International Study to Evaluate Two Programs of Support for Patients With Advanced Cancer and Their Families

Summary

Participation Requirements

Description

STUDY SETTING Both interventions (FOCUS+ and iFOCUS) will be administered in the homes of the patient-caregiver dyads (or in the location of the dyad's preference). The intervention will be conducted in six countries (Belgium, Denmark, Ireland, Italy, the Netherlands and the United Kingdom). In each...

STUDY SETTING Both interventions (FOCUS+ and iFOCUS) will be administered in the homes of the patient-caregiver dyads (or in the location of the dyad's preference). The intervention will be conducted in six countries (Belgium, Denmark, Ireland, Italy, the Netherlands and the United Kingdom). In each country, patients with advanced cancer and their primary family caregiver will be recruited and enrolled via participating hospitals. Inclusion criteria for hospitals participating in this study are having 1) to treat patients with advanced cancer and 2) an oncology care regimen. SAMPLE SIZE For the emotional functioning the EF10 subscale from the EORTC is used. For self-efficacy the CASE instrument is used. The investigators consider demonstration of an intervention effect (for each of both interventions) on at least one of these primary outcomes for either the patient or the caregiver at t1 as a success. A pre-determined strict fixed sequence (FS) procedure defines prospectively hierarchical ordering of the endpoints, for this study the hierarchical order is emotional functioning (1) and self-efficacy (2). Testing of null hypotheses proceeds according to their hierarchical order, that is, H(1)0 is tested first at a significance level of 5%, and if H(1)0 is rejected then H(2)0 is tested at the same significance level, otherwise H(2)0 is not tested at all. The strict FS approach has the highest power for testing the first hypothesis (outcome: emotional function) compared to the other methods, as it does not save any portion of alpha for testing later hypothesis. The reference mean value from EORTC for all cancer patients, stage III-IV is 71.5 (SD: 23.8). Alpha is set at 0.0125 instead of 0.05 to account for multiplicity (2 comparisons with control group * 2 participant groups [patients and caregivers]). 1-beta (i.e. statistical power) is set at 0.9. The expected difference between the control group and the intervention arms in the primary outcomes is 0.375 SD at t1 (12 weeks). With these parameters n= 203 is needed in each arm across all countries (i.e. 609 in total). Anticipating a 65% retention rate at t1, which is more conservative than found in previous studies in the USA on the FOCUS interventions due to the advanced cancer population included in this study, 938 dyads must be enrolled across the 6 countries (313 per group). This means n= 156 need to be enrolled in total in each country (n=52 in each of the 3 arms per country). Based on previous studies in the USA an enrolment rate of 55% is expected of those dyads referred to the study, meaning that about 282 dyads will need to be screened and identified in each country. The feasibility of recruitment has been evaluated based on previous research and discussions with clinicians in eligible hospitals. DATA ANALYSIS Four main quantitative analyses will take place: Primary hypotheses testing: Testing the null hypothesis of the first primary endpoint: emotional functioning. The effectiveness of the FOCUS+ face-to-face intervention and the iFOCUS web-based intervention will be compared with the standard care (control group) for each participant population (patients/caregivers) separately. In total, 4 comparisons are performed for one outcome variable (alpha=0.0125). The hypotheses related to the first primary outcome (emotional functioning) will be tested using a mixed model (per participant population) with the T1 measurement value for emotional functioning as outcome variable, recruitment center as random effect and randomization group and baseline measure of emotional functioning (T0) as predictor variables. Analyses will be performed on both 'intention-to-treat' and per-protocol principles. The primary principle is intention-to-treat. After completion of the baseline measurement (T0), dyads will be randomized to one of the trial arms. All randomized dyads will be included in the mixed model. Multiple imputation will be applied. Predictors for the imputation model will include the baseline measurement, randomization group, age and other variables (e.g. severity of the illness). The secondary principle is the per-protocol analysis that functions as a sensitivity analysis. The per-protocol population will be defined as dyads who have completed all sessions of the FOCUS+ or iFOCUS intervention (except for dyads in the control group) and T1 measurement. By including the baseline measurement as a predictor variable (ANCOVA), preexisting differences will be controlled, enhancing the sensitivity of the analyses. To interpret the magnitude of the effects for the different outcomes, effect sizes (Cohen's d) will be estimated. Testing the null hypothesis of the second primary endpoint: self-efficacy (the Lewis´ Cancer self-efficacy scale from FOCUS) As per the fixed sequence (FS) procedure, the null hypotheses of the second primary endpoint (self-efficacy) will only be tested if a significant result is found for the first primary endpoint (emotional function). The same strategy is then followed for the analyses as for the first primary endpoint, with an alpha level of 0.0125. Secondary hypotheses testing: All identified secondary endpoints (Quality of Life [including separate items of hopelessness, anxiety, depression], benefits of illness, coping, dyad communication, all at t1) will be evaluated by testing the FOCUS+ and iFOCUS will against care as