Patient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service Experience (PROMISE)–protocol for a randomised controlled trial of electronic self-reporting of symptoms versus usual care during and following treatment in patients with cancer Authors: Penelope M Webb, Amy Brown, Bena Brown, Louisa G Collins, Fiona Crawford Williams, Kerrie Doupain, Melissa Eastgate, Vicki Fennelly, Afaf Girgis, Gunter Hartel, Rahul Ladwa, Karen Martin, Robert Mason, Peter McGuire, Elizabeth Miller, Suzanne O’Brien, Rebecca Packer, Mark B Pinkham, Sabe Sabesan, Jasotha Sanmugarajah, Georgia Slapp, Doreen Tapsall, Jeanie White, Laurelie R Wishart, David Wyld, Raymond J Chan Published: 2024 ABSTRACT AbstractIntroductionRoutine collection of patient-reported outcome measures (PROMs) has the potential to inform and improve cancer care. It is now feasible for patients to complete PROMs electronically (ePROMs) providing information about their current levels of symptoms, side effects of treatment and other concerns. PROM scores can be tracked over time allowing more timely identification of problems and more appropriate intervention. Studies have reported clear benefits in patient–clinician communication when PROMs are used and trials in the USA and France found patients randomised to complete regular ePROMs reported better health-related quality of life, had fewer unplanned hospital visits and, importantly, significantly better survival than those randomised to usual care. However, information about the effects on health outcomes and, particularly, the cost-effectiveness of incorporating this information into practice is limited.Methods and analysisPROMISE (Patient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service Experience) is a multicentre, randomised hybrid effectiveness/implementation trial to evaluate the clinical and cost-effectiveness of using ePROMs in routine cancer care to improve patient outcomes. Participants (target sample=572; randomised 1:1 to intervention and control) are adults aged 18 years or older diagnosed with a solid cancer and starting treatment at one of the four study hospitals. The primary outcomes are unplanned hospital presentations and physical/functional well-being at 6 months. We hypothesise that, compared with usual care, patients randomised to use an ePROM tool will have fewer unplanned hospital presentations, report better health-related quality of life and greater satisfaction with their care and that the ePROM tool will be cost-effective. We will also assess implementation and process outcomes consistent with the RE-AIM (Reach Effectiveness Adoption Implementation Maintenance) Framework.Ethics and disseminationThis trial has been approved by the Metro South Human Research Ethics Committee (HREC/2020/QMS/67441). Participants provide written informed consent, including consent for record linkage, prior to completing the baseline questionnaire. Study results will be disseminated via peer-reviewed journals and presentations at scientific conferences and clinical meetings.Trial registration numberACTRN12620001290987. STRENGTHS AND LIMITATIONS OF THIS STUDY Hybrid effectiveness-implementation design to evaluate the use of electronic patient-reported outcome measures (ePROMs) in a close to real-world setting. Hybrid effectiveness-implementation design to evaluate the use of electronic patient-reported outcome measures (ePROMs) in a close to real-world setting. Conducted across four major centres: two in the state capital, one in a metropolitan region 1 hour from the capital and one 15 hours from the capital that serves a significant regional area. Conducted across four major centres: two in the state capital, one in a metropolitan region 1 hour from the capital and one 15 hours from the capital that serves a significant regional area. Comprehensive evaluation of implementation processes and both clinical and cost-effectiveness. Comprehensive evaluation of implementation processes and both clinical and cost-effectiveness. Qualitative evaluation of patient, carer and health professional perspectives. Qualitative evaluation of patient, carer and health professional perspectives. The current versions of the ePROM tools were only available in English, so non-English speaking patients had to be excluded. The current versions of the ePROM tools were only available in English, so non-English speaking patients had to be excluded. Introduction With the numbers of new cancer diagnoses and cancer survivors increasing,1 it is imperative we care for those affected in the most cost-effective way. There is considerable interest in the use of patient-reported outcome measures (PROMs) to inform cancer care. The hope is this will improve timely identification of symptoms and subsequent access to appropriate care, thereby enhancing physical and psychological well-being, improving health service experience, increasing treatment completion rates and, ultimately, improving survival. A patient-reported outcome (PRO) is an indicator of a patient’s health status, as reported by the patient without amendment or interpretation by a health professional or other person.2 PROMs provide information about a patient’s condition and response to treatment (eg, symptoms of underlying disease and treatment toxicities), functioning and