Open access Protocol Comparing the effectiveness and cost-­ effectiveness of self-­management interventions in four high-­priority chronic conditions in Europe (COMPAR-­EU): a research protocol Marta Ballester ‍ ‍,1,2,3 Carola Orrego,1,2,3 Monique Heijmans,4 Pablo Alonso-­Coello,5,6 Matthijs Michaël Versteegh,7 Dimitri Mavridis,8,9 O Groene,10 Kaisa Immonen,11 Cordula Wagner,4 Carlos Canelo-­Aybar,5,6 Rosa Sunol,1,2,3 for the COMPAR-­EU consortium To cite: Ballester M, ABSTRACT Orrego C, Heijmans M, et al. Strengths and limitations of this study Introduction Population ageing and increasing chronic Comparing the effectiveness illness burden have sparked interest in innovative care and cost-­effectiveness ►► The project will result in the largest network meta-­ models. While self-­management interventions (SMIs) are of self-­management analysis (NMA) of complex self-­management inter- drawing increasing attention, evidence of their efficacy interventions in four high-­ ventions (SMIs). priority chronic conditions is mostly based on pairwise meta-­analysis, generally ►► SMIs are inconsistently defined across the literature in Europe (COMPAR-­EU): a derived from randomised controlled trials comparing potentially generating a high level of heterogeneity research protocol. BMJ Open interventions versus a control or no intervention. As such, for the NMA, which we will mitigate by developing a 2020;10:e034680. doi:10.1136/ relevant efficacy data for comparisons among different validated taxonomy. bmjopen-2019-034680 SMIs that can be applied to specific chronic conditions are ►► The development of Core Outcome Sets with in- ►► Prepublication history for missing. Therefore, the relevance of the available evidence put from patients and other stakeholders for each this paper is available online. for decision-­making at clinical, organisational and policy chronic condition will ensure that outcomes as- To view these files, please visit levels is limited. sessed in the NMA are relevant to the target users. the journal online (http://​dx.​doi.​ Aim To identify, compare and rank the most effective ►► The comparative effectiveness analysis via NMA, org/​10.​1136/​bmjopen-​2019-​ and cost-­effective SMIs for adults with four high-­priority cost-­effectiveness and contextual factors evaluation 034680). chronic conditions: type 2 diabetes, obesity, chronic will provide new knowledge that should facilitate fu- obstructive pulmonary disease,and heart failure. ture implementation of successful SMIs. Received 01 October 2019 Revised 10 December 2019 Methods and analysis All activities will be conducted ►► An interactive platform will facilitate access to Accepted 19 December 2019 as part of the cost-­effectiveness of self-­management decision-­making tools relevant to the specific needs interventions in four high-­priority chronic conditions in of the different target users. Europe(COMPAR-­EU, Comparing effectiveness of self-­ management interventions in 4 high priority chronic diseases inEurope) Project, an European Union (EU)-­funded project designed to bridge the gap between current managers, policymakers and industry), and will be knowledge and practice on SMIs. In the first phase of the disseminated in a tailored multi-­pronged approach that project, we will develop and validate a taxonomy, and a will include deployment of an interactive platform. Core Outcome Set for each condition. These activities will inform a series of systematic review and network meta-­ analysis about the effectiveness of SMIs. We will also Introduction perform a cost-­effectiveness analysis of the most effective As population ageing accelerates worldwide, SMIs and an evaluation of contextual factors. We will finally chronic illness will place an increasing burden develop tailored decision-­making tools for the different on society and healthcare systems.1 Chronic © Author(s) (or their relevant stakeholders. employer(s)) 2020. Re-­use conditions affect over 80% of people aged Ethics and dissemination Ethical approval was obtained permitted under CC BY. from the local ethics committee (University Institute for over 65 in Europe and account for an esti- Published by BMJ. mated 77% of disease burden, as measured Primary Care Research - IDIAP Jordi Gol). All patients For numbered affiliations see and other stakeholders will provide informed consent by disability-­adjusted life years.2 Furthermore, end of article. between 70% and 80% of healthcare costs in prior to participation. This project has been funded by the Correspondence to EU Horizon 2020 research and innovation programme Europe can be attributed to chronic disease, Marta Ballester; (grant agreement no. 754936). Results will be of interest and the current €700 billion expenditure is ​mballester@​fadq.