How Coastal Hospice Improved Clinical Outcomes with AI

In case you missed it, Hospice News has launched a new specialty publication for palliative care professionals. You can subscribe to Palliative Care News here: Subscribe today!

Coastal Hospice & Palliative Care recently began exploring ways to improve clinical outcomes during the last days of life, an important quality measure in end-of-life care.

Established in 1980, Maryland-based Coastal Hospice serves four counties in its home state. In addition to hospice and palliative care, the nonprofit provider offers grief support, dementia patient and caregiver services and an advanced lung care program.


The organization has found success utilizing an AI program that analyzes patient data using a palliative performance scale (PPS) indicator, according to Monica Escalante, president and CEO at Coastal Hospice. This window into a patient’s conditions has helped hospice teams better assess approaches to care delivery, Escalante said.

“In the AI reports, we are looking for trends that help us staff better around patient needs, and they help us provide processes that will lead to better outcomes,” Escalante told Hospice News. “That data is incredibly important; it can never replace clinical judgment. It just gives [clinicians] a tool at their fingertips to assess certain situations. We can validate the data in their electronic medical records and the PPS is also based on the clinician making an assessment every time they see the patient.”

The palliative performance scale measures declines in a patient’s health status, including data around assistance with activities of daily living (ADLs), ambulation and mobility, changes in appetite, sleep and incontinence, among others.


The PPS scale is utilized by all interdisciplinary staff with an aim to increase hospice visit frequencies and predict additional family needs as a patient declines, according to Coastal Hospice Chief Clinical Officer Ann Lovely. Through the AI technology, patients with less than a 20% PPS score were earmarked for additional daily visits at the end of life.

The palliative scale aided in adjusting after-hours triage staff scheduling and optimizing operational efficiencies, Lovely stated.

“Understanding this scale is important as hospice staff can be proactive in their care,” Lovely said. “For example, a patient at a 20% [PPS] would be transitioning to the actively dying phase and may require additional phone check-ins that would guide clinical staff in knowing if a visit is warranted or a change in plan of care is needed.”

Since implementing the AI, Coastal Hospice has seen quality scores in the last days of life improve, according to Escalante. Roughly 71% of patients received three visits or more during the last days of life as a result of utilizing the PPS scale, she stated.

The hospice provider’s quality ratings have reached 23% above the national averages for hospice visits in the last days of life (HVLDL), according to Coastal Hospice. The HVLDL is part of the U.S. Centers for Medicare & Medicaid Services’ (CMS) quality measures in the Hospice Quality Reporting Program (HQRP) and the Hospice Item Set (HIS).

Much remains to be accomplished when it comes to improving hospice visits during the final stages of life, Escalante indicated.

“The support of using artificial intelligence guides the clinicians to help them see the whole disease progression and the decline,” Escalante said. “All those cues help in developing a trusting relationship with the caregivers to also voice what they observe. What I am most concerned about isn’t that 71% receiving more visits, it’s the 29% who aren’t. That means caregivers aren’t getting that support, and our goal is to make it 100% of patients receiving several visits.”

The palliative data tool has also measured clinical capacity needs, as well as improvements in patient and family satisfaction, according to Holly Murray, director of home hospice at Coastal Hospice.

“It has improved weekend staffing and scheduling because we are doing our best at anticipating our patient’s needs,” Murray said in an email. “Also, overall patient satisfaction has improved because of the better outcomes.”

The AI also has helped curb high-acuity utilization, such as preventing rehospitalizations, among others, according to Escalante. The technology has helped flag when patients have after-hours triage needs such as assistance with medication management and obtaining durable medical equipment and costly infusion therapies, she stated.

Financial impacts have come from the predictive analytics tool as well, Escalante stated.

“We’re able to bill for these visits and be more supportive as patients decline,” Escalante said. “It’s anticipating crises and supporting caregivers. It also affects our bottom line, because we can bill for these visits at a little over $20,000 a month in revenue.”

Companies featured in this article: