intervention, of $5,000 per patient.)
The Transitional Care Model (TCM), developed at the University of Pennsylvania, establishes a multidisciplinary team that is led by a master’s prepared transitional care nurse (TCN) to treat chronically ill high risk older patients before, during and after discharge from the hospital. This model has proven to decrease preventable hospital readmissions, improve health outcomes and reduce health care costs. A randomized evaluation of the program indicated that the total annual intervention cost was $1,743 per patient, producing a savings, above and beyond the cost of the intervention, of $1,364 per patient.)
Geriatric Resources for Assessment and Care of Elders (GRACE) is a physician/practice-based care coordination model. GRACE is conducted for a long-term/indefinite amount of time and requires a nurse practitioner and social worker.
GRACE has been tested for low-income individuals aged 65 years or older in primary care, including a group at high risk of hospitalization (as determined by the probability of repeated admission risk screen). A randomized study indicated the total annual intervention costs for high- risk patients to be $315,040 ($1,432 per patient). The study concluded the intervention to be cost-neutral for high-risk patients due to reductions in hospital costs.
If I use the same parallel on healthcare spent, we are talking about 34.4 billion dollars market size as far as transition care is concerned at the current rate of spending (7%) but given the current focus on healthcare there is a likelihood that the spending may increase and since the transition care market size is a derivative of the overall healthcare span. (For 2015-25, health spending is projected to grow at an average rate of 5.8 percent per year (4.8 percent on a per capita basis).
A solution for near future
In summary, there is a need to predict with artificial intelligence (AI) a new network to optimize and auto-reconcile individualized discharge follow-up plans called Transition of Care Actionable Data (ToCADA) with the use of Deep Learning (DL) that moves with patients longitudinally over time, in the transition of care between hospitals and patient’s home.
Right now, the principal objective is to evaluate the use of many AI functionalities in the context of your clinical setting and to gather feedback to prioritize feature developments. We’re just getting started and glad to shape many discharged plans.