Optimising cohort data in Europe

Optimising cohort data in Europe Methodological, practical, ethical and legal considerations

Editors

Ellen Vorstenbosch

The SYNCHROS consortium Ellen Vorstenbosch Josep Maria Haro Abad Anastassja Sialm

The SYNCHROS consortium consists of 11 partners from 6 European countries

Disclaimer: This document has been produced in the context of the SYNCHROS Project. The SYNCHROS Project has received funding from the European Union’s H2020 Programme under grant agreement Nº 825884. For the avoidance of all doubts, the opinions expressed in this document reflect only the authors’ view and in noway the European Commission’s opinions. The European Commission has no liability in respect to this book and is not responsible for any use that may be made of the information it contains.

First edition, 2022 © 2022, Editorial INDE www.inde.com editorial@inde.com editorialinde.tumblr.com facebook.com/INDEEditorial

@INDEEditorial TV Inde Editorial

© 2022, authors

This product is protected by intellectual property laws. All rights reserved. No part of this publication may be reproduced, in whole or in part, stored or transmitted in any form or by any means, mechanical, electrical, chemical, optical, recording or photocopying, without the prior permission of the publisher.

Editors:

Ellen Vorstenbosch Research, Innovation & Teaching Unit Parc Sanitari Sant Joan de Déu- CIBERSAM Carrer Doctor Antoni Pujadas 42 08830 Sant Boi de Llobregat (Barcelona), Spain Josep Maria Haro Abad Research, Innovation & Teaching Unit Parc Sanitari Sant Joan de Déu- CIBERSAM Carrer Doctor Antoni Pujadas 42 08830 Sant Boi de Llobregat (Barcelona), Spain

Anastassja Sialm Disability Policy & Implementation Research Unit

Swiss Paraplegic Research Guido A. Zäch-Strasse 4 6207 Nottwil, Switzerland

The SYNCHROS consortium

Index Chapter 1: SYNCHROS (SYNergies for Cohorts in Health: integrating the Role of all Stakeholders)

1. Background

2. Objectives

3. The role of stakeholders

4. Concluding remarks

Chapter 2: Methodological considerations for optimising the exploitation of cohort data in Europe

1. Introduction

2. Cohort data harmonisation and integration

3. Main methodological challenges

4. Best practices, solutions and implications for implementation

Chapter 3: Ethical and legal considerations for optimising the exploitation of cohort data in Europe

1. Introduction

2. Domain-related challenges to cohort research

3. Best practices, solutions and implications for implementation

Chapter 4: New data collection technologies: state-of-the art and impact on cohort research

1. Introduction

2. State-of-the-art

3. Ethical and legal considerations

4. Best practices, solutions and implications for implementation

Chapter 5: Strategic agenda for optimising cohort data exploitation in Europe

1. Introduction: Resources and Capabilities for an “Institute for Multi-study Optimising Integrative Cohort Research in Health”

2. Pillar I: Standards- Standards for variables, minimal set of variables

3. Pillar II: Guidance- Guidelines of harmonised legal, ethical and organisational conditions for access

4. Pillar III: Inclusion- Participatory collaboration and knowledge exchange

5. Strategic Agenda

6. Conclusions

SYNCHROS (SYNergies for Cohorts in Health: integrating the Role of all Stakeholders) Ellen Vorstenbosch and Josep Maria Haro Abad

1. Background 2. Objectives 3. The role of stakeholders

3.1. Stakeholder involvement 3.2. Stakeholder identification 3.3. Stakeholder engagement 3.3.1. Evidence-based synthesis and priority setting 3.3.2. Evidence-informed policy-making 3. Concluding remarks References

Ellen Vorstenbosch ( ✉ ) • Josep Maria Haro Abad Research, Innovation & Teaching Unit Parc Sanitari Sant Joan de Déu- CIBERSAM Carrer Doctor Antoni Pujadas 42 08830 Sant Boi de Llobregat (Barcelona), Spain ✉ ellen.vorstenbosch@sjd.es

