Effect of an Oncology Nurse-Led Primary Palliative Care Intervention on Patients with Advanced Cancer: The CONNECT Cluster Randomized Clinical Trial

Yael Schenker, Andrew D. Althouse, Margaret Rosenzweig, Douglas B. White, Edward Chu, Kenneth J. Smith, Judith M. Resick, Shane Belin, Seo Young Park, Thomas J. Smith, Marie A. Bakitas, Robert M. Arnold

Research output: Contribution to journalReview articlepeer-review

31 Scopus citations


Importance: Guidelines recommend early specialty palliative care for all patients with advanced cancer, but most patients lack access to such services. Objective: To assess the effect of CONNECT (Care Management by Oncology Nurses to Address Supportive Care Needs), a primary palliative care intervention delivered by oncology nurses, on patient outcomes. Design, Setting, and Participants: This cluster randomized clinical trial of the CONNECT intervention vs standard care was conducted from July 25, 2016, to October 6, 2020. Participants were adult patients with metastatic solid tumors who were undergoing oncological care and for whom an oncologist would agree with the statement "would not be surprised if the patient died in the next year." The trial was conducted at 17 community oncology practices in western Pennsylvania. Data analyses adhered to the intention-to-treat principle. Interventions: The CONNECT intervention included 3 monthly visits with an existing infusion room nurse who was trained to address symptoms, provide emotional support, engage in advance care planning, and coordinate care. Main Outcomes and Measures: The primary outcome was quality of life. At baseline and 3 months, participants completed assessments of quality of life (Functional Assessment of Chronic Illness Therapy-Palliative care: score range, 0-184, with higher scores indicating better quality of life), symptom burden (Edmonton Symptom Assessment Scale: score range, 0-90, with higher scores indicating greater symptom burden), and mood symptoms (Hospital Anxiety and Depression Scale [HADS]: score range, 0-21, with higher scores indicating substantial anxiety and depression). Linear mixed-effects models were used to estimate adjusted mean differences in 3-month outcomes. Preplanned, intensity-adjusted analyses were conducted. Results: A total of 672 patients were enrolled (mean [SD] age, 69.3 [10.2] years; 360 women [53.6%]). The mean (SD) number of CONNECT visits completed was 2.2 (1.0). At 3 months, no difference in mean (SD) quality-of-life score was found between the CONNECT and standard care groups (130.7 [28.2] vs 134.1 [28.1]; adjusted mean difference, 1.20; 95% CI, -2.75 to 5.15; P =.55). Similarly, there was no difference between groups in 3-month mean (SD) symptom burden (23.2 [16.6] vs 24.0 [16.1]; adjusted mean difference, -2.64; 95% CI, -5.85 to 0.58; P =.11) or mood symptoms (HADS depression subscale score: 5.1 [3.4] vs 4.8 [3.7], adjusted mean difference, -0.08 [95% CI, -0.71 to 0.57], P =.82; HADS anxiety subscale score: 5.7 [3.9] vs 5.4 [4.2], adjusted mean difference, -0.31 [95% CI, -0.96 to 0.33], P =.34). Intensity-adjusted analyses revealed a larger estimated treatment effect for patients who received a full dose (3 visits) of the CONNECT intervention. Conclusions and Relevance: This cluster randomized clinical trial found that a primary palliative care intervention that was delivered by oncology nurses did not improve patient-reported outcomes at 3 months. Primary palliative care interventions with a higher dose intensity may be beneficial for most patients with advanced cancer who lack access to palliative care specialists. Trial Registration: ClinicalTrials.gov Identifier: NCT02712229.

Original languageEnglish (US)
Pages (from-to)1451-1460
Number of pages10
JournalJAMA Internal Medicine
Issue number11
StatePublished - Nov 2021

ASJC Scopus subject areas

  • Internal Medicine


Dive into the research topics of 'Effect of an Oncology Nurse-Led Primary Palliative Care Intervention on Patients with Advanced Cancer: The CONNECT Cluster Randomized Clinical Trial'. Together they form a unique fingerprint.

Cite this