well-being (eg, physical functioning, social/emotional well-being) and psychological symptoms (eg, depression and anxiety). Advances in technology and internet access now allow patients to complete PROMs electronically (ePROMs), providing their clinical team with regular information about their current symptoms, well-being and other concerns. Their ePROM responses can be tracked over time, potentially allowing more timely identification of, and intervention to resolve problems. The routine use of ePROM tools in cancer care has attracted considerable interest since a US trial reported that patients starting chemotherapy for metastatic solid cancers who were randomised to complete ePROMs that generated email alerts when patients reported severe or worsening symptoms, had significantly improved survival compared with those randomised to usual care (median 31 vs 26 months, p=0.03).3 The authors had previously reported improved health-related quality of life (HRQoL) and longer treatment duration among the intervention group.4 A French trial has since reported that, among patients with non-progressive disease at the end of primary treatment for advanced lung cancer (n=121), survival was significantly better among the group randomised to complete weekly ePROM measures compared with usual care (median 22.5 vs 13.5 months, one-sided p=0.002)5 6 and the intervention was highly cost-effective.7 However, that study had to be terminated early when the survival difference exceeded preset criteria in a prespecified interim analysis. Three of four non-randomised studies have also reported a survival benefit, but a 2022 systematic review concluded that the current evidence was insufficient to conclude that routine PRO monitoring would improve survival.8 Studies have also reported benefits for other patient outcomes, but the results have not always been consistent, and heterogeneity between the interventions, reporting and outcomes assessed makes it hard to assess efficacy.9 Although the trials suggest routine use of ePROMS has potential to improve patient survival, they were conducted in highly controlled settings and more pragmatic trials are required to investigate ePROM implementation across a broader range of settings. One example is PROMPT-Care, an Australian system designed to collect ePROM data from patients with cancer to alert clinicians to problems and support patient self-management.10 Non-randomised studies in the real-world setting have demonstrated satisfactory implementation and acceptability, from the perspective of both patients and healthcare professionals11 and fewer ED presentations among PROMPT-CARE users compared with a matched control group.12 However, there is an overall lack of high-quality hybrid implementation-effectiveness trials in the literature. The traditional route to translation requires separate studies to evaluate clinical effectiveness and implementation of the intervention in practice. This stepwise process is necessarily time-consuming leading to designs that allow both steps to be conducted in a single study.13 PROMISE (Patient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service Experience) is a pragmatic hybrid effectiveness-implementation trial to test the effectiveness/cost-effectiveness and implementation of ePROMs in routine cancer care in Queensland, Australia. Through this large hybrid trial, we aim to add novel insights by examining factors affecting implementation success. Methods and analysis This manuscript was written in accordance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines.14 The full study protocol is on the PROMISE study website (https://www.qimrberghofer.edu.au/promise-study/). PROMISE is a pragmatic investigator-initiated multisite, two-arm randomised controlled hybrid effectiveness-implementation trial of two ePROM tools. Our primary hypotheses are that, compared with usual care, routine self-reporting of ePROMs will result in: fewer emergency presentations and unplanned hospital admissions, and better physical and functional well-being at 6 months. Secondary hypotheses are that ePROMs will lead to higher rates of treatment completion, better HRQoL, lower symptom burden and fewer out-patient visits and that use of the ePROM tools will be cost-effective. Additional exploratory hypotheses are that ePROMs will lead to a better patient care experience, lower rates of distress and unmet supportive care needs among patients’ partners/carers and, ultimately, better survival. We will also assess implementation and process outcomes consistent with the revised Reach Effectiveness Adoption Implementation Maintenance (RE-AIM) Framework15 16 (table 1). The RE-AIM framework16 and Consolidated Framework for Implementation Research (CFIR)17 will guide examination of the complex interplay between intervention components, implementation strategies and contextual factors in the context of the RE-AIM dimensions.18 Table 1Outcomes and information to be collected to address the RE-AIM indicatorsRE-AIM indicator and level of analysisCollection method/instrumentREACH and representativenessStudy and centreParticipation rates (%), comparison of participants and eligible non-participantsStudy database, hospital record auditPatient clinical and demographic characteristics (overall and by