Machine Learning / Predictive Analytics

Success Stories

Health Catalyst

Machine Learning Improves Accuracy of Risk Predictions and Improves Operational Effectiveness

The Centers for Medicare and Medicaid Services (CMS) readmission penalties are a significant concern for healthcare organizations, with over 2,500 hospitals being penalized each year, resulting in CMS withholding more than $500 million in payments.
For Westchester Medical Center Health Network (WMCHealth), hospital readmissions carried more than financial consequences. Care managers had to use multiple systems and time-consuming, manual processes to identify recently discharged patients at risk for readmission. These processes limited the effectiveness of the care management team, as care managers lost valuable time searching patient records for data needed to prioritize their workload and choose the right interventions.
To address this problem, the data analytics teams at WMCHealth and network member Bon Secours Charity Health System leveraged artificial intelligence and machine learning to develop a more accurate readmission risk prediction model that would enable care managers to use their time coordinating and engaging with patients more effectively. Results include:

A risk prediction model that is 17 percent more accurate than widely used readmission risk models in identifying patients at high-risk and low-risk for readmission within 30-days.
Care managers obtain follow-up appointments faster, usually within seven days, and connect patients with the services needed to prevent unnecessary visits to the emergency department and readmissions to the hospital.
1,327 hours per year saved, freeing up care managers to spend more time with patients.

Read more
My Folder
Health Catalyst

Machine Learning Automates Outpatient Coding

Accurate service line reporting is necessary for a healthcare organization to understand its total cost of care. Organizations that do not understand the total cost of care cannot be successful in risk-sharing and other forms of value-based payment, resulting in a loss of reimbursement.
In an effort to reduce costs, MultiCare Health System, an integrated delivery system serving Washington, decided to outsource all encounter coding, which eliminated the coding of outpatient encounters, negatively impacting service line reporting. To ensure accurate reporting, MultiCare asked its coders to assign an MS DRG code to all hospital-based outpatient encounters, which brought significant additional costs. To mitigate this, MultiCare utilized data analytics and machine learning to develop an algorithm that predicts the MS DRG code for hospital-based outpatient encounters.
By employing machine learning, MultiCare has achieved impressive results, including:

Successfully restoring service line reporting, enabling the organization to better understand the total cost of care, and supporting future participation in value-based care and risk-sharing agreements.
Ability to avoid additional labor costs that would be required to perform dual coding, saving more than $1M annually.

Read more
My Folder
Health Catalyst

Artificial Intelligence Improves Accuracy of Heart Failure Readmission Risk Predictions

A global pandemic, heart failure (HF) affects at least 26 million people worldwide, and its prevalence only continues to increase. Within the U.S. alone, 5.7 million adults live with HF, carrying a cost of nearly $30.7 billion each year. At 55 percent, HF represents the most common cause of Medicare readmissions, and HF accounts for 42 percent of total admissions for Medicare patients.
Readmissions for HF carry a heavy cost for patients and health systems, in addition to reimbursement penalties from CMS. This makes properly assessing the risk for readmission for patients with HF a top priority. MultiCare Health System leveraged artificial intelligence and machine learning to improve the accuracy of readmission risk predictions for patients with HF. Providing a more accurate risk score in a timely fashion gives care teams more time to intervene effectively and prevent avoidable readmissions.
Results: 

85 percent estimated accuracy for heart failure readmission risk predictor. (LACE accuracy around 62 percent)
Three-fold increase in the number of HF readmission risk-predictions made each day.

