Implementing Systematic Screening and Structured Care for Distressed Callers Using Cancer Council’s Telephone Services: Protocol for a Randomized Stepped-Wedge Trial

Background Structured distress management, comprised a 2-stage screening and referral model, can direct supportive care resources toward individuals who are most likely to benefit. This structured approach has yet to be trialed in Australian community-based services such as Cancer Council New South Wales (NSW) and Victoria Cancer Information and Support (CIS) 13 11 20 lines who care for a large community of cancer patients and caregivers. Objective The aim of this study was to evaluate the effectiveness of structured screening and referral in (1) increasing the proportion of distressed CIS callers who accept supportive care referrals and (2) reducing distress levels at 6-month follow-up. Methods In this stepped-wedge trial, Cancer Council NSW and Victoria CIS consultants are randomized to deliver structured care during inbound 13 11 20 calls in accordance with 3 intervention periods. Eligible callers are patients or caregivers who score 4 or more on the Distress Thermometer; NSW or Victorian residents; aged 18 years or older; and English proficient. Study data are collected via computer-assisted telephone interviews (CATIs) at 3- and 6-month follow-up and CIS record audit. CATIs include demographic and service use items and the General Health Questionnaire (GHQ-28) to assess distress. An economic analysis of the structured care model will be completed. Results The structured care model was developed by guideline review and identification of service characteristics to guide mapping decisions; place-card methodology; and clinical vignettes with think-aloud methodology to confirm referral appropriateness. The model includes an additional screening tool (Patient Health Questionnaire-4) and a referral model with 16-20 CIS services. Descriptive statistics will be used to assess referral uptake rates. Differences between GHQ-28 scores for structured and usual care callers will be tested using a generalized linear mixed model with fixed effects for intervention and each time period. The trial will recruit 1512 callers. The sample size will provide the study with approximately 80% power to detect a difference of 0.3 SD in the mean score of the GHQ-28 at an alpha level of .05 and assuming an intra-cluster correlation of .04. A random sample of recorded calls will be reviewed to assess intervention fidelity and contamination. To date, 1835 distressed callers have been invited to participate with 60.71% (1114/1835) enrolled in the study. A total of 692 participants have completed 6-month CATIs. Recruitment is anticipated to end in late 2019. Conclusions This trial is among the first to rigorously test the outcomes of a community-based structured approach to distress management. The model is evidence-informed, practice-ready, and trialed in a real-world setting. The study outcomes will advance the understanding of distress management internationally for both patients and caregivers. Trial Registration Australian New Zealand Clinical Trial Registry ACTRN12617000352303; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=372105&isReview=true (Archived by WebCite on http://www.webcitation.org/78AW0Ba09) International Registered Report Identifier (IRRID) DERR1-10.2196/12473