1. Background SYNCHROS (SYNergies for Cohorts in Health: integrating the Role of all Stakeholders) is an EU-funded Coordination and Support Action (H2020, ref. no. 825884; period January 2019- June 2022) aimed at developing a sustainable strategy for the integration and harmonisation of health cohort initiatives and networks across Europe and internationally. Through intensive collaboration with stakeholders (e.g. researchers, funding bodies, coordinators of cohort harmonisation and integration initiatives, policy-makers), SYNCHROS created a strategic agenda for an enhanced international coordination and sustainable recommendations for better collaboration of cohorts globally, towards the development of stratified or personalised medicine. An EU priority is the optimisation of health-related data to support personalised medicine (1) and efficient knowledge transfer from research to clinical practice. Precision or personalised medicine demands that, ideally, data on a multitude of individual behavioural and lifestyle determinants are combined with genetic and environmental information in order to refine predictions and tailor interventions to individual patients. This requires large amounts of evidentiary resources and sophisticated analytical methods, some of which have only recently become available. Population and patient cohort studies, including those from clinical trials, are particularly valuable sources of data because they can detect and quantify changes in health-related parameters. Identifying patterns in the specific disease outcomes and their determinants across these diverse population groups offers a robust evidence-based background for improving medical science and health care practices. Data sharing and optimisation initiatives have taken place at various levels in Europe. Several patient and population consortiums have joined their efforts to create an unparalleled information resource, which can reveal insights into disease processes and provide pointers for interventions (e.g. LifeCycle, CLOSER, ATHLOS, NETCAL and WW Finger). This integration of data from studies involving similar cohort populations is valuable as it increases the longitudinal power and range of health-related parameters from heterogeneous data sources. More complex is integrating data “horizontally“ across diverse cohort studies, either across different population cohorts, across patient cohorts for different diseases or across general population and patient cohorts. However, exploring commonalities across what are seemingly very different populations would offer a unique opportunity to fully exploit the information contained in these large datasets. In the case of combining data from relatively homogenous patient populations (e.g. participants in clinical trial data) or from heterogeneous populations (e.g. population and patient cohorts), in both cases, there are many obstacles related to the integration

(1) http://data.consilium.europa.eu/doc/document/ST-15054-2015-INIT/en/pdf

and harmonisation of data. To start, there are practical challenges as harmonising data across studies can be complex, and time and resource consuming. There are substantial administrative challenges, such as identifying the studies, getting access to the individual participant level data and addressing ethical and legal issues. Additionally, there are the methodological challenges in integrating highly heterogeneous data from widely diverse cohort populations that have not been designed ex ante to collect identical information. The point is that Europe is extremely well-served by a rich variety of population, patient and clinical trial cohorts, but that the enormous potential benefits of this rich information still remain not fully and optimally exploited. This is the reason why SYNCHROS set itself the challenge of defining a European/international approach for the optimisation of cohort data. In collaboration with a broad spectrum of stakeholders (e.g. principal investigators of large cohort studies, coordinators of harmonisation initiatives, representatives of ethics committees, data managers, legal advisers, policy-makers or representatives of funding agencies), SYNCHROS defined a strategy towards the integration of European and international health cohorts. 2. Objectives The main objective of the SYNCHROS project was to establish a sustainable European strategy for the development of the next generation of integrated cohorts, thereby contributing to an international strategic agenda for enhanced coordination of cohorts globally, in order to address practical, ethical, legal and methodological challenges to optimising the exploitation of current and future cohort data, towards the development of stratified and personalised medicine as well as facilitating health care policies. y y To map the cohort landscape in Europe and large international initiatives. y y To identify the best methods for integrating cohort data in order to enable the harmonisation of past ( ex ante harmonisation) and future ( ex post harmonisation) data collection. y y To identify solutions for addressing practical, ethical and legal challenges in integrating data across patients, clinical trials and population cohorts. y y To take stock of emerging and new data collection technologies and types of data, including new exposures and health risks, and their potential impact on the development of future cohort studies and the need to optimise the integration of data. Some of the more specific objectives of SYNCHROS included the following:

