Pragmatic Implementation Trial of a CF Primary Palliative Care Intervention

Overview

Five Cystic Fibrosis (CF) centers, key stakeholders, and a palliative care institute have collaborated to create a novel primary palliative care intervention for patients with CF, "Improving Life with CF: A Primary Palliative Care Partnership," and established the infrastructure and support necessary for a follow-on implementation study. This intervention provides a framework for a nationally generalizable model to improve best practices in generalist-level palliative care in CF. Objectives: Aim 1: Implement a primary palliative care intervention comprising screening-and-triage workflows, best practice treatment guides for high frequency problems, patient/family and provider education, and a quality improvement (QI) toolkit. Aim 2: Evaluate feasibility, uptake, and preliminary outcomes during a multisite pragmatic, implementation trial of the intervention at 5 diverse Cystic Fibrosis (CF) Centers. Subaim 2.1: Evaluate feasibility and uptake as measured by rates of screening and treatment delivery. Hypothesis 1: Related to feasibility and uptake of the intervention: 1. \> 80% of individuals with CF of all ages will receive an annual palliative care screening. 2. \> 25% of individuals with CF will receive a palliative care screening prompted by hospitalization, new diagnosis of CF-Related Diabetes, need for transplantation, or another disease- or treatment-specific trigger. Hypothesis 2: Related to provider education: a) \> 80% will access \>1 training(s) (on-demand webinars or in-service by trained site educators). Subaim 2.2: Evaluate data on preliminary outcomes for individuals with CF by comparing ratings on patient- and caregiver-reported outcome measures (e.g., Integrated Palliative Care Outcome Scale, Cystic Fibrosis Questionnaire-Revised, Memorial Symptom Assessment Scale-Cystic Fibrosis, and Brief Assessment Scale for Caregivers) during the trial to baseline ratings and exploring covariates of change (age, race/ethnicity, gender, disease severity, CFTR modifier treatment, psychological distress, and varied indicators reflecting intervention implementation).