participation)Study database; hospital records; comparison to state-wide data from QCCATEffectivenessPrimary outcomes:SystemThe rate of unplanned hospital presentations/admissionsHospital records; linkage to QHAPDC, QHNAPDC, QEDC, MBSPatientPhysical and functional well-being at 6 monthsSelf-report: FACT-G: physical and functional well-being subscales39Secondary outcomes:PatientRates of treatment completion and proportion of planned treatment received (dose reductions, delays)Medical recordsOverall well-being and change in well-beingSelf-report: FACT-G (33 items)39Distress (anxiety and depression)Self-report: HADS (14 items)40 41Symptom burdenSelf-report: MSAS (32 items)42Numbers of out-patient visitsHospital records; linkage to QEDC, MBSSystemCost-effectivenessQHAPDC, QHNAPDC, QEDC, MBS, PBS, NCDC, EQ-5D-5L43 44Exploratory outcomes:PatientPatient experience of careSelf-report: AHPEQS (12 items)45SurvivalHospital recordsPartner/carer distress and unmet needsSCNS-P&C46AdoptionCentreNumbers of clinics, medical, nursing and allied health clinicians invited to participate and number participating in implementing ePROMsHospital and study recordsImplementationPatientePROMs completed (N, %)ePROM platformePROM items completed (%)ePROM platformPatient feedback on system usabilityStakeholder interviews, System Usability Scale47CentreFrequency of clinician access to ePROM data (N, %)ePROM platformLength of medical appointments at the cancer centreHospital recordsClinical responses related to ePROMsHospital recordsGP and specialist visits (N, timing)MBS; hospital recordsHealth provider feedback on system usability and implementationStakeholder interviewsBenefits in quality-adjusted life yearsSelf-report: EQ-5D-5L43 44Maintenance and sustainabilityProportion of the clinics, clinicians and intervention participants using the ePROMs at the end of the study, 6 months and 12 months after study completionClinician surveys and ePROM platform usage statisticsAHPEQS, Australian Hospital Patient Experience Question Set; ePROMselectronic patient-reported outcome measuresEQ-5D-5L, EuroQol-5 Dimension 5-level; FACT-G, Functional Assessment of Cancer Treatment-General; GPgeneral practitionerHADS, Hospital Anxiety and Depression Scale; MBS, Medicare Benefits Schedule; MSAS, Memorial Symptom Assessment Scale; NCDC, National Cost Data Collection; PBS, Pharmaceutical Benefits Scheme; QCCAT, Qld Cancer Control Analysis Team; QHAPDC/QHNAPDC/QEDC, Queensland Hospital Admitted Patient/Non-Admitted Patient/Emergency Data Collection; RE-AIMReach Effectiveness Adoption Implementation MaintenanceSCNS-P&C, Supportive Care Needs Survey-Partners and Carers Outcomes and information to be collected to address the RE-AIM indicators AHPEQS, Australian Hospital Patient Experience Question Set; ePROMselectronic patient-reported outcome measuresEQ-5D-5L, EuroQol-5 Dimension 5-level; FACT-G, Functional Assessment of Cancer Treatment-General; GPgeneral practitionerHADS, Hospital Anxiety and Depression Scale; MBS, Medicare Benefits Schedule; MSAS, Memorial Symptom Assessment Scale; NCDC, National Cost Data Collection; PBS, Pharmaceutical Benefits Scheme; QCCAT, Qld Cancer Control Analysis Team; QHAPDC/QHNAPDC/QEDC, Queensland Hospital Admitted Patient/Non-Admitted Patient/Emergency Data Collection; RE-AIMReach Effectiveness Adoption Implementation MaintenanceSCNS-P&C, Supportive Care Needs Survey-Partners and Carers Study sites and study population The target population is adults (age≥18 years) diagnosed with a solid cancer and starting systemic anticancer or radiation therapy at Princess Alexandra Hospital (PAH), Royal Brisbane and Women’s Hospital (RBWH), Gold Coast University Hospital (GCUH) or Townsville University Hospital (TUH) in Queensland, Australia. These sites were selected because PAH and RBWH are quaternary public health hospitals in the state capital that, together, deliver outpatient care to approximately 3000 patients with cancer each week. They have developed the ePROM tools we will use (My Health My Way and AboutMe). GCUH and TUH are large tertiary centres: GCUH is in a metropolitan region 1 hour from the capital, while TUH is 15 hours from the capital and serves a more geographically dispersed rural population. For these reasons, these public hospitals are well placed to optimise and influence cancer care delivery in Queensland, and to inform a state-wide approach to the use of ePROMs. The study opened at PAH in August 2021, RBWH in October 2021 and at GCUH and TUH in June 2022 and recruitment closed at the end of June 2023; intervention and follow-up are ongoing. Eligibility and exclusion criteria To be eligible, participants had to meet all of the eligibility criteria and none of the exclusion criteria listed in table 2. Table 2Inclusion and exclusion criteriaInclusion criteriaExclusion criteriaAge at least 18 yearsDiagnosed with an invasive solid cancerAbout to start (within 2–3 weeks) or recently started (eg, received no more than 1 cycle of chemotherapy or <25% of planned radiation) treatment (primary or for recurrence) at one of the study sitesSite-specific criteria:Will not return to the site for regular treatment/follow-upAlready using or would normally use an ePROM tool at that siteCancer type is the target of a pre-existing trial of patient care at that