3. The role of stakeholders 3.1. Stakeholder involvement

Stakeholder involvement was a key factor for the success of the SYNCHROS project. Stakeholders can be defined as an individual, group or organisation who may affect, be affected by or perceive itself to be affected by a decision, activity or outcome of a project (Project Management Institute, 2017). Their involvement or collaboration is sought for reasons such as helping to solve arising problems, achieving goals, increasing the availability of resources and improving the quality of products, services or outcomes (Krstic, 2014). Most opportunities associated with stakeholder involvement are related to participatory and decision-making processes (Mielke et al., 2016). Stakeholder involvement is considered an effective strategy for reducing the gap between research, practice and policy (Slunge et al., 2017; Nyanchoka et al., 2020), helping projects to remain close to the real needs and priorities within a specific field of research. Also, stakeholder involvement can lead to empowerment and joint ownership of research projects (Peters et al., 2017), motivating sponsors, promoting transparency in decision-making of the projects, enhancing support once projects have been completed and, ultimately, leading to sustainability of outcomes (Peters et al., 2017; Ray and Miller, 2017; Martinez et al., 2019). Another reason for involving stakeholders beyond the scientific community is the complexity of certain topics. Stakeholders can provide expertise based on information from their field (e.g. academia, funding agencies and regulatory bodies), which helps to understand and reduce project constraints and risks, and increase the potential success and acceptance. Stakeholder involvement is also indicated to increase relevance, legitimacy and credibility as well as effectiveness and (social) accountability of research (Mielke et al., 2016). SYNCHROS systematically addressed the factors that contribute to the gap between optimal cohort health-data exploitation and practice by understanding the context and identifying barriers to and solutions for the delivery of a sustainable, effective strategy that will maximise positive outcomes for a large number of stakeholders. To this purpose, the following actions were necessary: 1. To assess the current state-of-the-art concerning cohort data access, sharing, harmonisation and integration, thereby focussing on the methodological, practical, ethical and legal implications as well as potential issues related to the use of new data collection technologies. 2. To define the way forward regarding the implementation of potential solutions to overcome the identified obstacles, thereby presenting windows of opportunities and barriers.

3. To formulate a robust, stable and sustainable strategic agenda for optimising the benefits and exploitation of health-related cohort data, as the final aim of the project. The functional role of stakeholder involvement in SYNCHROS is related to both processes and results, making them intermediaries for the project outcomes. Nonetheless, each project phase needed different types of stakeholders, with different approaches and levels of engagement. Not all stakeholder categories required equal involvement during the entire project. At the same time, for certain groups of stakeholders, it was considered necessary to continuously engage and use in-person contact to build relationships, while taking into account a flexible approach that adapted to their practical needs (Forsythe et al., 2016). In summary, stakeholder involvement in SYNCHROS was guided by two main purposes, each related to another phase of the project: 1. Evidence-based synthesis and priority setting. 2. Evidence-informed policy-making. 3.2. Stakeholder identification In order to coordinate the stakeholder involvement in SYNCHROS, during the first phase of the project, it was important: y y To define the type of stakeholders relevant for the methodological as well as practical, ethical, and legal aspects. y y To define the type of stakeholders relevant for the implementation facets. y y To determine the inclusion and exclusion criteria (taking into account gender and geographical balance). y y To estimate the number of representatives for each subarea. y y To define the methodology for selecting/detecting stakeholders, including means of communication. Based on an initial analysis carried out during the project’s proposal phase, the following stakeholder groups were considered relevant: researchers, leads of cohort harmonisation initiatives, funding agencies, and officials of societies of longitudinal and cohort studies such as the Society for Longitudinal and Lifecourse Studies. However, it was considered that a broader plethora of stakeholders would have an interest in patient, clinical trial and population cohort studies with an emphasis on harmonisation and integration approaches, in addition to the related methodological, ethical and legal aspects. Therefore, four types of selection methods (purposive selection, snowballing, open call and systematic selection) were combined.