​org to relevant stakeholder groups (patients, professionals, expected to rise.3 Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 1 Open access Self-­management support has become a key strategy for meta-­analysis, however, leaves a crucial gap, because it addressing chronic disease burden,4 contributing to the requires the RCT’s that are pooled to have included the paradigm shift from a paternalistic model where patients same interventions. To provide decision-­ makers, clini- are viewed as passive recipients of care, towards more cians and patients with solid evidence on how effective equitable and collaborative models of clinician–patient an SMI is for a given outcome and disease, multiple inter- interaction.5 ventions need to be compared. In COMPAR-­EU, we plan Cost-­ effectiveness of self-­management interventions to do this using NMA to assess the relative effectiveness of in four high-­ priority chronic conditions in Europe SMIs in four chronic conditions. (COMPAR-­ EU) is an EU-­ funded project designed to NMA synthesises direct and indirect evidence across bridge the gap between current knowledge and prac- a network of multiple interventions. This method has tice on self-­management interventions (SMIs). For the numerous advantages: it provides more precise effect purpose of this project, we define self-­ management estimates, allows for the estimation of relative effective- as ‘actions that individuals, families, and communi- ness between interventions that have not been compared ties engage in to promote, maintain, or restore health directly and provides a ranking of interventions by and cope with illness and disability, with or without the effectiveness, presenting thus a potential analytical support of health professionals, and including but not advantage.18–22 NMA has been in use for some years, an limited to self-­prevention, self-­diagnosis, self-­medication, empirical evaluation of 456 NMAs published up to 2015 and coping with illness and disability’.6 showed that just 16% of these addressed complex inter- Self-­management of a chronic condition requires ventions23 such as SMIs.24 self-­efficacy, largely understood as a person’s confi- To support future reimbursement decisions based on dence in their ability to cope with their illness.7 To be estimates of the long-­term effects of SMIs, we will develop self-­efficacious, people need special skills to cope with simulation models in the four disease areas which will the consequences of the disease, including monitoring translate benefits as identified by the NMA into long-­term symptoms and clinical markers, understanding the impli- health benefits expressed in quality-­ adjusted life years cations of these, and adjusting behaviours treatment (QALYs), which enables us to compare not only SMIs accordingly. within but also across the four conditions. The simula- SMIs are supportive interventions systematically deliv- tion models will also include a cost-­effectiveness compo- ered or led by healthcare staff or other patients with nent. Cost-­effectiveness analyses are important as they the aim of building patients’ confidence and equip- help to prioritise health expenditure. There is evidence ping them with the necessary skills. Their purpose is to that certain SMIs are cost-­ effective. Weight reduction actively engage patients (and informal caregivers where in obesity, for example, can produce short-­term savings appropriate) in the management of their disease.8 As and increase the chances of remaining in employment,25 such, they are more than merely didactic, instructional resulting in additional societal gains, while COPD manage- programmes, as their primary objective is to bring about ment interventions can improve quality of life at generally changes in behaviour and trigger a sequence of positive acceptable societal costs,26 and in some cases even result knock-­on effects.9 in short-­term healthcare cost savings.27 Secondary cardio- SMIs are complex interventions,9 typically character- vascular risk prevention programmes have the potential ised by multiple factors (components, formats, settings, to reduce direct healthcare expenditure and improve target behaviours, etc) that interact over time as partic- health outcomes.28 While these results have not yet been ipants move back and forth between the intervention structured into a consolidated body of knowledge, they processes and everyday life, which can entail challenges in do indicate that investing in SMIs may be cost-­effective. measuring effectiveness. Despite this, there is promising A better understanding of facilitators and barriers to evidence that SMIs, under given conditions, can improve successful programme implementation is also essential. clinical outcomes in numerous chronic conditions, such Contextual factors at various levels (patient, patient– as diabetes (reduction of haemoglobin A1c (HbA1C) provider interaction, organisation and system) can levels),10 obesity (reduction of weight loss),11 12 chronic all influence SMI uptake, engagement and success. obstructive pulmonary disease (COPD) (improvement of Improving our understanding of how these factors influ- dyspnoea)13 and heart failure (reduction of mortality).14 ence the effectiveness and cost-­effectiveness of complex SMIs have also been associated with improvements in SMIs for chronic illness is much needed. patient-­reported outcomes, such as quality of life, and more specific disease measures,15 such as self-­efficacy16 Aim and adherence.17 COMPAR-­EU is a multimethod interdisciplinary project Evidence on the efficacy of SMIs to date has mostly come that will run from 2018 to 2022. The project has been from systematic reviews that have employed pairwise meta-­ designed to help bridge the gap between current knowl- analysis of randomised controlled trials (RCTs). System- edge and practice in SMIs for chronic illness. Its aim is atic reviews pool evidence from RCTs comparing the to identify, compare and rank the most effective SMIs for same interventions and have long been considered as the adults living with four high-­priority chronic conditions