Comparing Three Ways to Help Cancer Survivors Plan for Follow-Up Care [Internet]

Excerpt

Background: In 2005, the Institute of Medicine (now the National Academy of Medicine) issued, “From Cancer Patient to Cancer Survivor: Lost in Transition,” documenting deficiencies in cancer care during the transition from treatment to long-term survivorship. The report recommended that patients receive a survivorship care plan (SCP) that summarizes a patient's diagnosis, the cancer treatments received, and the possible implications of treatment, and that also outlines future health care needs. In 2015, the American College of Surgeons' Commission on Cancer made provision of SCPs an accreditation requirement, though this was subsequently relaxed in 2019. SCPs have been widely endorsed, but uptake has been limited. Previous trials, primarily using patient-reported outcome (PRO) end points, have failed to demonstrate impact on outcomes such as functional status and well-being. This study's purpose was to identify the simplest, most effective, and feasible SCP process, for diverse settings and cancer types, to promote appropriate survivorship care. Because SCPs are intended to help survivors understand and receive needed follow-up care, the study's primary end point was adherence to SCP-recommended care.

Objectives: This research was organized around 3 aims:

1. Aim 1: Compare 3 models of SCP implementation for 3 cancer types in 2 health systems

2. Aim 2: Compare the effectiveness of the 3 models of SCP delivery

3. Aim 3: Assess stakeholders' perspectives on the value of survivorship care planning

Methods: A randomized controlled trial (RCT) comparing 3 models of SCP implementation of various intensities was conducted and complemented with exploratory focus groups and postintervention qualitative interviews. The RCT compared the feasibility, effectiveness, and patient centeredness of the following 3 models in 2 cancer centers: (1) an urban, academic tertiary medical center with multiple oncology clinics based on disease site and (2) a rural, regional community medical center with a single general oncology clinic. Thus, a total of 4 clinics participated across 2 cancer centers, with breast, prostate, and colorectal cancer clinics at the urban-academic site and a single general oncology clinic at the rural-community site. To inform the RCT, we conducted 6 focus groups with cancer survivors and caregivers. In the RCT, participants were randomized to 1 of 3 interventions:

1. Participant was sent a written SCP by mail (“mail” group) and the SCP was mailed to their primary care provider (PCP)

2. Participant received an SCP during a survivorship visit (“1-visit” group) and the SCP was mailed to their PCP

3. Participant received an SCP during a survivorship visit, plus a 6-month follow-up visit (“2-visit” group), and the SCP was mailed to their PCP

Trial participants were followed for 18 months. We collected health services use at 6, 12, and 18 months and the following PROs at baseline and at 6, 12, and 18 months: Preparing for Life as a (New) Survivor, Assessment of Survivor Concerns, and Follow-up Care Use among Survivors– Information Needs Module. The primary outcome was receipt of SCP-recommended care, assessed based on the patient-reported health service use and medical record abstraction. Eligibility criteria were: age 21 years or older; diagnosed with nonmetastatic (ie, stages I-III) breast, colorectal, or prostate cancer; completed acute treatment within 3 months (ongoing long-term maintenance therapy allowed); no evidence of disease; and a participating site was considered “responsible” for survivorship care. After 18 months of follow-up data collection, we conducted qualitative interviews with trial participants about perceived intervention value and implementation experience.

Results: We randomized 378 survivors of breast (n = 159 [42%]), prostate (n = 142 [38%]), and colorectal (n = 77 [20%]) cancer to 3 arms: 126 for mail, 125 for 1-visit, and 127 for 2-visit; 47 survivors and caregivers took part in preintervention focus groups, and 40 trial participants were interviewed after the trial's completion. The RCT and subsequent interviews identified considerable interest in survivorship care planning, but they also documented implementation challenges; overall, 76.8% of participants received the intervention per protocol. Feasibility challenges were more pronounced at the urban-academic site than at the rural-community site (61.6% vs 93.4% of participants receiving the intervention per protocol, respectively). We found no difference across study arms in the proportion of participants who received all plan-recommended care across the 18-month observation period. In the mail arm, 45.2% received SCP-recommended care, along with 50.5% in the 1-visit arm and 42.7% in the 2-visit arm (P = .60). The differences between arms did not vary by cancer type (P_interaction = .56) or recruitment site (P_interaction = .86). The proportions of participants who had the recommended number of clinic visits were 53.4% for mail, 54.8% for 1-visit, and 53.7% for 2-visit (P = .99). The proportions undergoing SCP-recommended tests/procedures over the 18-month observation period were 78.8% for mail, 77.2% for 1-visit, and 73.0% for 2-visit (P = .62). The proportion receiving recommended nonoral cancer medications was 66.7% for mail, 75.0% for 1-visit, and 73.7% for 2-visit (P = .87). Our exploratory analysis of possible testing overuse showed no significant difference by arm (P = .65). For individual patients, 28 (9%) across the 2 sites received at least 1 test consistent with overuse, most frequently in patients with breast cancer and least frequently in patients with prostate cancer. Across all arms and time points, PRO scores (a secondary outcome) were high and information needs low, indicating that regardless of study arm, study participants were generally knowledgeable, confident, and not worried about follow-up care. Our primary PRO analysis found no statistically significant differences between the study arms at 12 months. From the interviews, we learned that SCP documents and visits were not transformative for patients' experience of survivorship care in the 18 months post–acute treatment, but they had value when they supported care coordination, and they provided opportunities for continued support and space for questions and dialogue about ongoing needs.

Conclusions: Our study did not find differences in receipt of recommended survivorship care or PROs by SCP delivery approach. The 2 sites differed in delivering the SCPs as specified in the study protocol, with fewer participants receiving the SCP per protocol at the urban-academic site. Interview data were consistent with the Commission on Cancer's recent emphasis on SCPs as only 1 part of intentional survivorship care. For services to be patient centered, it is important to understand both health systems' needs to tailor SCP delivery to available resources, and patients' needs and preferences once the acute phase of treatment is ending. Patients need accurate and accessible information, and health systems need survivorship programs that are workable with other clinic demands. As we found no clear indication that 1 mechanism of delivering SCPs was superior to others, survivorship programs can prioritize feasibility and acceptability when designing SCP approaches.

Limitations: Limitations include being conducted in only 2 health systems and requiring that participants have health insurance and speak English. Neither study participants nor research or clinic staff were blinded to study arm assignment. We did not include a no-SCP arm and cannot comment on how any of the SCP delivery approaches compare with no SCP at all. Further, because this was not an equivalence trial, we cannot conclude that the 3 approaches are equally effective.