Introduction: Palliative care is an interdisciplinary model of care intended to mitigate the suffering and illness burden experienced by patients with serious illnesses, and their families, from the time of diagnosis forward. The illness burden experienced by patients with CF and their caregivers is characterized by problems in many domains (e.g., physical symptoms, emotional distress, caregiver burden, and need for goal-setting). Interventions to address the goals of palliative care must be undertaken throughout the course of the illness by CF professionals-an approach known as generalist-level or primary palliative care. At present, there are no standardized, generalizable models of primary palliative care for CF. New CF Foundation guidelines recommend that CF Centers screen all adults and children from the time of diagnosis, along with caregivers of individuals with CF of all ages, for unmet palliative care needs annually and after "triggers" such as hospitalization or other major events (Kavalieratos D et al. J Palliat Med. 2020 Sep 16. Epub ahead of print). The goal is to identify high frequency problems, such as symptom distress and the need for advance care planning (ACP). The proposed intervention will involve a multi-site implementation trial of a new care model that applies the CF Foundation Palliative Care Guidelines (Kavalieratos D et al., 2020) at five CF Centers across the country. We will collect longitudinal data from patients and family caregivers on palliative care outcomes and also evaluate uptake of the care model. Research conduct: At baseline, each of the 5 participating CF Centers will recruit all eligible adolescent and adult CF patients plus one identified caregiver per patient of any age. Participants will have the opportunity to have their questions answered by a member of the site's study staff and will have at least 24 hours to decide whether they wish to participate. Those who agree to participate will be assigned a study ID number and their status tracked in a secure enrollment log. Outcome Assessment: Prior to the start of the system-level palliative care intervention involving individuals with CF and their primary caregivers at each site, patients and caregivers consenting to participate in the outcome assessment will complete a baseline questionnaire packet. This packet will include validated questionnaires that will be repeated during the outcome assessment. Participants will be reimbursed $20 for the baseline assessment, which is expected to take \< 30 minutes to complete. The baseline assessment will include sociodemographic items that assess age, gender, racial/ethnic status, primary language, marital status, educational level, employment status, annual household income, and insurance status, and will be completed by patients and caregivers to profile the cohort. During the last 3 months of trial implementation, patients and caregivers who consented to participate in the baseline assessment will be asked to complete a second questionnaire packet (identical to the first). This will occur approximately 2 years following baseline, in study months 25-27. Participants will be reimbursed $20 for the outcome assessment, which is expected to take \< 30 minutes to complete. After the baseline research assessment, all study sites will seek to implement new CF Foundation guidelines recommending annual and triggered palliative care needs assessments for individuals with CF and their primary caregivers (Kavalieratos D et al., 2020). As each of the CF Centers in the study adopt this new system-wide care model, the proposed study will conduct an intervention trial to support practice change, involving: 1) practice-level screening-and-triage workflows, 2) best practice treatment guides for management of high frequency problems; 3) provider education in generalist-level palliative care, supported by on-demand webinars and 4) train-the-trainer approach (education and manuals); and 5) a QI toolkit including resources for patient and family education, the monitoring of quality indicators for successful implementation of routine screening, and focused studies related to specific objectives of palliative care. Statistical Plan: Process measures collected in the trial will include data on uptake of the intervention, including rates of screening, triage, use of best practice guides, and provider use of education (on-demand webinars and on-site in-services and training) at each site. For process measures, the electronic medical record or other tracking systems at each site will be queried quarterly under a waiver of written authorization for medical record abstraction. This de-identified, aggregate information, such as percentage of center patients who receive an annual or clinically triggered palliative care screening (e.g., Subaim 2.1, hypotheses 1a and 1b), will be entered by the Research Coordinator into a REDCap® database for the project. All data from the participating sites (process data on intervention uptake from the electronic medical record, and quantitative data from patients and caregivers from REDCap®) will be merged into a SPSS database by the Research Manager at the Data Coordinating Center, MJHS Institute. The process measures will provide information related to the feasibility and uptake of the proposed intervention. The analysis will evaluate a) the proportion of patients and caregivers who completed annual screenings vs. patients and caregivers who are eligible; b) the proportion of screenings completed by patients and caregivers following specific triggers: hospitalization, new diagnosis of CF-Related Diabetes, and referral/declined referral for transplant evaluation; c) the proportion of patients and caregivers whose screening identified a need for further assessment for various symptoms or needs; and d) the proportion of patients who underwent a change in care plan based on the triage approach for different symptoms or needs. The analysis will evaluate each of these outcomes by patient age group (adults, adolescents, older children, and young children). Uptake also will be assessed by the number of Care Team members who access on-demand webinars or on-site in-service training. At each site, the new model will be judged to be feasible or not, based on whether the Center achieves all of the following benchmarks: 1) the structural components can be implemented and sustained over time and 2) by the outcome assessment, \>80% of patients due for an annual screening are screened, and \>25% of patients who need a 'triggered' screening are screened. The overall feasibility of the model relative to national distribution will be affirmed if at least four of the practices demonstrate that the model is feasible at the site. This evaluation of feasibility will be complemented by exploratory quantitative analyses. In the quantitative analyses, all study variables will be described, and distributions evaluated. Generalized logistic linear mixed effects models will evaluate changes in all uptake variables. Uptake variables (1=screened, 0=not screened) will serve as the outcome in separate models. All study variables will be described and distributions evaluated. Linear mixed effects models will evaluate changes in the typical primary outcomes (CFQ-R physical and emotional functioning subscale scores) and secondary outcomes (IPOS total score, MSAS-CF total score, total scores for the BASC, PHQ, GAD, PSS, and other CFQ-R subscales (e.g., vitality, digestive, respiratory symptoms, and role subscale scores)) as preliminary data for a fully-scaled randomized clinical trial. Time will serve as the fixed effect. Individual intercepts, site, and time will serve as random effects. The best fitting covariance structure will be selected using Akaike's Information Criterion. Additional separate models will evaluate patient- (sociodemographics, disease severity-, and psychological distress-related variables), provider- (use of education sessions), and site-level predictors of uptake (size of practice). SPSS version 25 will be used for all analyses. All analyses will be two-tailed with alpha set at .05. With baseline data prior to the intervention, longitudinal data from the outcome assessment, and process data about uptake of the intervention by each site, we also will be able to explore relationships between the intervention "dose" trends (e.g., the frequency of screening and triage across patients and the type of clinical interventions recommended and utilized) and effects on patient symptoms, function, or HRQL. Monitoring and Quality Assurance: Monitoring of source data will occur at each study site and by the Data Coordinating Center (the MJHS Institute). Process data on uptake of the intervention at the CF Centers and individual self-report or parent proxy report questionnaires, will be checked by research staff using a Standard Operating Procedure (SOP) Manual specifying the standard practices, methods, and protocol. A member of the study staff will cross-check all data and review all self-report measures to ensure their completeness. Any errors in completion will be reviewed to determine if directions or procedures for the assessments need to be altered. In this case, permission from the IRB will be requested to change any procedures. In addition, study personnel will use IRB-approved eligibility checklists to ensure that all patients and caregivers who are approached for participation in study assessments are eligible. Research Coordinators at each site will be in weekly contact with the Site PIs and Research Manager at the Data Coordinating Center (the MJHS Institute) to review any issues with the data (i.e., missing data) and ensure data integrity. Given the multi-center nature of the study, the study PIs (Anna Georgiopoulos, MD, and Lara Dhingra, PhD) will also meet regularly by telephone and videoconference with the Site PIs to review study conduct at all sites. Sites will be required to review their progress regularly and report unanticipated events, protocol deviations, and protocol violations at least monthly to the study PIs. Dr. Lara Dhingra, Co-PI at the MJHS Institute, the Data Coordinating Center for this study, and her research team will oversee the conduct of monthly protocol compliance audits. These will focus on informed consent procedures, participant eligibility, baseline and outcome assessments, and quarterly review of process measures. Protocol compliance will consist of 100% verification of the protocol specific measures identified in the protocol being available and accurately noted for the participant being reviewed. All study questionnaires will be audited for completeness by Dr. Dhingra, who also will conduct quality checks of the database that will receive this information; these data verification audits will be conducted for 100% of cases. Safety Monitoring: The proposed study has been reviewed by the Cystic Fibrosis Foundation Data Safety Monitoring Board, with the determination that the program is minimal risk and that there is no need for a Data Monitoring Committee for this study. Study PIs will be responsible for safety monitoring. Site PIs or their designees will be available at all times to respond to any question of safety.