highest standard in evidence-­based healthcare. Pairwise (type 2 diabetes, obesity, COPD and heart failure), and 2 Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 Open access among these interventions, categorise the most cost-­ decision-­making tools to facilitate access to and use of the effective and feasible SMIs. The results of the project most effective SMIs among key target end users. will facilitate informed decision-­ making and support the implementation of best practices in different health- care contexts through an interactive platform featuring Methods and analysis decision-­making tools, and other end products for poli- The COMPAR-­EU project is divided into seven phases cymakers, guideline developers, researchers, healthcare following to our specific objectives. Each phase is further professionals, patients and industry. described and shown in figure 1 (COMPAR-­EU Phases The specific objectives are to (1) validate a taxonomy and main tasks). of SMIs; (2) identify and prioritise SMI outcomes from the perspective of both patients and practitioners, culmi- Phase 1: refinement and validation of a taxonomy of SMIs nating in a Core Outcome Set (COS) for each condition; Taxonomies are formal systems for classifying multifac- (3) carrying out systematic reviews to synthesise existing eted, complex phenomena according to a set of common evidence on SMIs from RCTs; (4) compare the relative conceptual domains and dimensions; their use increases effectiveness of SMIs through NMA; (5) model the cost-­ clarity in defining and comparing complex phenomena.29 effectiveness impact of SMIs; (6) analyse contextual Several taxonomies for SMIs have been developed in the and implementation factors and (7) develop and pilot literature, but so far have focused in a specific area of the Figure 1 COMPAR-­EU Phases and main tasks, presents a visual summary of the main phases of the project as described in this protocol. COS, Core Outcome Set; NMA, network meta-­analysis; RCTs, randomised controlled trials; SMI, self-­management intervention. Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 3 Open access intervention (eg, self-­management behaviour) and have the synthesis of results from the literature review. We will not been validated. In the first phase of COMPAR-­EU, we establish criteria to address potential discrepancies across will produce a conceptual map and classification system stakeholders. that will be evaluated by self-­management experts and stakeholders (including patients) through a modified Phase 3: systematic review—descriptive synthesis of the Delphi technique consisting of a two-­round online survey. evidence The list of candidate participants will include authors on Preparations for data collection self-­ management or related topics taxonomies, profes- We will develop a protocol following Cochrane guid- sional experts in self-­ management and patient repre- ance33 34 and hold training sessions for those respon- sentatives. The resulting feedback and suggestions for sible for collecting data. Before the extraction process, refinement will be integrated into a new version that will all reviewers will be trained and undergo calibration be tested by the research team when classifying the SMIs to ensure inter-­ rater agreement. Additionally, in the reported in the RCTs that will be included in the NMA. extraction process, all data collected will be reviewed by an independent researcher to ensure quality. Phase 2: development of core outcome sets for each condition Interventions can only be compared across studies when Literature search and screening they share at least some common outcomes, and appro- To identify relevant RCTs, we will draw on the databases priate selection of outcomes is essential if research is to of previous European project (PRO-­STEP) that identi- guide decision-­making and inform policy. We will there- fied hundreds of systematic reviews on SMIs for diabetes, fore develop a COS for each of the four chronic condi- obesity, COPD and heart failure. We will use these RCTs tions included in the project. published from 2000 up to 2015 (last date of SR publi- cation included in those projects) in our project, and Outcome catalogue update this data set through new searches in MEDLINE, The first step is to create an exhaustive database listing CINAHL, Embase, Cochrane and PsycINFO. We will outcomes reported in previous EU projects (PRO-­STEP30 include RCTs that compared SMIs in adults with at least and EMPATHIE31) and COMET, a COS database and one of our conditions of interest (type 2 diabetes, obesity, other relevant organisations (using a snowballing tech- COPD or heart failure) and are published in English or nique). For each condition, outcomes will be classified into Spanish. The search will be focused in RCTs published different categories, such as clinical outcomes, patient-­ from 2015 to 2018 to complement the findings of the SR reported outcome measures or resource utilisation. included in PRO-­STEP, with the possibility of including previous years if the systematic reviews from PRO-­STEP Systematic review of how patients value self-management have not covered those previous years sufficiently. An outcomes total estimate of 4000 RCTs, based on the results of a To identify patient priorities for self-­management in the previous overview,30 will be included for the four priori- selected conditions, we will conduct a scoping review, tised