This clinical trial evaluates an educational program called Algorithm-Enabled Patients Activated in Cancer Care Through Teams (A-PACT) for reducing unplanned hospital visits and improving goals of care conversations with providers among patients with solid cancers. A-PACT is an educational program where lay health workers (educators) help patients talk with their health care team about issues that matter most to them (goals of care). During A-PACT sessions, patients receive assistance in formulating health care and end of life care preferences, assistance in completing advance directives, guidance on how to engage in these conversations with family members, friends, and clinical teams, and encouragement to discuss these topics with their clinical team. A-PACT may reduce unplanned hospital visits and improve goals of care communication with providers among patients with solid cancers.
Participants must have a diagnosis of a solid tumor malignancy of any stage
Participants must be identified as high risk, defined as having 6 month mortality estimate from machine learning (ML) algorithm >= 20%
Participants must not be receiving or have pre-existing plans to enter hospice care at the time of study registration
Participant must be actively receiving or planning to receive systemic anti-cancer therapy (defined as any oral, injection, or intravenous therapy against cancer) within 3 months after registration
None
Parallel
Best Practice
Other interventionReceive usual care
Educational Intervention
Other interventionReceive A-PACT education
Electronic Health Record Review
Other interventionAncillary studies
Interview
Other interventionAncillary studies
Questionnaire Administration
Other interventionAncillary studies
Telephone-Based Intervention
BehavioralReceive A-PACT phone calls
Proportion of patients with any hospitalization
The primary analysis will be conducted using multivariable logistic regression, adjusting for stratification factors as covariates.
Proportion of patients with 0 versus >= 1 of the combined endpoint of hospitalizations or emergency room (ER) visits
Logistic, linear, and negative binomial regression models will be used, as appropriate for the given endpoint. Each regression analysis will include covariate adjustment for the stratification variables.
Number of hospitalizations as a continuous outcome
Logistic, linear, and negative binomial regression models will be used, as appropriate for the given endpoint. Each regression analysis will include covariate adjustment for the stratification variables.
Number of hospitalizations or ER visits
Logistic, linear, and negative binomial regression models will be used, as appropriate for the given endpoint. Each regression analysis will include covariate adjustment for the stratification variables.
Intensive end-of-life care
Logistic, linear, and negative binomial regression models will be used, as appropriate for the given endpoint. Each regression analysis will include covariate adjustment for the stratification variables.
Goals of care communication
Abstracted from the electronic health record. Will assess differential impact by income and race and ethnicity.
Emergency room visits
Will assess differential impact by income and race and ethnicity.
Patient-reported anxiety
Measured using Patient Reported Outcomes Measurement Information System (PROMIS) version 1.0 Short Form Anxiety. Will assess differential impact by income and race and ethnicity.
Patient-reported depression
Measured using PROMIS version 1.0 Short Form Depression. Will assess differential impact by income and race and ethnicity.
Click on the lead sponsor to view all of their trials.
SWOG Cancer Research Network
Lead sponsor
National Cancer Institute (NCI)
Collaborator