For purposive selection, all consortium members were invited to send a list of potential stakeholders. For the “evidence-based synthesis and priority setting” phase, eligible candidates could contribute to 1) setting priorities related to harmonisation methodology, 2) setting priorities related to practical, legal and ethical issues, and 3) principal investigators (PIs) from population-based, patients cohorts and clinical trials to discuss on both topics. Moreover, SYNCHROS’ project officer shared a list of relevant H2020 projects including the coordinators’ contact information. For the “evidence informed policy-making” phase, suggested candidates had to meet the following description: “A representative who is in the position of facilitating research data harmonisation (methodologically and/or ethically/legally) in a health context, and who can, directly or indirectly, implement/make use of the SYNCHROS recommendations.” In the snowballing approach, suggestions were made by key stakeholders and candidates who responded to invitations to participate, either confirming their participation or apologising for not being able to participate, were asked if they could suggest alternative stakeholder candidates that could have an interest in SYNCHROS or could be of interest for SYNCHROS. The open call selection method was based on stakeholder participation advertised publicly, published on the SYNCHROS website and in newsletters. In the systematic selection, the leaders of the different work packages of the project were instructed to keep track of authors that (repeatedly) published on topics that were of particular relevance for SYNCHROS. The four methods combined resulted in a dynamic database of stakeholder candidates. This database was updated during the course of the SYNCHROS project. The stakeholder categories identified and the main reasons for engagement are shown in Table 1. Table 1. Stakeholder categories and motivations for engagement

Stakeholder category

Main motivation for engagement. They…

… have experience/knowledge in cohort research management and can provide information about practical issues (during the data collection phase). … have experience/knowledge in cohort harmonisation and can identify main issues (either methodological, practical, legal or ethical).

IPs from cohort studies

Coordinators of cohort harmonisation initiatives

Stakeholder category

Main motivation for engagement. They…

… have experience/knowledge in cohort harmonisation and can identify main methodological issues and can provide information on the current-state-of-the-art, best practices and potential solutions. … can identify main ethical issues related to data sharing and harmonisation, and provide information on the current-state of-the-art, best practices and potential solutions. … can identify main legal issues related to data sharing and harmonisation, and provide information on the current-state of-the-art, best practices and potential solutions. … have high-level experience/knowledge in cohort harmonisation and can identify main issues (either methodological, practical, legal or ethical), provide information on the current-state-of-the-art, best practices and potential solutions, and can give the consortium direction about next steps. … have experience/knowledge in aspects related to data management at national or international level, and can provide information on systems, services and organisation. … may have an interest in optimising Return-On-Investment (ROI) by implementing or funding the implementation of SYNCHROS’ recommendations, and can inform about relevant ongoing initiatives and projects. … are funding the SYNCHROS project, may have an interest in optimising Return-On-Investment (ROI) by implementing or funding the implementation of SYNCHROS’ recommendations, and can inform about relevant ongoing initiatives and projects. … might have a mutual interest; can learn from SYNCHROS findings and SYNCHROS can learn from their projects, findings and, for instance, working groups on specific topics related to cohort data harmonisation such as Research Data Alliance’s Working Group on Metadata Standards Directory. … might have a mutual interest; can learn from SYNCHROS’ findings and SYNCHROS can learn from their project findings. … may have an interest in optimising ROI by implementing or funding the implementation of SYNCHROS’ recommendations.