chronic conditions. of quantitative and qualitative studies, and a series of specific overviews for each condition about what patients, Data extraction and collection and their caregivers’ value on SMIs. We will extract data including patient characteristics, disease characteristics and comorbidities, intervention Delphi survey with patient representatives characteristics (guided by the taxonomy), outcomes We plan to include patient participation in our outcome (guided by the four COSs), results, and information on selection as it has been described as a good practice to study design and risk of bias. Specific attention will be minimise the influence of power differentials between paid to subgroups of patients according to comorbidity, stakeholders.32 We will use a modified Delphi survey gender and socioeconomic variables (eg, health literacy). administered to a convenience sample of patients, and patient representatives to ensure that our research Descriptive analysis and summary of SMIs and outcomes addresses outcomes that matter to patients (and other SMIs and outcomes identified in the RCTs will be stakeholders). Four panels of patients and carers, for described and summarised to provide information on each disease separately (5–8 members) will be given the type and number of interventions, outcome results, task of prioritising relevant outcomes on a Likert scale. patient characteristics and presence of comorbidities for each of the four conditions. Consensus workshop The proposals for each of the four COSs will be presented Phase 4: systematic review—NMA and certainty of evidence to a panel of patients and other stakeholders (healthcare Network meta-analysis professionals, policymakers and researchers) in a work- We will initially develop theoretical models that make shop that aims to achieve consensus across groups on explicit the mechanisms through which the different the most important outcomes to include in each COS. SMIs operate on a given outcome. Based on our taxonomy Participants will be previously provided with the results of of interventions, these models will identify hierar- the patient prioritisation process, and with the results of chies of elements (type of support provided, expected 4 Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 Open access self-­ management behaviour, mode of delivery of the gains and changes in caregiver burden and costs in life-­ intervention, provider, etc) that operate on the outcome years gained).37 By expressing health benefits in QALYs, to identify which components or which combinations of we will be able to produce a ranking of the most cost-­ components are most effective. effective SMIs, within and across conditions. The base Additionally to the standard NMA models, we will cases for the models will have a societal perspective, a employ component NMA35–37 to identify key intervention life-­time time horizon, apply differential discounting at components and create a ranking of SMIs according to 4% for costs and 1.5% for health (with sensitivity anal- their effectiveness. During this process, it is crucial that ysis for equal discounting at 3%). The incremental cost-­ major assumptions of NMA like transitivity (that there effectiveness ratios will be evaluated against the so-­called must be no relevant discrepancy or inconsistency between v-­threshold that denotes willingness to pay (WTP) for a direct and indirect evidence38) are satisfied, as the validity QALY.43 The WTP in the base case is assumed to be the of results will depend on the plausibility of the assump- median WTP of €24 226, per QALY as found in the system- tions made. We will also explore the distribution of effect atic review on WTP thresholds by Ryen and Svensson,44 modifiers (eg, comorbidities, gender and socioeconomic with sensitivity analyses using the mean WTP threshold factors) across the various comparisons. This will include from that same review of € 74,159,44 and country-­specific differentiation, if possible, between the various forms of thresholds where available. Reporting will include ‘usual care’ reported in the included studies. scenario analyses to evaluate structural uncertainty as well as univariate and probabilistic sensitivity analysis to eval- Evaluation and summary of the certainty of the evidence uate parameter uncertainty. To guide users in knowing how much confidence they can place in the summarised evidence; we will rate the Phase 6: analyse the contextual factors that promote certainty of evidence, obtained through the network implementation in real life contexts meta-­analysis, for each outcome of interest, using the For each of the most effective SMIs identified for the Grading of Recommendations, Assessment, Development four chronic conditions, we will perform a realist system- and Evaluations (GRADE) approach.39–41 Additionally, we atic review45 to identify key determinants of success (or will apply an alternative approach, the CINeMA frame- failure), such as intervention settings (eg, whether on work, to assess the confidence in the results by exploring a primary care level or hospital level care) and mecha- how information flows in the network and how much nisms (eg, engagement processes) that produce specific studies at high/unclear risk of bias affect the network outcomes. At the patient level, special attention will meta-­analysis estimates. We will explore how the assess- be paid to comorbidities, gender and