The algorithm helped identify patients who would need end-of-life care communication from their oncologists.
Patient-provider communication about end-of-life wishes is important for making patients comfortable as they face serious illness, but knowing which patients could benefit from those conversations is often difficult for providers.
A group of researchers from Penn Medicine is looking to address that conundrum, announcing the effectiveness of a new predictive algorithm that can help providers identify patients for whom it would be appropriate to hold an end-of-life wishes conversation.
“On any given day, it’s actually pretty difficult to identify which patients in my clinic would benefit most from a proactive advanced care planning conversation,” said the study’s lead author Ravi Parikh, MD, an instructor of Medical Ethics and Health Policy at the University of Pennsylvania and a staff physician at the Corporal Michael J. Crescenz VA Medical Center.
Want to publish your own articles on DistilNFO Publications?
Send us an email, we will get in touch with you.
“Patients oftentimes don’t bring up their wishes and goals unless they are prompted, and doctors may not have the time to do so in a busy clinic. Having an algorithm like this may make doctors in clinic stop and think, ‘Is this the right time to talk about this patient’s preferences?’”
Separate research corroborates this. Data from a 2017 John A. Hartford survey revealed that 99 percent of providers agree advance care planning and end-of-life planning is important, and another 95 percent agreed with the 2017 move from CMS to reimburse providers for having these types of conversations.
But only 14 percent of providers actually billed for end-of-life planning, indicating that very few are actually holding these conversations with patients. And while hospital logistics and protocol can be cumbersome, most providers say they shy away from advance care planning because those conversations are uncomfortable.
Many providers don’t know which patients would appreciate these conversations, and many providers saying they are concerned their patients will believe their doctors have given up on them.
This newest innovation from Penn Medicine aims to turn those concerns on their head. The predictive algorithm looks at a number of patient data points, including their personal and healthcare demographics, to determine whether their providers should begin discussing end-of-life care.
The model correctly predicted that 51 percent of high-risk patients — or patients who would benefit from advance care planning — died within six months of the evaluation. Sixty-five percent died within 18 months of evaluation.
This comes compared to only four percent of low-risk patients — or patients who would not benefit from or appreciate advance care planning — who died within six months of evaluation.
These results indicate that the model can accurately predict which patients would benefit from having an end-of-life care conversation with their providers and family members.
This newest algorithm was adapted from one used to refer patients to palliative care, called PalliativeConnect. That platform was able to increase palliative care consultations by 74 percent, leading more patients to receive comforting and cost-effective care toward the end of their lives.
This latest algorithm is novel in that it can influence patient-provider communication between the patient and the oncologist specifically.
“We’re excited about the scalability of this decision support method for providers, and not just in oncology,” Parikh said. “Our process of using machine learning to flag high-risk patients in real time is broadly applicable, and our approach risk-stratifies patients in a usable way that just hasn’t been available to us before.”
Next steps for Parikh and the research team include a second pilot at a medical center that is separate from the original test site to determine the true scalability of the system.
Other researchers have likewise suggested at the importance of advance care planning and end-of-life conversations. A 2019 report from Dana Farber Cancer Institute researchers outlined a training system that can help doctors and nurses feel more comfortable engaging patients in end-of-life care conversations.
The guide specifically outlines how to assess patient understanding of prognosis, patient values and personal goals, fears and sources of strength, capabilities that are essential to good quality of life, and the types of treatments patients are willing to endure for more time.
Following introduction to the guide, providers began engaging patients in advance care planning an average of 2.4 months earlier than they typically would. Ninety percent of patients ended up discussing their personal care goals and values for the end of life, which in turn improved quality of life and reduced anxiety.
“We know how beneficial patient-centered conversations are, and our goal is to ensure that they happen for all patients, earlier in the course of illness and focus on what matters most to the patients,” said study author Rachelle Bernacki, MD, MS, associate director of the Serious Illness Care program and a palliative care doctor at Dana-Farber Cancer Institute.
“These results are exciting because they show us that more, better and earlier conversations are possible and they can reduce symptoms of depression and anxiety in our most vulnerable patients.”
Source: Patient Engagement HIT