Transform discharge planning and save lives with data science
As the use of data science and analytics grow in the healthcare industry, they are transforming how providers make everyday decisions and improve the quality of care. Today’s blog post features Shanu Sushmita, PhD, Head of Data Science at KenSci and Affiliate Assistant Professor at the University of Washington Tacoma. Dr. Shanu sat down with us to discuss one of her recent research papers and explain how data science is driving better discharge planning that minimizes hospital readmissions, cuts costs, and ultimately saves lives. Learn more about KenSci’s Clinical Analytics solution here.
After a patient receives care in a hospital, the care teams’ work is far from over. Discharge planning is a difficult, delicate dance between releasing patients too soon and keeping them too long. If a patient is discharged too early, they are at risk for serious complications and readmission. In a single year in the U.S., over 3 million patients were readmitted within 30 days of discharge at a cost of more than $41 billion1. On the other hand, keeping someone hospitalized too long also comes with costs – an increased risk of hospital-acquired conditions, blocking a bed for someone in greater need, and wasting the hospital’s and patient’s resources. The National Health Service in the U.K. found that, while readmissions within 30 days cost £2.4 billion, delayed discharged resulted in the loss of 830,000 acute bed days2. Knowing exactly when to discharge patients and how to optimize their post-hospitalization care is critical to improving the quality of health outcomes and containing the rapidly rising costs of care.
Today discharge planning is too much art and not enough science
With countless lives and constrained healthcare resources on the line, patients and providers can ill-afford poor discharge planning. Unfortunately, individualized discharge decisions that are effective and efficient have been constrained by limited resources and inadequate tools. With care teams’ time stretched thin by numerous tasks, from completing paperwork to treating patients, detailed discharge planning can fall by the wayside. Many care teams also lack the proper tools to assess all of the patient variables and the myriad aftercare options available that optimize release and transition for each individual. Even with plenty of time and powerful tools, predicting readmission risks with a high degree of accuracy, maximizing patient flow through hospitals with perfectly timed discharge, and creating optimal, personalized aftercare plans for each patient has been extremely difficult. As a result, discharge planning for many patients ends up as too much art and not enough rigorous science – with only a few static factors and risks considered. Fortunately, data science is poised to revolutionize discharge planning.
Optimize hospital release and transition with advanced analytics
While care teams only have the time and tools to look at a limited number of variables about a patient to determine the best discharge plan, machine learning can sift through oceans of data to bring valuable insights to the surface. Machine learning automates analytics, enabling care teams to leverage deep analysis of clinical data in their risk predictions and discharge planning, combining factors such as the likelihood a patient will be readmitted in the next 30 days, and the anticipated cost or length of readmittance3. Integrating highly accurate readmission analysis with precise cost and length of stay predictions empowers care providers to maximize efficiency at the population level, managing beds more effectively and treating more people in less time. Plus, with advanced analytics stratifying patients into highly accurate risk categories, care teams can make better decisions for their individual patients at discharge.
Coordinating discharge planning around accurate patient risk stratifications enables care teams to know when to release patients and how to optimize their aftercare to improve long-term health outcomes. If analytics surface that a patient’s likelihood of readmittance is high, doctors can take a deeper look at both the modifiable and unmodifiable risk factors that may cause the patient to return sooner than expected – and intervene. When it comes to discharge, precise risk analysis enables care teams to make data-driven decisions about ordering additional clinical tests, setting follow up appointments, organizing check-ins to ensure adherence to lifestyle changes, determining the appropriate aftercare facilities, or deciding the patient isn’t ready for transfer and extending their hospitalization. With machine learning surfacing granular risk predictions, care teams are empowered to effectively leverage different degrees of care in their discharge plans and maximize for individual health outcomes as well as hospital efficiency.
Transform discharge planning today with KenSci Clinical Analytics
KenSci Clinical Analytics, built on Microsoft Cloud Technology, is the world’s first vertically integrated advanced analytics platform for healthcare, leveraging 180+ pre-built machine learning models. The solution not only delivers highly accurate risk stratifications, it also operationalizes them to enhance discharge planning with data-driven recommendations personalized for every patient. Clinical Analytics transforms discharge planning, empowering care teams with detailed actions to prescribe in the care transition process to prevent readmission, maximize efficient, effective care, and ultimately save lives.
Start basing discharge decisions on dynamic analysis that gets more precise every day. Try Clinical Analytics on Microsoft AppSource.
_______________________________________________________________________________________________________
1Hines AL (Truven Health Analytics), Barrett ML (ML Barrett, Inc), Jiang HJ (AHRQ), and Steiner CA (AHRQ). Conditions With the Largest Number of Adult Hospital Readmissions by Payer, 2011. HCUP
Statistical Brief #172. April 2014. Agency for Healthcare Research and Quality, Rockville, MD.
http://www.hcup-us.ahrq.gov/reports/statbriefs/sb172-Conditions-Readmissions-Payer.pdf
2National Audit Office (2013), Emergency admissions to hospital: managing the demand, https://www.nao.org.uk/wp-content/uploads/2013/10/10288-001-Emergency-admissions.pdf
3Chun Pan Hon, Mayana Pereira, Shanu Sushmita, Ankur Teredesai, and Martine De Cock. 2016. Risk Stratification for Hospital Readmission of Heart Failure Patients: A Machine Learning Approach. In Proceedings of the 7th ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics (BCB ’16). ACM, New York, NY, USA, 491-492. DOI: https://doi.org/10.1145/2975167.2985648
![[4:56 PM] Jordan Davis (Wipro Designit Services, Inc.) Green background with colored swoops on the top and an image of a young adult helping older adult with a computer.](https://www.microsoft.com/en-us/industry/blog/wp-content/uploads/2024/04/MSFT_Azure_APR08_328468_Blog_Blogheader_240408_V01-1-300x169.webp)