socioeconomic ment differs between these two approaches. dimensions such as health literacy to better understand how these influence the implementation of the selected Phase 5: model the cost-effectiveness of effective SMIs SMIs. We will then use a modified Delphi method, with For the cost-­ effectiveness analysis, data on short-­ term experts on self-­management and/or implementation of effects on intermediate outcomes (eg, body mass index, healthcare interventions, to establish the importance of HbA1C) from the NMA will serve as input for simula- these contextual factors in the target countries (Belgium, tion models that extrapolate these effects into long-­term Germany, Greece, the Netherlands and Spain). Experts health effects expressed in QALYs. These health benefits will be asked to rate the magnitude of the influence of a will be combined with costs (in 2019 Euro’s) to estimate contextual factor of a list developed by researchers. The the incremental cost per QALY gained. The development final list of the contextual factors will be produced in a of the health economic models will follow a stepped final expert discussion. approach. First, a conceptual model of the disease will be developed, informed by a review of the literature. Phase 7: development and piloting of the COMPAR-EU The conceptual model will inform which statistical tech- information technology platform, and preparation for future niques and models to use (eg, discrete-­ event models, implementation Markov models, patient-­ level simulation models). We will develop an online platform that will integrate These decisions will be guided by good practice guide- all the information and evidence synthesised during lines from pharmacoeconomic communities of medical the different phases of the project. The aim is that the decision-­making and the International Society for Phar- resources included facilitate decision-­making for target macoeconomics and Outcomes Research42 available data end users (patients, healthcare professionals, policy- and clinical expertise available within the consortium. In makers, researchers and small and medium-­sized enter- all four disease models, the natural course of each of the prises (SMEs)). The platform will include the following four study conditions will be modelled with incorpora- GRADE-­based tools: tion of background events and disease-­specific mortality ►► Evidence profiles and Interactive Summary of Find- and morbidity. Intervention costs will take into account ings tables46: these presentations will provide informa- societal, healthcare, and patient perspectives, and where tion in different formats about the quality of evidence, possible, special attention will be given to societal costs and magnitude of relative and absolute effects for that have historically been understudied (eg, productivity each of the core outcomes identified. Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 5 Open access Figure 2 COMPAR-­EU platform, decision-­making tools and other end products, presents a visual summary of the projects’ foreseen main products and how they relate to the key stakeholders. COS, Core Outcome Set; NMA, network meta-­analysis; SMI, self-­management intervention. ►► Evidence to Decision (EtD) frameworks: using semi- the five participating countries. Piloting will be organised automatic templates, interactive EtD frameworks47 will in focus groups and user experience tests in simulated be completed for a number of priority questions that settings using techniques such as thinking-­out-­loud. Pilots will take into account the magnitude of desirable and with patient associations and clinicians will focus on the undesirable effects, stakeholder views on the impor- actual use of the decision aids, where pilots with industry tance of different outcomes, information on resource and policymakers will focus on the potential reimburse- use and cost-­effectiveness, impact on equity, and other ment and integration into existing IT tools. Results of the aspects like acceptability or feasibility of the interven- piloting will be used to further improve the tools. tions. The frameworks will include draft recommen- dations that could be then applied or adapted to Preparation of future implementation different settings. Even with well-­developed and user-­friendly decision tools, ►► Patient decision aids: will be developed in plain-­ integration in existing policy and regulatory frameworks language for all selected situations identified in is crucial for successful uptake of our recommendations. the previous phases of the study. The aids will be Using expert networks, we will assess the grey literature, produced in six languages (English, French, German, current guidelines and legislation to identify and analyse Spanish, Dutch and Greek) and will include informa- relevant policy and regulatory frameworks and stan- tion obtained in some of the phases included in the dards on patient participation, health technology assess- project. ment agencies, eHealth/mHealth and other topics that might influence the uptake of the self-­management and Development of COMPAR-EU online platform decision-­making tools designed within COMPAR-­ EU. The COMPAR-­EU online platform will feature structured The purpose is to ensure that current legislative, regu- interfaces tailored to the needs of different end users. It latory and reimbursement decisions at EU level are will be designed so that after answering