Scientific Quality (Refer to the information available in the Research
well written and well structured proposal The need for economic assessment ewas mentioned/implied but was not further discussed or developed This proposal addresses a defined need in cancer populations (i.e., psychological distress). There are two hypotheses (1. predicting lower distress in the Intervention group and 2. predicting that costs of the intervention will be appropriate in terms of improved client outcomes). Hypothesis 1 predicts a very small group difference, raising the question of the value of such an intervention. Clinically it is questionable whether a difference of 0.2 of a Standard Deviation would be meaningful. Hypothesis 2 lacks clarity as the terms "appropriate" and "improved client outcomes" are not well defined. Another concern is the lack of control of the content of Usual Care. It does not appear that the specific content of this condition will be monitored in the study (although the length of the phone consultations will be timed). As Usual Care is "unstructured" there is a possibility that one of the consultants could provide recommendations not dissimilar to that of the Intervention operators -there does not seem to be any mechanism outlined that would capture this situation. It is also possible that the longer time spent by consultants with Intervention participants may lead to differences in outcomes. Since sessions are being recorded it should be possible to consider time spent as a possible covariate. It is noted that the consultants will be trained and monitored to ensure protocol adherence, but it is not clear how can this be achieved with the unstructured usual care group. It is estimated that 1400 individuals will consent to be in the study, which is appropriate given the small effect size hypothesised, and appears feasible given that more than 15,000 individuals contact the Cancer Council's CIS services each year. However, it is not clear what percentage of these callers are likely to have a distress level of 4 or greater (a requirement for study entry). Some greater detail on these numbers would improve the case for this study being both feasible and representative. Since there is a move towards mandatory distress screening for callers to CIS services, this study has the potential to provide timely data regarding the most effective means by which to provide support and management of individuals experiencing heightened distress. This research will help inform best practice for community-based cancer support services, and may lead to international adoption due to links with CI Jacobsen (USA). One aspect that has been underplayed in this protocol is an argument why harmonisation of services across the nation is of value. This is alluded to but a stronger argument could have been put forward. Given the hypothesised small effect size, I don't find that the case for the significance of this project is compelling. This research proposal has been put forward by a multidisciplinary group of researchers from Australia, with one highly experienced CI from the USA (Jacobsen). The team is comprised of individuals with extensive experience in conducting randomised controlled trials, the assessment of support needs amongst cancer populations, and the provision of psychosocial support for cancer patients. Adding strength to this team is CI O'Brien who has extensive experience working at the Cancer Council NSW at the operational level, and CI Boltong, who is based at Cancer Council Victoria. Statistical and research design advice will be provided by the very experienced CI McElduff. I applaud the researcher's inclusion of an ECR researcher (CI Boyes) who has demonstrated in a short space of time an excellent track record of publication and grant funding. Overall, this team is highly experienced and diverse, addressing the necessary skills to undertake the proposed research. In addition, several of the team members have previously collaborated, resulting in several well received publications in related areas.

Budget Comments
Generally, the budget is appropriate for the scope of the proposed project. It is not particularly clear why the postdoc is required for 40% time only, nor how many telephone interviewers will be employed for the stated 50%, 100% and 50% requested budget. The requested budget for the database amendments is appropriate for this scope of work. Some more detail on the need for the visits to and from the Cancer Council offices would be helpful (assume that some of this is for the team meetings?).

Overall Comments
Overall this is an interesting proposed project that has the capacity to improve support services for individuals with cancer, particularly for the thousands of cancer survivors. However, some limitations of the study design outlined previously detract from the overall proposal. This application has the potential to address an important issue in cancer care. However the researchers provide no pilot data to indicate the proposed methodology is acceptable to cancer council clients and feasible in the context of cancer council's models of care. Particular issues that may need to be addressed is the potential for contamination between consultants within the usual care and intervention groups given they sit in the same workspace. The number of screening questions may also be problematic both in terms of time to complete when a non-consented and distressed participant is asking for support with the potential for consultants to deviate from protocol and/or impact on the quality of service delivery/uptake of services. Although I acknowledge that these factors will be assessed in the RCT preliminary data guiding the study design has not been provided and the methodology differs sufficiently from that used in QLD that there may be issues with acceptability. With respect to the primary outcome measure GHQ-28, it is not clear how baseline levels of distress will be controlled for given multiple distress measures are being used at b/l and 3/6months. Reliance on GHQ-28 at 3 months to control for individual differences at 6 months limits the ability to determine whether improvements were directly attributable to the structured intervention. The impact of the research could be strengthened by the inclusion of a more robust cost benefit analysis rather than subjective assessment using focus groups. The team has the experience to deliver the proposed intervention. It is a good balance between experienced methodologists, leaders in psychosocial distress research, implementation and service delivery. the team would benefit from health economic expertise to better define how to achieve aim 2 of the research.Inclusion of a number of ECRs is appropriate and builds research capacity in psycho-oncology.

Budget Comments
The budget is appropriate for the proposed work. Level of PSP appointments proposed are also appropriate for skill level required

Assessor 3
Overall Comments This application addresses and important area of cancer care but the application may be premature given the lack of pilot data to support such a large RCT as proposed.