Preventing Long-Term Care Placement: Predicting Who Is at Greatest Risk

View presentations

Nearly two-thirds of Americans over the age of 65 are estimated to be candidates for long-term care (LTC) facility placement. However, alternatives to LTC often exist, and recent studies have shown that many elderly individuals would prefer to receive care in their home or other community-based institutions – rather than in an LTC facility – when possible. As a result, health care providers have begun to look at new approaches to caring for the elderly that avoid unnecessary placements in LTC facilities.

Duke University, Connecticut Community Care, Inc. (CCCI), and Amerigroup have developed a predictive model to identify Medicaid beneficiaries that are at highest risk for LTC placement, as well as examine the participant characteristics and program services which are associated with LTC placement. Given provisions in the Patient Protection and Affordable Care Act to expand Medicaid eligibility and incentivize states that implement programs to reduce unnecessary LTC placements, this model may help inform state and local policy efforts to improve targeted interventions for those highest at risk to enter LTC facilities.

Join the Partnership to Fight Chronic Disease and WellPoint for an in-depth discussion about this promising pilot and to examine strategies and tools for understanding the risk factors for LTC placement and which care management programs may prevent LTC placement – especially among the population of high-risk, elderly Medicaid beneficiaries. 

Featured speakers include:

  • Ken Thorpe, PhD, Chairman, Partnership to Fight Chronic Disease
  • Isao Iwata, MD, PhD, EdM, Duke University (Assistant Professor, Division of Geriatric Medicine, University of North Carolina at Chapel Hill)
  • Kate Massey, Vice President, Public Policy Institute, Amerigroup
  • Sheila Molony, PhD, APRN, GNP-BC, Director of Quality Improvement, Connecticut Community Care, Inc.
  • Matt Salo, Executive Director, National Association of Medicaid Directors


Wednesday, June 26, 2013
11:45 a.m. Lunch available
12 noon - 1:30 p.m. Panel Discussion

Lunch will be served in compliance with the Widely Attended Event exception to the Congressional Gift Ban.


538 Dirksen Senate Office Building

RSVP online or Email


Predicting Long-term Care Placement Among Participants in Managed Long-Term Services and Supports

Understanding Risk Factors for Long Term Care


Connecticut Community Care, Inc.

Connecticut Community Care, Inc. (CCCI) boasts an impressive record in the development of high quality, cost effective solutions to the community long term care needs of frail elders and individuals with disabilities.

Since its incorporation in 1980, CCCI has engineered a comprehensive, inter-disciplinary care management practice. Nurse and social service professionals assess the strengths and needs of the individual. Consumers are empowered to fully participate in this process and in the subsequent development of a holistic plan of care. In FY 2013 approximately 9,000 individuals benefitted from these services. CCCI is the pioneer Connecticut care management organization for the statewide Connecticut Home Care Program (CHCP).

In addition to this publically funded model, CCCI offers customized care management services for any Connecticut resident through our private Care Management Associates (CMA) Division. Attorneys, trust officers, probate court judges are among the most common referral sources.

As an approved access agency for The Connecticut Partnership for Long Term Care, CCCI provides clinical support to long term care insurance carriers and their customers. The Connecticut Housing and Finance Authority turns to CCCI for assessment and care planning services under the Reverse Annuity Mortgage (RAM) program. The CT Chapter of the MS Society contracts with CCCI for care management support for individuals confronted with this devastating diagnosis. A major insurer, ConnectiCare, engages CCCI care managers with their most compromised Medicare Advantage members.

In the absence of comprehensive care management software, CCCI developed CyberCAM, technologically responsive care management system that provides accuracy and consistency in data collection and reporting at a reduced investment of time and money. This software enabled CCCI to collect, manage and interpret the data essential to the development of the Predictive Modeling program.

In addition to the provision of care management and related services, CCCI is a highly credible partner in state and national public policy venues.

In partnership with the Robert Wood Johnson Foundation, CCCI convened a panel; of national care management experts to identify and define care management practice standards. Two decades later, “The Guidelines for Case Management Practice Across the Continuum” is still recognized as a seminal publication in this profession.