siteLife expectancy less than 3 monthsTreated with surgery onlyCurrent participation in another study that requires completion of similar PROMsUnable to provide informed consentUnable to use the ePROM tools (eg, no internet access, a comorbid condition affecting vision, cognition or dexterity)Unable to complete questionnaires in EnglishUnwilling to consent to data linkageePROMselectronic patient-reported outcome measuresPROMspatient-reported outcome measures Inclusion and exclusion criteria Will not return to the site for regular treatment/follow-up Will not return to the site for regular treatment/follow-up Already using or would normally use an ePROM tool at that site Already using or would normally use an ePROM tool at that site Cancer type is the target of a pre-existing trial of patient care at that site Cancer type is the target of a pre-existing trial of patient care at that site ePROMselectronic patient-reported outcome measuresPROMspatient-reported outcome measures Recruitment Participant flow is shown in figure 1. Site Coordinators identified all patients attending participating clinics during the study period who met the inclusion criteria. Patients who met one of the exclusion criteria were flagged and the reason for exclusion recorded. The Site Coordinators approached the remaining patients and, if patients were interested in participating, performed a final eligibility screen and recorded any additional reasons for exclusion. They then gave patients who passed the screening the PROMISE Participant Information sheet and Consent Form (PICF, see online supplemental material), explained the study to them and answered any questions. All participants were asked to consent to allow the study team access to their medical record, for linkage to hospital and emergency department databases and, separately, for linkage to the Australian Medical Benefits Schedule (MBS, private outpatient consultations, investigations and procedures) and Pharmaceutical Benefits Scheme (PBS, prescription data) via Services Australia. When a patient consented, this was recorded in the Study Database and they were assigned a unique study ID number. This initiated a personalised link asking them to complete the Baseline Study Questionnaire. This provides sociodemographic information (eg, education level, post code of usual residence, marital and employment status, household income, private health insurance status, dependents/household structure), medical/health history (eg, height, weight and existing comorbidities) and baseline measures of well-being (see ‘Outcomes and timing of assessments’, below). If they did not complete this within 7 days, they were sent a reminder, followed by a telephone call 7 days later. If the questionnaire had still not been completed after a further 7 days, the participant was withdrawn from the study. PROMISE trial design. *Linkage to hospital and emergency records when all participants have reached this time-point. PROMISE, Patient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service Experience. Randomisation and blinding Randomisation (1:1) was at the patient level, using randomly varied permuted blocks, stratified by sex, hospital, the ‘surprise question’ (would the clinician be surprised if the patient died within 1 year) as a validated measure of prognosis,19 and computer literacy (defined as regular access to a computer/tablet/smart phone and use of email≥1/week vs less often) as in previous trials some benefits were greater for those with lower computer literacy.3 4 When a participant completed the Baseline Questionnaire, the central web-based database automatically applied the next relevant record in the allocation sequence, as we have done previously.2022 All study staff/investigators were blinded to the allocation sequence ensuring allocation concealment until study entry. Due to the nature of the intervention, participants, their treating clinicians and the Site Coordinators could not be blinded to group allocation; however, the other investigators and project staff, including those conducting analyses, will remain blinded until after the main analyses have been finalised and the first papers are being prepared for publication. PROMISE-carers substudy After randomisation, participants were invited to nominate their partner/carer to participate in the carer substudy to evaluate whether the ePROMS would also reduce their distress. Nominated partners/carers were given a copy of the PROMISE carer PICF (see online supplemental material and invited to participate. When a partner/carer consented, they were assigned a study ID number linked to the patient and this initiated a personalised link asking them to complete the Baseline Carer Questionnaire with sociodemographic information similar to that collected for patients and baseline measures of well-being (see ‘Outcomes and timing of assessments’, below). If they did not complete the questionnaire within 7 days, they were followed up as for patients (above). Intervention The full PROMISE intervention combines the ePROM tools that participants complete and the associated implementation strategies. In addition to their usual care (below), participants randomised to the intervention were oriented to the use of the relevant ePROM tool (My Health My Way at