Read more
My Folder
Health Catalyst

Accuracy of Readmission Risk Assessment Improved by Machine Learning

Hospital readmissions carry significant financial costs and are associated with negative patient outcomes. While the reasons behind patient readmissions are multi-factorial, and the specific rates vary by institution, nearly 20 percent of all Medicare discharges nationwide led to a readmission within 30 days. Preventing even 10 percent of these readmissions could save Medicare $1 billion.
North Carolina’s only not-for-profit, independent community healthcare system, Mission Health, is comprised of seven hospitals, 750 employed/aligned providers, and one of the largest Medicare Shared Savings ACOs in the nation. Mission had been using the LACE index to predict risk for readmission, and while it was helpful, Mission’s patient population was different than the population used to develop the LACE index, leaving the health system with some uncertainty regarding the readmission risk of its patients. With the help of data analytics, Mission developed its own predictive model for assessing readmission risk, aimed at preventing readmissions and improving outcomes for patients.
Results:

The area under the curve (AUC) for Mission’s readmission risk predictor is 0.784, outperforming LACE, and meeting the organization’s goal for performance.
Mission’s readmission rate is 1.2 percentage points lower than its top hospital peers.

Read more
My Folder
Health Catalyst

Machine Learning, Predictive Analytics, and Process Redesign Reduces Readmission Rates by 50 Percent

The estimated annual cost of readmissions for Medicare is $26 billion, with $17 billion considered avoidable. Readmissions are driven largely by poor discharge procedures and inadequate follow-up care. Nearly one in every five Medicare patients discharged from the hospital is readmitted within 30 days.
The University of Kansas Health System had previously made improvements to reduce its readmission rate. The most recent readmission trend, however, did not reflect any additional improvement, and failed to meet hospital targets and expectations.
To further reduce the rate of avoidable readmission, The University of Kansas Health System launched a plan based on machine learning, predictive analytics, and lean care redesign. The organization used its analytics platform, to carry out its objectives.
The University of Kansas Health System substantially reduced its 30-day readmission rate by accurately identifying patients at highest risk of readmission and guiding clinical interventions:

39 percent relative reduction in all-cause 30-day.
52 percent relative reduction in 30-day readmission of patients with a principle diagnosis of heart failure.

Read more
My Folder
Health Catalyst

DKA Risk Prediction Tool Helps Reduce Hospitalizations

Each year, more than 12,700 pediatric patients are diagnosed with diabetic ketoacidosis (DKA), a life threatening complication of diabetes. Texas Children’s Hospital sought a way to accurately predict risk of DKA in time for care team members to intervene before these patients suffered a severe episode.
The health system ultimately formed a multidisciplinary high risk diabetes team to devise pre- and post-discharge strategies, and DKA risk prediction tools aided by the Health Catalyst Analytics Platform built using the Late-BindingTM Data Warehouse.
Results:

30.9 percent relative reduction in recurrent DKA admissions per fiscal year.
90 percent of all patients with new onset type 1 diabetes at the Medical Center Campus have a documented RIPGC in their medical chart.
100 percent of patients with type 1 diabetes have a risk index for DKA documented every 6 months.

Read more
My Folder
Health Catalyst

Reduce Readmissions with Predictive Analytics and Process Redesign

With nearly 20 percent of elderly patients released from a hospital being readmitted within 30 days, Allina Health is focused on providing patients optimum care and support post discharge to minimize readmissions. Focusing on 30-day potentially preventable readmissions (PPRs) as its global outcome measurement, Allina Health used key clinical variables to derive the clinical relationships between hospitalizations that determine PPRs. It further built analytic capabilities to identify opportunities for improvement in care management and to test quality improvement ideas.
Allina Health’s multipronged solution included redesigning care management processes, implementing predictive analytics to identify at-risk patients, using analytics to measure the impact of its interventions, and educating patients, families, and clinicians.
These efforts are driving measurable improvements including: 10.3 percent overall reduction in PPRs, 27 percent reduction in PPRs for patients with clinic follow-up within 5 days, and $3.7 million reduction in variable costs due to avoided readmissions.

Read more
My Folder
Health Catalyst

How to Improve Sepsis Diagnosis and Outcomes through Innovative Healthcare Analytics

Sepsis affects more than 750,000 hospitalized patients and results in 570,000 ED visits per year. Learn how this large medical center is tackling their sepsis care challenges by leveraging their EDW and healthcare analytics. They defined and built a sepsis registry and analytics platform in 10 weeks to measure 6 interventions and 4 outcome measures — including mortality rates, length of stay, total hospital stay and 30-day readmits.

Read more
My Folder