usual (control group) for each participant population (patients/caregivers) separately. In total, 4 comparisons are performed for each outcome variable. For each secondary outcome variable a mixed model is applied (per participant population) with the T1 measurement value as outcome variable, recruitment center as random effect and randomization group and baseline measurement of the variable (T0) as predictor variables. Analyses will be performed on both 'intention-to-treat' and per-protocol principles, applying the same principles as described above. By including the baseline measurement as a predictor variable (ANCOVA), preexisting differences will be controlled, enhancing the sensitivity of the analyses. To interpret the magnitude of the effects for the different outcomes, effect sizes (Cohen's d) will be estimated. All statistical tests will be two-sided and considered significant if p< 0.0125. All primary outcomes and secondary outcomes as listed above will also be analysed at T2 (6 months) to evaluate longer term effects, using the same analysis procedures. The cost-effectiveness of the interventions will be determined by analyzing patterns and costs of healthcare utilization and effects on quality of life (measured by the EORTC, EQ5D5L, FACT G and CQOLC). Data will also be collected on the types and amounts of informal care provided to patients in each arm of the study, to investigate if amount or patterns of informal care change as a result of the intervention. The outputs will be mean costs of care for patients in each arm of the study, cost per year of life gained (if survival is affected significantly by the intervention and the costs in the intervention groups overall are higher) and (if appropriate) the additional costs of achieving better quality of life outcomes (including estimates of cost per quality adjusted life year gained). Exploratory hypotheses testing: For all exploratory endpoints, two-sided statistical tests will be considered significant if p< 0.05 For the outcomes that are measured identical for the patient and the caregiver, the effect on the dyad as a whole (i.e. both patient and family caregiver) will be assessed. For the outcome instruments that led to comparable estimated differences between FOCUS+ and standard care and iFOCUS and standard care, the effect will be assessed on the dyad as a whole by adding an extra level (dyad) to the linear regression model. For each of the primary and secondary endpoints, subgroup analyses will be performed using formal interaction tests to explore the extent to which the outcomes of the trial differ by country, gender and socioeconomic status. Interaction terms between respectively country, gender and socioeconomic status on the one hand and the trial arms on the other hand will be added to the analysis models. For the country variation a multilevel mixed model analyses will also be performed to additionally account for potential clustering by country (i.e. participants nested within a country). Outcomes will be analyzed with country as random factor. Other analyses: Background reports describing care as usual for people with advanced cancer will facilitate the understanding of the results of the between-country comparisons. Process evaluation of the implementation of the interventions will be analyzed following the MRC framework for evaluating complex interventions, integrating normalization process theory (NPT) and the RE-AIM framework. Data analysis for the process evaluation will include a) standard statistical descriptions of the quantitative data from the intervention checklist and routine monitoring to describe adherence to the implementation. This analysis will determine cut-off points for good intervention adherence and, hence, inform the per-protocol analyses; b) analyses of the qualitative data (semi-structured interviews with patients and their family caregiver and post-intervention interviews with the nurses who delivered the face-to-face FOCUS+ intervention) will be performed (see below - Qualitative analysis). Qualitative analysis: With the transcription of interviews into the local language, the analysis process will involve a collaborative process involving researchers from each partner site collecting data. Thematic analysis allows for both inductive and deductive analysis and can be implemented with a range of computer-based software to support the management of the analysis process (e.g., NVIVO, MAXQDA). Deductive analysis will be informed by semantic information sought from the interview (i.e., were participants satisfied, where particular elements of the programmes described as positive or negative) and themes evident in previous evaluations of the FOCUS intervention. This will involve developing themes in advance of the analysis process and assessing the presence or absence of these themes across the data. Inductive analysis will be structured using the objectives of the process evaluation to target key topics, with more latent or interpretative themes isolating more experiential findings from the data. Qualitative analysis will be conducted at two levels, an initial assessment of themes in each data source (stakeholders, staff, researchers, patients and carers, different language groups) followed by a higher-level analysis of superordinate themes of convergence and divergence evident across groups. Additional strategies for managing the potential impact of multilingual analysis are recommended, including peer debriefing during the process of coding and the development of candidate themes, triangulation across researchers and language sources. Analysis will be informed by open discussion of conceptual issues in the data to explore variations in interpretation and identify shared meaning relevant to the focus of the process evaluation.

Tracking Information