2017 AHC/NCAL Annual Convention & Expo

Using Data-Driven Models to Drive Optimal Post-Acute Care Placement and Reduce Readmission Risk

Tuesday, October 17
8:00 AM - 9:30 AM
Location: Lagoon I&J
CE: Nurses: 1.5 | NAB: 1.5

Preventing readmission following an acute hospital stay benefits residents and has been identified by policymakers and providers as an opportunity to reduce overall health care costs through improved quality. Post-acute care (PAC) providers can capitalize on predictive analytics and multiple sources of data to inform post-acute care placement decisions that will minimize likelihood of readmission based on patient factors, such as length of stay, ICU days, demographic and socioeconomic factors, comorbidities, and functional/cognitive status.

This session will review findings of a large predictive analytic model that used claims data from CMS Medicare Fee-for-Service (FFS) files and a nationally representative sample of Medicare Advantage (MA) data, supplemented with information from assessment tools (MDS 3.0, IRF-PAI, OASIS). Predictive models were developed to minimize readmission risk and overall cost of care within 90 days of acute discharge based on placement in one of four PAC settings: long term acute care hospital, skilled nursing facility, inpatient rehabilitation facility, or home health agency. The models were then used to simulate results of data-driven placement for the top 10 diagnostic resource groups (DRGs). Compared with actual placement, following recommended placement would reduce readmission risk for patients with major joint replacement by 16.6 percent and total 90-day payment amount by $2,589 on average. Readmission risk for patients with septicemia was reduced by 7.9 percent, simple pneumonia 4.8 percent, kidney and urinary tract infection 7.1 percent, and heart failure 5.1 percent. Models of optimal care pathways show that data-driven transitions can reduce readmission risk by 46 percent (from 13 percent to 7 percent) for the top 10 DRGs following placement from first PAC setting, with only a small increase in 90-day cost of care ($2,134 to $2,801 on average). These models are being used by a large post-acute care organization in a new PAC placement platform to optimize first PAC setting and subsequent transitions.

Learning Objectives:

Christie Teigland

Vice President, Advanced Analytics
Avalere Health

Christie Teigland, PhD, Vice President of Advanced Analytics at Avalere Health, is expert in the design and implementation of statistical studies focused on comparative effectiveness, predictive analytics, and performance measure development. Prior to joining Avalere, she served as Senior Director of Statistical Research at Inovalon where she managed quality projects awarded by the Commonwealth Fund, National Committee on Quality Assurance (NCQA), Pharmacy Quality Alliance (PQA), URAC, and other national organizations. In 2014-15, she directed an impactful study investigating disparities in outcomes in dual eligible and socioeconomically disadvantaged Medicare beneficiaries. Dr. Teigland serves on the National Quality Forum (NQF) Disparities Standing Committee and the Pharmacy Quality Alliance (PQA) Quality Metrics Expert Panel. She was appointed to the NQF Quality Innovation Expert Panel on the Impact of Social Determinants of Health in 2017 and invited to serve as an expert panel discussant in workshops sponsored by the National Academies of Science, Engineering and Medicine on their report series “Accounting for Social Risk Factors in Medicare Payment.”

Prior to joining Inovalon, Dr. Teigland conducted long term care research for 15 years with the Foundation for Long Term Care where she directed the development of innovative technology solutions to advance the use of data-driven decision making to improve outcomes and reduce healthcare costs. She served on several CMS expert panels including the Nursing Home Quality Measures Technical Expert Panel, Five Star Quality Rating System Workgroup and field-testing of AHRQ CAHPS satisfaction and patient safety tools, and on national expert panels including a RAND expert panel that prioritized national patient safety measures; AHRQ Care Planning Expert Panel; MDS 3.0 Validation Panel, a CMS funded project to set national performance goals, NYS QIO Steering Committee, and the American Medical Director’s Association (AMDA) Technology Work Group. Dr. Teigland received her Ph.D. and M.S. in Econometrics from the University of New York at Albany.


Send Email for Christie Teigland

Zulkarnain Pulungan

Director, Health Economics Outcomes Research
Avalere Health

Zulkarnain Pulungan, PhD, Director at Avalere Health, is expert in econometric and statistical modeling, predictive analytics, quality measurements and health outcomes. Dr. Pulungan works with a team of healthcare research professionals in the design and implementation of health outcomes studies, comparative outcomes analyses, risk modeling, predictive analytics, performance measurement, survey analysis and disease management evaluation. Prior to joining Inovalon, Dr. Pulungan specialized in quality measurement and improvement research at Leading Age New York and the Foundation for Long Term Care where he developed, validated, and implemented quality measures and predictive models as part of innovative technology solutions to advance the use of data-driven decision making. Dr. Pulungan served as data analyst and evaluator for large health outcome studies funded by the national Alzheimer’s Association, New York State Department of Health, HRSA, and New York State Health Foundation focused on development of quality measures and risk models designed to predict adverse health outcomes along with patient centered risk profiles to guide preventive actions. Dr. Pulungan incorporated these models into internet-based data analytic software readily accessible to providers. He was involved in the development of metrics used in the New York State (NYS) Pay-for-Performance System and investigator in a large National Institute of Health funded health disparity study. Dr. Pulungan teaches an econometrics analysis course at the State University of New York at Albany where he received his Ph.D. in economics.


Send Email for Zulkarnain Pulungan


Using Data-Driven Models to Drive Optimal Post-Acute Care Placement and Reduce Readmission Risk


MP3 Audio

Presentation Slides



Handout 1 Handout 2



Attendees who have favorited this

Send Email for Using Data-Driven Models to Drive Optimal Post-Acute Care Placement and Reduce Readmission Risk