PAH, TUH; AboutMe at RBWH, GCUH) at their first visit postrandomisation. Both ePROM tools include two validated instruments (Distress Thermometer (DT)23 and Edmonton Symptom Assessment Scale (ESAS)24) that have been used in previous studies and are used in the PROMPT-care system in New South Wales.10 11 25 Both systems automatically generate alerts if patients report symptoms that exceed the defined thresholds and they generate summary reports detailing the most recent symptoms entered and their severity as well as trends over time. Hospital-specific protocols (codesigned with clinical staff) identified the frequency with which patients were sent reminders to complete the ePROMs, threshold scores, management pathways to respond to scores above these thresholds, and clinically appropriate actions and recommendations for patient review/referral. It was, however, left to the treating clinicians to decide and prioritise management according to individual patient need. My Health My Way operates via a secure Qualtrics platform and is managed by PAH. The tool can be accessed from anywhere by patients (via unique links sent to them) and approved healthcare providers. The system has been adapted for PROMISE to identify trial participants and allow access by appropriate staff at TUH. Built-in reminders prompt participants to complete the tool at prespecified intervals: weekly at PAH/fortnightly at TUH for the first 3 months, and then monthly. If a patient’s responses to the ePROM tool exceed predefined thresholds for various disciplines (Nursing, Medical, Dietetics, Speech Pathology, Physiotherapy, Occupational Therapy and Social Work), the system automatically sends an email alert to the relevant discipline to review and take appropriate action. For example, weight loss above a set threshold would send an alert to the Dietetics team, with the urgency of the alert (low, medium, high, urgent) depending on the amount of weight lost. Alerts relating to allied health were managed directly by allied health staff and those related to medical or nursing were monitored by the Site Coordinators who then alerted the relevant clinicians. These alerts are routinely monitored, and actions performed in response to the alert are recorded. As a safeguard, patients who enter scores that exceed the thresholds are advised that the ePROM tool is not monitored outside working hours and they should also contact their clinical team. AboutMe is managed at RBWH and operated via a Queensland Government QGov account to ensure data are stored and transmitted securely. When participants are randomised to the intervention group, the local Trial Coordinator helps them set up a QGov log-in and shows them how to use the AboutMe system. As for My Health My Way, the tool can be accessed from anywhere by patients and approved healthcare providers. Reminder messages are sent monthly for patients to the ESAS and 3-monthly for the DT; however, AboutMe users can also login and complete the surveys more frequently if they wish to. The system automatically identifies scores that breach the set thresholds. Site Coordinators monitor the system daily and notify the appropriate clinic staff when breaches occur. Patients who entered scores that breached thresholds were also provided contact details to seek support. Usual care Prior to commencing treatment, all participants receive education regarding the possible symptoms and side effects relevant to their cancer type and treatment, as per standard care. They attend their routine medical and supportive care appointments during treatment and follow-up, with regular medical reviews (by the treating oncologist) and nursing/allied health intervention at standard intervals depending on cancer type and treatment. As per standard care, patients who experience symptoms of concern are advised to contact their treatment team. Implementation The implementation strategy was informed by the revised RE-AIM framework16 and identified priorities in relation to ePROM implementation.26 The process and project personnel responsible for each stage in this process are summarised in table 3. Table 3Implementation strategyPretrial1Establish project and prepare environment (site PIs)Baseline survey: current workflows and use of PROMsUse a codesign approach to finalise core content of ePROM tools with site PIsObtain support from key opinion leaders (eg, medical and nursing executives)Present an overview of the PROMISE intervention to the clinical staffBegin the preparation for on-boarding/trainingIdentify clinical champions and develop strategy to maximise RE-AIM2Implementation planning (site PIs)Codesign to reach consensus on local procedures, management pathways and recommendations for actionEnsure infrastructure is in place3Develop appropriate guidance and procedures documentation (PM)Ensure documentation (practice guidance, trial and implementation procedures flowcharts, etc) is congruent with those agreed in step 24Implement on-boarding/training with all clinical staff (PM)Undertake group-based, face-to-face meetings with all participating clinicians (offering individual sessions to those who cannot participate in group sessions)During trial5Clinical implementation (PM and Site Coordinators)Facilitate and implement clinical workflowProvide clinical