A lack of coordination of patients from one level of care to the next, whether to a post-acute care setting or home, can lead to preventable readmissions. According to a 2009 New England Journal of Medicine article titled, “Rehospitalizations Among Patients in the Medicare Fee-for-Service Program,” avoidable hospital readmissions directly impact a patient’s health and are burdensome and costly.

In Michigan, hospitals and post-acute settings in metropolitan Detroit drive up the statewide readmission rate and suffer a disproportionate share of readmission penalties. Because they serve a diverse population, including a large proportion of low-income, minority, and vulnerable people, health care providers in southeast Michigan are trying to slow the pace of the readmission-related reimbursement cuts through a variety of quality improvement programs.

A key component in these efforts includes widespread collaboration and targeted strategies to improve care coordination for successful transitions of patients to the community.

Tri-County Collaborative

Beginning in 2015, five major health systems and more than 100 skilled nursing facilities (SNFs) in southeast Michigan recognized the opportunity to align goals and resources and formed the Tri-County SNF Collaborative with the intent to improve care transitions, enhance communication, provide higher quality of care, and reduce readmissions.

Participating SNFs shared a broad range of characteristics, with a variation in ownership types (for example, for-profit corporation, for-profit individual, for-profit partnership, government-county, nonprofit church related, nonprofit corporation, nonprofit other) and the Centers for Medicare & Medicaid Services (CMS) Five-Star Quality Ratings ranging from one to five.
This material was prepared by Lake Superior Quality Innovation Network, under contract with the Centers for Medicare & Medicaid Services (CMS), an agency of the U.S. Department of Health and Human Services. The materials do not necessarily reflect CMS policy.

MPRO, a quality improvement organization based in Farmington Hills, Mich., represents Michigan in Lake Superior Quality Innovation Network, serving Michigan, Minnesota, and Wisconsin under the CMS Quality Improvement Organization Program. MPRO leveraged established relationships throughout the Southeast Michigan community to engage, build, and organize the Tri-County SNF Collaborative with the aim of improving health outcomes for Medicare beneficiaries.

This unique position led to diverse leadership among the collaborative partners and helped create an environment that aligns resources to achieve low-cost, high-yield impact. All participants agreed that data would serve as a key driver of their efforts; process and outcome measures were tracked and reported by the SNFs to drive continuous quality improvement.

Aims of the Collaborative

Upon formation, the aims of the collaborative were to improve care transitions, enhance communication, provide higher quality of care, and reduce readmissions. Participating facilities committed to working with each other on aligned care transition measure reporting and shared best practices, with the united goal of improving care for all patients as a shared model.

SNFs reported data metrics on a quarterly basis; site-level feedback reports were shared with SNFs to drive quality improvement. Medicare Fee-for-Service Part A claims data were also used to calculate the percentage of 30-day readmissions for discharges from the health systems to the SNFs. Additionally, metrics related to acuity, transitions, and quality were evaluated to identify opportunities for improvement. Metrics included:
  • Acuity: Average daily census; facility average length of stay;
  • Transition: Time frame for first physician history and physical evaluation upon admission; seven-day follow-up primary care physician (PCP) visits for patients discharged from SNF to community; discharge summary provided to PCP within seven days;
  • Quality: Nosocomial urinary tract infection rate; nosocomial pneumonia rate; falls with major injury rate; short-stay new/worse pressure ulcer rate; 30-day all-cause readmissions with a SNF admission index diagnosis of sepsis; antipsychotic medication rate in long-stay patients; and
  • Readmissions: 30-day all-cause readmission rate.
Participants met quarterly to review progress and key outcome data findings. Best practices were identified and shared, and this resulted in grassroots education and the development of a multifaceted portfolio of market-driven strategies to address the needs of the high-risk populations served by the collaborative.

Some examples of the types of strategies that were used include:
  • Using data to drive decisions: A quality metric dashboard was created to address key focus areas for the collaborative.
Regular reviews of the metrics and data allowed for the discussion and agreement on which measures provided the most valuable information. System goals were intentionally made accessible to every member of the collaborative, which gave everyone the opportunity to offer insights. These insights were turned into action by the collaborative to improve the quality of care for patients.
  • Expansion of the collaborative: The collaborative realized its work impacted more than just hospitals and SNFs, and that it needed to include the entire care continuum if change were truly going to be made. The collaborative developed (and is currently piloting) a Home Health Agency (HHA) workgroup that is actively developing its own quality metrics dashboard, with the goal of collecting data that will offer insight into where the group can take action and make change for HHAs as well.
  • Early recognition and management of sepsis program and toolkit: The five regional health systems partnered to facilitate a sepsis program for SNFs in southeast Michigan. Participating SNFs attend five monthly meetings and two coaching calls where they receive education on the prevention of pneumonia, urinary tract infections, and central line-associated bloodstream infections. To date, 33 SNFs have successfully completed the session and another 60 SNFs planned to engage in the program in fall 2018.

Better Outcomes

The readmission rates are not adjusted or standardized in any way. Calculated rates include both scheduled and unscheduled readmissions due to difficulties in defining and removing “scheduled” readmissions. Data for July 2017 through August 2017 are not fully matured, and rates are subject to change.

Part A Medicare Fee-for-Service standard analytic claims data were used to calculate the 30-day SNF readmission rate. A readmission was defined as a discharge from a hospital to a nursing facility participating in the collaborative and readmitted to a hospital within 30 days of discharge.

A subsequent claim for admission to a SNF within two days of the inpatient discharge was used to determine the SNF to which the beneficiary was discharged. To help quantify results, comparison groups also were formed among the SNFs. SNFs that are participating in the collaborative were grouped into the collaborative SNF group, while all other SNFs in Michigan were grouped into the noncollaborative SNF group. The CMS Certification Number was used to identify the SNF in the dataset.

The SNF collaborative readmission results have fluctuated over time, with a period of downward trending during the first few months, followed by upward movement and proceeded by downward movement in the most recent months of the intervention. The SNF noncollaborative readmission rate showed an increasing trend over time. Figure 1 shows details of the SNF readmission measure.

Next Steps

With increasing acuity levels, medical complexities, and behavioral health issues within the SNF population, all the major health systems in the collaborative formed a regional network to strengthen their approach and quality improvement efforts.

Data outcomes have helped prioritize the need for a change in strategy and new approaches to care coordination. Subsequently, insights from this large and successful collaborative have led to measure modifications and targeted action plans, which are the result of root cause analysis and other quality improvement strategies.

Through re-evaluation of the collaborative needs, the health systems’ leadership decided to invest time, resources, and education in a more systematic and teachable approach for larger assemblies of providers. 
The health systems have proactively provided training programs led by subject matter experts for specific diagnoses like sepsis and heart failure. With this new collaborative approach, there has been a renewed surge of dedication and support for ongoing transformation and quality improvement.

Improved Quality of Care

Since the collaborative began, metrics related to acuity, transitions, quality, and readmissions have demonstrated that joint efforts can impact patient safety and clinical quality. The most recent months have also shown a reduction in SNF readmissions. In turn, the collaborative has fostered the development of sustainable, market-driven best practices to improve quality of care and care coordination. Further study deepened understanding of variation in quality across health care organizations and highlighted the value of the shared responsibility for transforming care delivery.
Elizabeth Waldman is director, Department of Applied Epidemiology and Evaluation, MPRO, part of Lake Superior Quality Innovation Network. She can be reached at 248-465-7390 or ewaldman@mpro.org.