In 2012 CCCI was awarded a significant CMS funding opportunity for the Compass to Care (ComPass2C) Program. In nine major hospitals in north central and eastern CT, CCCI coaches complement traditional hospital discharge planning functions with care transition intervention strategies to reduce hospital readmissions.

CCCI has been hailed as “The Best” In Home Care by the Connecticut Law Tribune. CCCI ranked Number One among the 50 Best Mid-size Nonprofits to Work For in the Non-Profit Times Best Places to Work recognition program.

CCCI lives and breathes our mission: Connecticut Community Care, Inc. identifies choices and provides services to help people of all ages, abilities and incomes to live at home.

Academy Health Abstract: Predicting Nursing Home Placement among Participants in a Community-based Care Management Program

Conference: June 23-25, 2013

Melissa A. Greiner, MSa, Laura G. Qualls, MSa, Isao Iwata, MD, PhD, EdMb, Heidi White, MDc, Sheila Molony, PhD, APRN, GNP-BCd, Terry Sullivan, RN, MSW, MSNd, Bonnie Burke, MSe,, Soko Setoguchi, MD, DrPHa,c

a Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina
b Department of Medicine, University of North Carolina School of Medicine, Chapel Hill, North Carolina
c Department of Medicine, Duke University School of Medicine, Durham, North Carolina
d Connecticut Community Care, Inc., Bristol, Connecticut
e Amerigroup Corporation, Virginia Beach, Virginia

Research Objective: Several states offer publicly funded care management programs to prevent long-term care (LTC) placement in high-risk Medicaid beneficiaries. Understanding risk factors for LTC placement and which services may prevent LTC placement can facilitate efficient allocation of resources among program participants. We used administrative databases for elderly participants in a state and waiver-funded home- and community-based services (HCBS) program to develop a prediction model to identify participants that are at highest risk for LTC placement, and examine participant characteristics, and program services which are associated with LTC placement.

Study Design: We used de-identified data for the Connecticut Home Care Program for Elders (CHCPE) participants who had an annual reassessment between 2005 and 2010 and selected the earliest available per person. We analyzed data collected by Connecticut Community Care, Inc. (CCCI) primary care managers on medical history, functional dependency, cognitive ability, medications, social support, and financial assistance; data on service use and facility admission in the prior 12 months; and LTC placement within 180 days or 1 year of the reassessment. Logistic regression models with random effects were used to predict nursing home placement in a 66% random derivation sample; the models were validated in the remaining 34% sample. We evaluated the calibration and discrimination of all models in both samples and refit the models on the entire sample.

Population Studied: CHCPE participants 65 years and older.

Principal Findings: There were a total of 10,975 study participants (median age 75-79 years, 74% female and 74% white). Common medical diagnoses included hypertension (59%), diabetes (33%) and Alzheimer’s disease (20%). Common functional dependencies included bathing (80%), meal preparation (92%), and money management (75%). Over two-thirds of participants were Medicaid eligible. Within 1 year of reassessment, 11.4% (1,249) of clients had LTC placement. Risk factors for 1-year LTC placement included Alzheimer’s disease (OR 1.30 [1.18, 1.43]), money management (OR 1.33 [1.18, 1.51], living alone, (OR 1.53 [1.31, 1.80]), number of prior SNF visits (OR 1.46 [1.31, 1.62]) and English primary language (OR 2.22 [1.66, 2.97]). Participants with monthly medical service costs above the median ($511) had on average 7% higher odds of LTC placement. Use of a personal care assistant service was associated with 46% lower odds of LTC placement. The models were well-calibrated. The c-statistic was 0.79 in the the derivation cohort and 0.76 in the validation cohort. In 6-month models, c-statistics were higher (0.82 and 0.79).

A model using information commonly collected in a care management program had strong discrimination to predict which participants will require LTC placement. The personal care assistant service was effective in preventing LTC placement while higher monthly medical service cost was associated with higher risk of LTC placement.

Implications for Policy, Delivery, or Practice:
Provisions in the Patient Protection and Affordable Care Act expand Medicaid eligibility and offer monetary incentives to states who implement programs to prevent nursing home placement. This study may inform efforts to make community care programs more effective through implementation of targeted interventions and services for those at highest risk.