Loading ...

TOPLINE:

A default palliative care referral intervention increased palliative care consultations by more than fivefold and decreased end-of-life systemic therapy by more than half among patients with advanced cancer being treated in the community oncology setting.

METHODOLOGY:

• Early palliative care among patients with cancer can enhance quality of life and reduce intensive end-of-life care and is recommended by national guidelines. However, there are barriers to making these referrals and limited strategies to increase access in the community oncology setting.
• Researchers developed an intervention — an algorithm-based default palliative care referral, embedded in an electronic health record (EHR) — to help overcome barriers to palliative care referrals. The EHR algorithm assigned scores to each patient based on their palliative care risk factors to identify those eligible for palliative care.
• The researcher conducted a trial across 15 community oncology clinics and evaluated 562 patients with advanced lung cancer or noncolorectal gastrointestinal malignant tumors who were randomly assigned to either the control group (n = 266; eight clinics) or the EHR algorithm-based default palliative care group (n = 296; seven clinics).
• Clinicians at the control sites received weekly reports on palliative care referral rates compared with their peers and referred patients at their own discretion. Those at intervention sites also received default palliative care orders using the EHR.
• The primary outcome was completed palliative care consultation within 12 weeks of enrollment; exploratory outcomes included quality of life, feeling heard and understood, and markers of intensive end-of-life care.

TAKEAWAY:

• Overall, 43.9% of patients in the intervention group completed palliative care consultations compared with 8.3% of those in the control group, highlighting a significant increase in completed palliative care visits using the algorithm-based default referral (adjusted odds ratio [AOR], 8.9; P < .001).
• Among patients completing palliative care visits, both groups showed similar engagement in terms of the median number of visits (2.9 in the intervention group vs 3.4 in the control group) and the percentage of individuals with more than one visit (84.2% vs 89.3%).
• Among patients who died at the 24-week follow-up, those in the intervention group had lower rates of systemic therapy within 14 days of death than those in the control group (6.5% vs 16.1%; AOR, 0.3; P = .05). No significant differences in overall survival or late hospice referrals were observed between the two groups.
• At 24 weeks, no significant differences were observed between the groups in terms of health-related quality of life (mean difference, −0.5; P = .68) and feeling heard and understood (mean difference, −0.1; P = .51).

IN PRACTICE:

“In this randomized clinical trial, an intervention combining algorithm-based automated identification of patients eligible for PC [palliative care] with default PC referral led to an increase in PC visits and a decrease in end-of-life systemic therapy among patients with cancer,” the authors wrote. “This study provides guidance for scalable, algorithm-driven PC implementation across community oncology settings,” they added.

SOURCE:

This study, led by Ravi B. Parikh, MD, MPP, Emory University School of Medicine, Atlanta, was published online in JAMA Network Open.

LIMITATIONS:

The findings may not be generalizable to other community-based systems without embedded palliative care. The palliative care algorithms were limited to depression and distress screening surveys, lacking external validation despite being based on national guidelines. Additionally, patient-reported outcomes may have been underpowered because of low completion of assessments.

DISCLOSURES:

This study was supported by Emerson Collective. Several authors reported receiving grants or personal fees and having other ties with various sources. Additional disclosures are noted in the original article.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.