support to maximise the uptake of agreed clinical workflowsOversee implementation process, monitor challenges, provide feedbackIdentify and resolve issues6Periodic audit and review of RE-AIM outcomes (PIs, PM, PROMISE team)Report to site PIs and PROMISE leadership group at regular intervalsPost-trial7Facilitate maintenance beyond trial—if found to be effective (CIs and PM)Apply trial learnings to improve implementation of ePROMS (site-specific)Facilitate establishment of ePROMS as a routine procedure at the clinicsCI, chief investigator; e-PROMselectronic patient-reported outcome measuresPM, project managerPROMISEPatient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service ExperienceRE-AIMReach Effectiveness Adoption Implementation Maintenance Implementation strategy Baseline survey: current workflows and use of PROMs Baseline survey: current workflows and use of PROMs Use a codesign approach to finalise core content of ePROM tools with site PIs Use a codesign approach to finalise core content of ePROM tools with site PIs Obtain support from key opinion leaders (eg, medical and nursing executives) Obtain support from key opinion leaders (eg, medical and nursing executives) Present an overview of the PROMISE intervention to the clinical staff Present an overview of the PROMISE intervention to the clinical staff Begin the preparation for on-boarding/training Begin the preparation for on-boarding/training Identify clinical champions and develop strategy to maximise RE-AIM Identify clinical champions and develop strategy to maximise RE-AIM Codesign to reach consensus on local procedures, management pathways and recommendations for action Codesign to reach consensus on local procedures, management pathways and recommendations for action Ensure infrastructure is in place Ensure infrastructure is in place Ensure documentation (practice guidance, trial and implementation procedures flowcharts, etc) is congruent with those agreed in step 2 Ensure documentation (practice guidance, trial and implementation procedures flowcharts, etc) is congruent with those agreed in step 2 Undertake group-based, face-to-face meetings with all participating clinicians (offering individual sessions to those who cannot participate in group sessions) Undertake group-based, face-to-face meetings with all participating clinicians (offering individual sessions to those who cannot participate in group sessions) Facilitate and implement clinical workflow Facilitate and implement clinical workflow Provide clinical support to maximise the uptake of agreed clinical workflows Provide clinical support to maximise the uptake of agreed clinical workflows Oversee implementation process, monitor challenges, provide feedback Oversee implementation process, monitor challenges, provide feedback Identify and resolve issues Identify and resolve issues Report to site PIs and PROMISE leadership group at regular intervals Report to site PIs and PROMISE leadership group at regular intervals Apply trial learnings to improve implementation of ePROMS (site-specific) Apply trial learnings to improve implementation of ePROMS (site-specific) Facilitate establishment of ePROMS as a routine procedure at the clinics Facilitate establishment of ePROMS as a routine procedure at the clinics CI, chief investigator; e-PROMselectronic patient-reported outcome measuresPM, project managerPROMISEPatient Reported Outcome Measures in cancer care: a hybrid effectiveness-Implementation trial to optimise Symptom control and health service ExperienceRE-AIMReach Effectiveness Adoption Implementation Maintenance Outcomes and timing of assessments The trial outcomes will be assessed according to the RE-AIM framework which includes five key components: Reach, Effectiveness, Adoption, Implementation and Maintenance.15 16 Effectiveness and cost-effectiveness will be measured via a series of primary, secondary and exploratory outcomes at the patient and system level. These will be assessed via self-reported online questionnaires, clinical data collection and linkage to routine data collections. Reach, Adoption, Implementation and Maintenance will be assessed via recruitment and process data, qualitative data from stakeholder interviews and ePROM platform usage statistics. The outcomes and source(s) of information that will inform them are summarised in table 1. Questionnaire data These are collected via Qualtrics. At 3, 6, 12, 18 and 24 months, all participants (Intervention and Control) are sent a link and asked to complete the appropriate Follow-Up Questionnaire online (figure 1). If they do not complete this within 7 days, they are sent a reminder. If the questionnaire has still not been completed after a further 7 days, a project officer at QIMR Berghofer calls the participant to encourage them to complete it. All data are stored on a secure server with questionnaire data kept separate from personal information, which is only accessible to key project staff. Clinical data collection Information about the participant’s cancer (histology, grade and stage of disease at diagnosis), date of diagnosis, treatment (surgery, chemotherapy and radiotherapy doses and dates, side effects experienced and response to treatment), recurrence of disease (basis of diagnosis, dates and treatment) and vital status (date and cause of death, if relevant) is obtained from medical records at baseline and annually thereafter. Information about hospital admissions, emergency department presentations, procedures and prescriptions will be obtained via linkage to the Queensland Hospital Admitted Patient, Non-Admitted Patient and Emergency Data Collections and the national MBS and PBS databases. Stakeholder interviews Samples of patients, partners or carers and healthcare workers are invited to take part in a one-off, individual semistructured qualitative interview regarding their experiences using the ePROM tool. When they initially consented to the study, participants and partners/carers were advised that they might be invited to participate in an interview but that they were not obliged to agree to this. Healthcare providers invited to take part in interviews are given a separate Participant Information Sheet explaining what this will involve. Approximately 30 participants will be selected for the qualitative interviews based on maximum variation sampling. Maximum variation sampling is used when a range of participants is required and involves selecting information-rich cases to capture a broad range of perspectives. The following characteristics will be sought: role (patient, carer, health professional (medical, nursing, allied health), clinical site, time in trial and clinical adopters versus non-adopters. Interviews are conducted either face-to-face or via tele/video-conferencing by a researcher skilled in qualitative interview techniques. Participants are encouraged to speak openly and to describe their experience with the ePROM tools/system. The interview guide was informed by the Consolidated Framework for Implementation Research (CFIR) Interview Guide Tool (https://cfirguide.org/guide/app/#/). Where appropriate, interview questions may be adapted, omitted and/or elaborated on an individual basis to encourage participants to reflect openly about their experience. Interviews will be transcribed verbatim. It is anticipated that interviews will take no longer than 30 min. Sample size Our sample size has been calculated to achieve a minimum 80% power (a=0.05) for our primary outcomes. Previous studies have reported reductions in emergency presentations of 17%–30%4 (personal communication). Assuming a similar baseline rate (1.09/person-year), we require 242 participants per group to detect a 20% reduction with 18 months follow-up. For self-reported measures of well-being, we require 70–150 patients per group to detect minimally important differences (or 0.5 SD if there is no defined minimally important difference)27 depending on the outcome. A sample of 200 per group will give >90% power for the primary outcomes in the full population and ~80% power for analyses stratified by ePROM tool. A sample of 572 participants (286 per group) allows for up to 30% dropouts (death or withdrawal). The recruitment targets are 150–175 at PAH and RBWH and 110–145 at GCUH and TUH. Statistical analysis Data analysis will be guided by a formal statistical analysis plan that will be finalised and approved by all investigators prior to any analyses being conducted. The primary analyses will be intention-to-treat, but per-protocol analyses will also be performed and both results will be reported. To preserve the blinding during preliminary analyses, the Data Manager will generate data sets with participants’ true allocation code replaced by a random dummy code. This will be replaced by the true code when we perform the final analyses (as in previous studies).22 Effectiveness We will use Poisson regression to calculate and compare rates of unplanned hospital presentations and outpatient visits (per 100 patient-days) during follow-up in each group. For measures of well-being, distress and symptom burden, we will calculate the relevant scores at each time point; missing items will be treated according to the relevant scoring guidelines. We will compare mean physical and functional well-being scores from Functional Assessment of Cancer Treatment-General (FACT-G) in the two groups at 6 months (our primary outcome) using t-tests, followed by general linear models to incorporate baseline covariates, including site, ePROM tool, type of cancer, type of treatment (eg, radiation therapy or chemotherapy). A difference of 0.5 SD on each scale is considered clinically meaningful.27 Differences in the secondary/tertiary outcomes, for carers and at other time-points will also be assessed and we will explore changes over time and differences between treatment arms using mixed effects repeated measures models. Model for post-randomisation time-points will be adjusted for baseline score and stratification factors. To assess treatment completion, we will calculate the fraction of planned primary treatment that participants actually received. We will compare the mean (or median, as appropriate) fraction received and the proportion of participants who received >85% of their planned treatment between the groups using t-tests and χ2 tests, respectively. For patient care experience, we will use χ2 tests to compare the proportions of participants in each group that answer always or very good (the ‘top box’ responses) to individual items on the Australian Hospital Patient Experience Question