a few simple ques- appropriately considered in exploiting the results of the tions, users will be guided to the tool or product that best COMPAR-­EU project. We will also hold workshops with suits their needs (figure 2 illustrates the main products to industry, pharma and SMEs to identify business oppor- be integrated into the platform and expected end-­users). tunities resulting from the research findings that could result in long-­term sustainable dissemination. Piloting and refinement of decision-making tools with different stakeholders The COMPAR-­EU platform will be piloted to gauge poten- tial barriers to its implementation and understand how to Patient and public involvement position the decision-­making tools in the target health- Patients are a key component of the COMPAR-­EU project care markets. Decision aids will be piloted among patients from start to finish and their interests are represented by and health professionals and EtD frameworks and other the European Patient Forum (EPF), which is one of the products will be piloted with other stakeholders (eg, poli- consortium partners. Together with other stakeholders, cymakers, industry and representatives of patient associa- they will be involved in different aspects of the project tions) that may be involved in implementing the tools in including prioritisation of core outcomes and testing of 6 Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 Open access 6 interactive tools. A core group of partners, led by the EPF, CIBER de Epidemiología y Salud Pública (CIBERESP), Barcelona, Spain 7 will also be created to ensure co-­production and establish- Institute for Medical Technology Assessment, Erasmus University Rotterdam, Rotterdam, The Netherlands ment of explicit criteria to incorporate patient views into 8 Department of Primary Education, University of Ioannina, Ioannina, Greece project outcomes, products and communication inter- 9 Sorbone Paris Cité, Universite Paris Descartes Faculte de Medecine, Paris, Île-­de-­ ventions. Plain-­language material will also be developed France, France 10 to support the main end products of the project. OptiMedis AG, Hamburg, Germany 11 European Patients' Forum, Brussels, Belgium Ethics and dissemination Twitter Marta Ballester @Inst_Donabeidan The project coordinator (Avedis Donabedian Research Acknowledgements The COMPAR-­EU group for their valuable assistance in conducting this proposal. Institute) requested the overall ethical approval for the project to our local Clinical Research Ethics Committee Collaborators The COMPAR-­EU group: Ana Isabel González (Avedis Donabedian Research Institute (FAD) and, Red de investigación en servicios de salud en (CEIC) (the University Institute for Primary Care enfermedades crónicas (REDISSEC), Aretj-­Angeliki Veroniki (University of Ioannina, Research—IDIAP Jordi Gol). Ethical approval was Department of Primary Education), Claudia Valli (Iberoamerican Cochrane Centre granted on March 2018. Results are of interest to several – Biomedical Research Institute Sant Pau (IIB Sant Pau)), Claudio Alfonso Rocha stakeholder groups (patients, professionals, managers, Calderón (Iberoamerican Cochrane Centre – Biomedical Research Institute Sant Pau (IIB Sant Pau)), Ena Niño de Guzmán (Iberoamerican Cochrane Centre – Biomedical policymakers and industry) and will be disseminated Research Institute Sant Pau (IIB Sant Pau)),Estela Camus (Avedis Donabedian in a tailored multi-­pronged approach, including the Research Institute (FAD)), Giorgos Seitidis (University of Ioannina, Department of creation of an interactive platform. The data generated Primary Education), Hector Pardo-­Hernandez (Iberoamerican Cochrane Centre – by the project will be managed following the Golden Biomedical Research Institute Sant Pau (IIB Sant Pau) and CIBER de Epidemiología y Salud Pública (CIBERESP)), Jany Rademakers (Netherlands Institute for Health Open access as defined by the European Commission for Services Research, (NIVEL)), Jessica Beltrán (Iberoamerican Cochrane Centre – Horizon 2020 research projects.48 Biomedical Research Institute Sant Pau (IIB Sant Pau)), Karla Salas (Iberoamerican Cochrane Centre – Biomedical Research Institute Sant Pau (IIB Sant Pau)), Kevin Pacheco-­Barrios (Avedis Donabedian Research Institute (FAD)), Lyudmil Ninov Data sharing statement (European Patients' Forum), Maria Petropoulou (University of Ioannina, Department of Primary Education), Marieke van der Gaag (Netherlands Institute for Health COMPAR-­EU project will make all its anonymised data Services Research, (NIVEL)), Montserrat León (Iberoamerican Cochrane Centre – available on reasonable request, where required this will Biomedical Research Institute Sant Pau (IIB Sant Pau)), Nina Adrion (OptiMedis AG), be at aggregated level. The main data dictionaries and Rune Poortvliet (Netherlands Institute for Health Services Research, (NIVEL)), Stella databases generated by the project will be available on Zevgiti (University of Ioannina, Department of Primary Education), Jessica Zafra (Avedis Donabedian Research Institute (FAD)), Valentina Strammiello (European requests for uses related to research and quality improve- Patients' Forum). ment, and potentially for commercial exploitation, subject Contributors MB contributed to the design of the project and prepared the to approval by the consortium. Data availability and access manuscript of the paper. CO, MH, PA-­C, MV, DM, OG, KI, CW, CC and RS contributed is governed by the COMPAR-­EU Data Management Plan to the design of the project and reviewed and contributed to the manuscript of the which is aligned with the EU Open Data Initiative and paper. the FAIR Principles.49 Further details and information on Funding This work was supported by European Union’s Horizon 2020 research and how to access the data will be available from COMPAR-­ innovation programme under grant agreement No 754936. EU’s project website (https://​self-​management.