Established in 1977 as an Area Agency on Aging serving Northeastern Massachusetts, Elder Services of Merrimack Valley supports the independence of older people and adults with disabilities living in the community by providing information and referrals to these residents and/or their caregivers. Elder Services of Merrimack Valley also oversees several programs to address behavioral health needs and promote alternatives to nursing home care.
The impetus for this program stemmed from Elder Services of Merrimack Valley’s participation in the Community-based Care Transition Program (CCTP), created through Section 3026 of the Affordable Care Act. This program supports promising service delivery models that reduce readmissions.
Supported by mobile technology, trained health coaches at Elder Services of Merrimack Valley (an Area Agency on Aging in Northeastern Massachusetts) visit recently discharged Medicare patients in their homes and monitor them via telephone to identify and address declines in health status that increase the risk of readmission. Administered in partnership with area hospitals, the 4-week program begins with an inhospital visit to determine the risk of readmission.
Patients at medium or high risk for readmission receive an in-home visit within 48 hours of discharge and a weekly phone call for each of the next 3 weeks. During each encounter, the coach uses a tablet-based application that provides suggested questions written in lay language based on the patient’s diagnoses, treatment, and overall risk profile. If the answers indicate a decline in health status, the system sends a real-time alert to a nurse care coordinator, who subsequently uses a different component of the software to help the patient and coach address the issue within 24 hours, including arranging for any needed services. The use of health coaches supported by the tablet-based software significantly reduced readmissions among at-risk Medicare patients, as compared with use of health coaches without the software. This reduction generated substantial cost savings for partner hospitals and the health care system as a whole.
Key program components are described below:
Inhospital screening and enrollment of at-risk patients: Health coaches meet with discharge planning staff at partner hospitals and/or skilled nursing facilities to identify Medicare fee-for-service (FFS) patients about to be discharged to their home. The health coach administers an evidence-based risk screening and stratification tool [developed by EQ-Health, the Medicare quality improvement organization (QIO) for Louisiana] to identify and enroll those at moderate or high risk of readmission. Factors used to identify those at risk include diagnoses, level of available in-home and community support, complexity of the medication regimen, risk of falling, and level of cognitive function.
Warm handoff to health coach: Before discharge, a health coach meets with the enrolled patient, family members or other caregiver (as appropriate), and discharge nurse to review post-discharge instructions and explain how the coach will support the patient over the next several weeks.
Technology-supported home visit and telephone monitoring over 4-week period: Aided by tablet-based software, the health coach conducts an in-home visit and checks in on a weekly basis by telephone for a period of 3 weeks to monitor the patient’s health status. During each encounter, the coach uses the software application to ask questions designed to detect a decline in health status and increased risk for readmission. Based on the initial screening and diagnosis and the patient’s living situation and answers to previous questions, the software suggests 10 to 15 questions written in lay language appropriate for the health coach and patient. If the patient has more than one diagnosis, the software prioritizes the questions to ask about the most recent diagnosis and/or those conditions most likely to lead to readmission. For example, the software will encourage the health coach to ask a patient with congestive heart failure about weight gain or shortness of breath, both significant risk factors for readmission. The system also lists important questions not associated with a particular condition, such as questions to gauge the patient’s ability to access appropriate food and medications.
Automatic alerts to nurse care coordinator for those at risk: After asking the questions, the health coach enters the patient’s responses into the application and submits them to the system via a secure connection. If the patient’s answers during any encounter indicate a decline in health status, the system automatically generates an alert that is immediately sent to a nurse care coordinator. The alerts are categorized (mild, moderate, and high) depending on the urgency and severity of the situation. During the first 6 months of the program, 22 percent of the 1,902 surveys administered to patients triggered an alert.
Nurse care coordinator intervention to address problems: After receiving the alert, the nurse care coordinator accesses the system with a computer or smart phone through a separate portal that uses higher level clinical language and features additional support related to patient follow-up and monitoring. The care coordinator reviews the situation and then, as appropriate, pushes a button on his or her smart phone to contact the health coach, patient, family member, and/or caregiver to discuss the problem and come up with a plan to resolve it within 24 hours. As appropriate, the care coordinator facilitates the provision of needed services, such as primary care, specialty care, a change to the medication regimen, a home visit from a nurse, and/or an emergency department visit. During the first 6 months of the program, the vast majority (98 percent) of alerts led to a telephone follow-up with the patient, health coach, or primary care physician. Less than 1 in 10 necessitated a home visit by a nurse, and 3 percent required an urgent, unscheduled visit with a primary care provider or specialist. Once the episode has been resolved, the care coordinator clicks “no further action” and the alert is archived.
System-supported documentation, patient prioritization, and staff evaluation/training: The software-based system documents all care coordination activities, including surveys administered, alerts triggered, and responses to the alerts. The system uses these data to identify patients in need of more frequent telephone check-ins or additional home visits. The system also monitors and reports on staff performance, such as whether health coaches visit patients within 48 hours of discharge or nurse care coordinators resolve alerts within 24 hours. Care coordinators use information from the system to provide ongoing feedback to the health coaches on various aspects of their performance, such as their skills related to asking patients questions and listening to their answers.
The use of health coaches supported by the tablet-based software significantly reduced readmission rates among at-risk Medicare patients, as compared with health coaches who did not have access to the software. This decline generated substantial cost savings for partner hospitals and the health care system as a whole.
- Significantly fewer readmissions: Based on hospital-generated admissions data from hospitals involved in the program, the addition of the software to the health coach protocol reduced 30-day readmissions by 39.6 percent among at-risk patients eligible for a health coach, as compared with the previous model in which health coaches did not have access to the software. This translates to a 5.9 percent reduction in all-cause readmissions among Medicare beneficiaries, including low-risk patients who did not get assigned a coach. In addition, an expected seasonal spike in readmissions during the winter (due to an increase in respiratory illnesses) did not occur.
- Significant savings: The net savings generated by the reduction in readmissions averaged $109 per patient per month, with gross savings of roughly $600,000 and net savings of approximately $370,000 during the 6-month trial period involving 561 patients.