Set (AHPEQS)28 and t-tests to compare the mean number of items rated as always/very good. We will also compare the proportions of participants who report experiencing physical and/or emotional distress as a part of their care and, among those who report distress, the proportions who reported that healthcare staff discussed this with them. To evaluate survival, we will calculate the mortality rate (per person-year) in each group and use Cox proportional hazards and flexible non-parametric survival models to compare the two study groups. We will repeat the primary and secondary analyses above stratifying participants by (1) study site, (2) ePROM tool (My Health My Way vs AboutMe), (3) type of treatment (curative/palliative), (4) sex and (5) intervention adherence. This is to examine effect modification and identify scenarios where ePROM tools are potentially most effective and those where their effectiveness could be improved. We recognise that statistical power will be more limited for some of these comparisons. Cost-effectiveness We will conduct a cost-utility analysis to compare the costs and health benefits associated with the ePROM intervention from a health provider perspective. A Health Economics Analysis Plan will be developed prior to data analyses and the evaluation will conform to the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) Checklist.29 Healthcare resource use captured from hospitalisation and MBS/PBS data will be quantified and valued in 2024 Australian dollars, adjusted for inflation where necessary. Health utilities will be assessed using the EuroQol-5D 5-level version (EQ-5D-5L) and the Australian scoring algorithm. Utility scores are multiplied by survival time to generate quality-adjusted life years (QALYs). Using the area-under-the-curve method and health utility scores at each time point, a single QALY for each person will be calculated at the last time point.30 Incremental cost per QALY ratios will be calculated by dividing the mean difference in costs between the study groups by the mean difference in QALYs. A resampling method using single imputation nested in bootstrapping will be used to assess uncertainty around the ratio values.31 CIs (95%) around ratios for the difference in mean costs and QALYs will be generated through the bootstrap percentile method using 1000 bootstrap replications.32 A willingness-to-pay threshold of AU$50 000 per QALY gained will be used to interpret the incremental cost per QALY ratio, with costs below this considered cost-effective in Australia.33 Implementation To maximise implementation learnings, we will conduct a range of analyses. To address the RE-AIM indicators, quantitative measures (table 1) will be summarised and reported using descriptive statistics, aggregated overall and by study site. Qualitative interviews will be analysed using both inductive and deductive methods to enhance our ability to explore reasons for successes and challenges to effective implementation. Inductive analysis will use reflexive thematic analysis as described by Braun and Clarke.34 35 Deductive analysis will be conducted utilising both using the CFIR and RE-AIM.17 36 Coding will be performed with the assistance of the NVivo V.12 software programme. To ensure rigour and agreement in the data, two researchers will code the transcripts independently and discuss any discrepancies. After all interviews have been coded, the analysis will be reviewed by a third researcher. Final results will be discussed between the three researchers to reach consensus on data interpretation and major themes. Drawing on data from multiple sources (descriptive statistics, problem logs and study documentation from the Site Coordinators and Project Manager, qualitative interviews and meeting notes), we will use fishbone analysis, also known as Ishikawa diagrams,37 38 to identify potential factors associated with implementation. We will do this for the trial overall and, where there are sufficient data, for each site individually. The ‘mouth’ of the fish will represent the implementation outcome, and the ‘bones’ will display various categories including contextual factors, intervention components, and implementation strategies. Patient and public involvement Two consumer partners with lived experience actively contributed to the protocol development and were involved in all research processes including research team meetings. One has now had to withdraw (MD) but the other (EM) will continue to be involved in all discussions. Ethics and dissemination This trial has been approved by the Metro South Human Research Ethics Committee (HREC/2020/QMS/67441) and participants provide written informed consent, including consent for record linkage, prior to completing the baseline questionnaire. When they enrol for the study, participants and partners/carers are asked if they wish to receive a copy of the main study results when these are available. Those who request this will be sent a summary written in lay language, to coincide with the first formal presentation of the scientific results. Study results will be disseminated via peer-reviewed journals and presentations at scientific conferences and clinical meetings. A publication policy has been prepared and approved by all investigators. supplementary material