​eu/). Competing interests None declared. Patient consent for publication Not required. Provenance and peer review Not commissioned; externally peer reviewed. Patient and public involvement Open access This is an open access article distributed in accordance with the EPF, a key umbrella organisation in patient representa- Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits tion and European level is a partner of this project and others to copy, redistribute, remix, transform and build upon this work for any as such has contributed to the project from its inception purpose, provided the original work is properly cited, a link to the licence is given, and design. and indication of whether changes were made. See: https://​creativecommons.​org/​ licenses/​by/​4.​0/. Furthermore, we have planned for patient participa- tion in key stages of the project, including the selection ORCID iD of outcomes for the Core Outcome Sets, advise towards Marta Ballester http://​orcid.​org/​0000-​0002-​7803-​8354 patient end products and participating in the design and piloting of the COMPAR-­EU platform and related References decision-­making tools. 1 Chang AY, Skirbekk VF, Tyrovolas S, et al. Measuring population ageing: an analysis of the global burden of disease study 2017. Lancet Public Health 2019;4:e159–67. Author affiliations 2 Patient empowerment-­who empowers whom? Lancet (London, 1 Avedis Donabedian Research Institute (FAD), Barcelona, Spain England) [Internet]., 2012. Available: http://www.​ncbi.​nlm.​nih.​gov/​ 2 Universitat Autònoma de Barcelona (UAB), Barcelona, Spain pubmed/​22559881 [Accessed 4 Oct 2016]. 3 3 Anderson G, Hopkins J. Responding to the growing cost and Red de investigación en servicios de salud en enfermedades crónicas (REDISSEC), Barcelona, Spain prevalence of people with multiple chronic conditions, 2007. 4 4 Bycroft JJ, Tracey J. Self-­Management support: Awin-­win solution Netherlands Institute for Health Services Research (NIVEL), Utrecht, The for the 21st century. N Z Fam Physician 2017;33:243–8. Netherlands 5 Aujoulat I, Marcolongo R, Bonadiman L, et al. Reconsidering patient 5 Iberoamerican Cochrane Centre, Biomedical Research Institute Sant Pau (IIB Sant empowerment in chronic illness: a critique of models of self-­efficacy Pau), Barcelona, Spain and bodily control. Soc Sci Med 2008;66:1228–39. Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680 7 Open access 6 European Commission. Pilot project on the promotion of self-­care in 28 Tsiachristas A, Burgers L, Rutten-­van Mölken MPMH. Cost-­ chronic diseases in the European Union [Internet]. Available: http://​ Effectiveness of disease management programs for cardiovascular ec.​europa.​eu/​dgs/​health_​food-​safety/​funding/​docs/​call_​sante_​2015-​ risk and COPD in the Netherlands. Value Health 2015;18:977–86. d2-​021-​tech_​specs_​en.​pdf [Accessed 7 Apr 2017]. 29 Bradley EH, Curry LA, Devers KJ. Qualitative data analysis for health 7 Bandura A. Self-­Efficacy: toward a unifying theory of behavioral services research: developing taxonomy, themes, and theory. Health change. Psychol Rev 1977;84:191–215. Serv Res 2007;42:1758–72. 8 Department of Health. The expert patient: a new approach to chronic 30 PRO-­STEP Consortium. Promoting self-­management for disease management for the 21th century. London, 2011. chronic diseases in the EU-­PROSTEP project, 2015: Reference: 9 Pigott T, Shepperd S. Identifying, documenting, and examining SANTE/2015/D2/021-­SI2.722481. heterogeneity in systematic reviews of complex interventions. J Clin 31 EMPATHiE Consortium. Final summary report: EMPATHiE, Epidemiol 2013;66:1244–50. empowering patients in the management of chronic diseases, 2014. 10 Umpierre D, Kramer CK, Leita CB. Physical Activity Advice Only or Available: http://​ec.​europa.​eu/​health//​sites/​health/​files/​patient_​ Structured Exercise Training and Association With HbA 1c Levels in safety/​docs/​empathie_​frep_​en.​pdf Type 2 Diabetes. JAMA 2011;305:1790–9. 32 Kirwan JR, Minnock P, Adebajo A, et al. Patient perspective: fatigue 11 Armstrong MJ, Mottershead TA, Ronksley PE, et al. Motivational as a recommended patient centered outcome measure in rheumatoid arthritis. J Rheumatol 2007;34:1174–7. interviewing to improve weight loss in overweight and/or obese 33 Higgins JP, Green S. Cochrane Handbook for systematic reviews patients: a systematic review and meta-­analysis of randomized of version 5, the Cochrane collaboration, 2008. Available: http://​ controlled trials. Obes Rev 2011;12:709–23. onlinelibrary.​wiley.​com/​doi/​10.​1002/​9780470712184.​fmatter/​ 12 Silveira JACda, Taddei JAdeAC, Guerra PH, et al. The effect of summary participation in school-­based nutrition education interventions 34 Chaimani A, Caldwell DM, Li T, et al. Additional considerations are on body mass index: a meta-­analysis of randomized controlled required when preparing a protocol for a systematic review with community trials. Prev Med 2013;56:237–43. multiple interventions. J Clin Epidemiol 2017;83:65–74. 13 McLean S, Nurmatov U, Liu JLY, et al. Telehealthcare for chronic 35 Caldwell DM, Welton NJ. Approaches for synthesising complex obstructive pulmonary disease. Cochrane Libr 2011;163. mental health interventions in meta-­analysis. Evid Based Ment Health 14 Inglis SC, Clark RA, Dierckx R, et al. Structured telephone support or 2016;19:16–21. non-­invasive telemonitoring for patients with heart failure. Cochrane 36 Freeman SC, Scott NW, Powell R, et al. Component network meta-­ Database Syst Rev 2015;116. analysis identifies the most effective components of psychological 15 Steinsbekk A, Rygg L, Lisulo M, et al. Group based diabetes self-­ preparation for adults undergoing surgery under general anesthesia. management education compared to routine treatment for people J Clin Epidemiol 2018;98:105–16. with type 2 diabetes mellitus. A systematic review with meta-­ 37 Welton NJ, Caldwell DM, Adamopoulos E, et al. Mixed treatment analysis. BMC Health Serv Res 2012;12:213. comparison meta-­analysis of complex interventions: psychological 16 de Jongh T, Gurol-­Urganci I, Vodopivec-­Jamsek V, et al. Mobile interventions in coronary heart disease. Am J Epidemiol phone messaging for facilitating self-­management of long-­term 2009;169:1158–65. illnesses. Cochrane Database Syst Rev 2012;12:CD007459. 38 Tonin FS, Rotta I, Mendes AM, et al. Network meta-­analysis: a 17 Tricco AC, Ivers NM, Grimshaw JM, et al. Effectiveness of quality technique to gather evidence from direct and indirect comparisons. improvement strategies on the management of diabetes: a Pharm Pract 2017;15:943. systematic review and meta-­analysis. Lancet 2012;379:2252–61. 39 Yepes-­Nuñez JJ, Li S-­A, Guyatt G, et al. Development of the 18 Caldwell DM, Ades AE, Higgins JPT. Simultaneous comparison of summary of findings table for network meta-­analysis. J Clin multiple treatments: combining direct and indirect evidence. BMJ Epidemiol 2019;115:1–13. 2005;331:897–900. 40 Brignardello-­Petersen R, Bonner A, Alexander PE, et al. Advances 19 Salanti G. Indirect and mixed-­treatment comparison, network, or in the GRADE approach to rate the certainty in estimates from a multiple-­treatments meta-­analysis: many names, many benefits, network meta-­analysis. J Clin Epidemiol 2018;93:36–44. many concerns for the next generation evidence synthesis tool. Res 41 Guyatt GH, Oxman AD, Schünemann HJ, et al. Grade guidelines: a Synth Methods 2012;3:80–97. new series of articles in the Journal of clinical epidemiology. J Clin 20 Mavridis D, Giannatsi M, Cipriani A, et al. A primer on network meta-­ Epidemiol 2011;64:380–2. analysis with emphasis on mental health. Evid Based Ment Health 42 Cummins BGM. The International Society for pharmacoeconomics 2015;18:40–6. and outcomes research. Value Heal 2013;16:i. 21 Efthimiou O, Debray TPA, van Valkenhoef G, et al. GetReal in network 43 Brouwer W, van Baal P, van Exel J, et al. When is it too expensive? meta-­analysis: a review of the methodology. Res Synth Methods cost-­effectiveness thresholds and health care decision-­making. Eur J 2016;7:236–63. Health Econ 2019;20:175–80. 22 Dias S, Sutton AJ, Welton NJ, et al. Evidence synthesis for decision 44 Ryen L, Svensson M. The willingness to pay for a quality adjusted life year: a review of the empirical literature. Health Econ making 3: heterogeneity--subgroups, meta-­regression, bias, and 2015;24:1289–301. bias-­adjustment. Med Decis Making 2013;33:618–40. 45 Pawson R, Greenhalgh T, Harvey G, et al. Realist review - a 23 Petropoulou M, Nikolakopoulou A, Veroniki A-­A, et al. Bibliographic new method of systematic review designed for complex policy study showed improving statistical methodology of network meta-­ interventions. J Heal Serv Res Policy 2005;10:21–34. analyses published between 1999 and 2015. J Clin Epidemiol 46 Guyatt G, Oxman AD, Akl EA, et al. Grade guidelines: 1. Introduction-­ 2017;82:20–8. GRADE evidence profiles and summary of findings tables. J Clin 24 Petticrew M, Rehfuess E, Noyes J, et al. Synthesizing evidence Epidemiol 2011;64:383–94. on complex interventions: how meta-­analytical, qualitative, and 47 Alonso-­Coello P, Schünemann HJ, Moberg J, et al. Grade evidence mixed-­method approaches can contribute. J Clin Epidemiol to decision (ETD) frameworks: a systematic and transparent 2013;66:1230–43. approach to making well informed healthcare choices. 1: 25 Obesity RAR, Loss W. Weight loss, and employment prospects: introduction. BMJ 2016;353:i2016. evidence from a randomized trial. J Hum Resour 2015;50:759–810. 48 Guidelines to the rules on open access to scientific publications 26 Hoogendoorn M, van Wetering CR, Schols AM, et al. Is and open access to research data in horizon 2020, 2017. Available: interdisciplinary community-­based COPD management (INTERCOM) https://​ec.​europa.​eu/​research/​participants/​data/​ref/​h2020/​grants_​ cost-­effective? Eur Respir J 2010;35:79-87. manual/​hi/​oa_​pilot/​h2020-​hi-​oa-​pilot-​guide_​en.​pdf 27 Boland MRS, Tsiachristas A, Kruis AL, et al. The health economic 49 Wilkinson MD, Dumontier M, IjJ A, et al. Comment: the fair guiding impact of disease management programs for COPD: a systematic principles for scientific data management and stewardship. SCI data, literature review and meta-­analysis. BMC Pulm Med 2013;13:40. 2016. 8 Ballester M, et al. BMJ Open 2020;10:e034680. doi:10.1136/bmjopen-2019-034680