Experts on methodological aspects related to cohort harmonisation Experts on ethical aspects related to cohort harmonisation Experts on legal aspects related to cohort harmonisation

External Advisory Boards

Research Infrastructures

Funding agencies

European Commission, e.g. EU Policy-makers

International Networks, Associations or Organisations

Related H2020 projects

European Parliament

Stakeholder category

Main motivation for engagement. They…

… might have a mutual interest; can learn from SYNCHROS findings and SYNCHROS can learn from their project findings, and might be interested in an optimal exploitation of cohort data, to increase their ROI and enhance scientific progress. … can help to advocate for optimal exploitation of cohort data, to increase its ROI and enhance scientific progress on the health context of their interest. ... can exploit SYNCHROS’ findings by developing services and technologies that could facilitate (cohort) data harmonisation and integration for a specific health topic. ... can exploit SYNCHROS’ findings by developing services and technologies that could facilitate (cohort) data harmonisation and integration. … might be interested in helping national research institutes to improve their ROI, hence, can help to implement SYNCHROS’ findings at a local/regional level. … can help to advocate for optimal exploitation of cohort data, to increase its ROI and enhance scientific progress on the health context of their interest. … might be interested in helping local research institutes to improve their ROI, hence, can help to implement SYNCHROS’ findings at a local/regional level. ...are indirectly paying for the initiative so must be kept informed.

Academia/Research institutes

Health Nongovernmental Organizations (NGOs)

Health Trade Associations

Industry

National Health Ministries

Patient organisations

Regional Government

General Public / Press

During the identification process, it was essential to find stakeholders that best met the profile of interest, in particular during consultations and dialogues, at the time of which the consortium considered: 1. To aim for diversity in the stakeholder groups (i.e. sectors, professions, etc.). 2. To ensure geographical representation (e.g. East Europe should be present at least with one expert). 3. To ensure gender balance (e.g. at least 20 % of experts should be women or men). 4. To take into account potential conflicts between stakeholders. 5. To encourage and seek stakeholders who traditionally have been excluded from the process.

Of note, all individual stakeholders of each category were approached individually with maximum flexibility, thereby adapting the contact as a format to their availability and needs. 3.3. Stakeholder engagement Engagement begins when the researcher communicates and interacts with stakeholders, and if done properly, finally results in informed decision-making concerning the selection, conduct and dissemination of research findings so that the desired outcomes are achieved (Pandi-Perumal et al., 2015). The level and format of engagement depend on the role that the researcher foresees for the particular stakeholder (or stakeholder category), including multiple “roles” as stimulator, initiator, broker/ mediator, concept refiner, legitimator, educator, context enabler, or impact extender, as well as the evidence-related activities (priorisation, generation, synthesis, integration, dissemination and application, feedback and assessment), tasks and responsibilities assigned to the main project phases (Concannon et al., 2012; Ovretveit at al., 2014; Goodman et al. 2017). The AA1000 Stakeholder Engagement Standards (2) link the envisaged role of the stakeholder with the level of engagement, describing different approaches. In the SYNCHROS project, the principle-based frameworks of the AA1000 Stakeholder Engagement Standards were adapted and combined in the two main phases of SYNCHROS: evidence-based synthesis and priority setting, and evidence informed policy-making. 3.3.1. Evidence-based synthesis and priority setting During this project phase, SYNCHROS aimed to 1) collect the existing evidence on the international and European cohort landscape, as well as international cohort harmonisation initiatives; 2) identify the most recent information on harmonisation and integration methods, and analytic approaches to maximise the value of cohort data, thereby taking into account relevant barriers and facilitators; 3) create an overview of the practical, ethical and legal challenges in the context of cohort data harmonisation, plus the identified barriers and facilitators for collecting, sharing and integrating cohort data; and 4) take stock of emerging and new data collection technologies and types of data, including new exposures and health risks, and their potential impact on the development of future cohort studies and the need to optimise the integration of data. The approaches that were considered most efficient are summarised in Table 2.

(2) https://www.accountability.org/standards/aa1000-stakeholder-engagement-standard/

Table 2. Stakeholder approaches during the phase of evidence-based synthesis and priority setting (3)

Level of engagement

Approach

Who

Activities

Products

− Report of the European and international landscape of population-based cohorts, control trials − SYNCHROS cohort repository (3) − Report on harmonisation and integration methods − Report on analytical approaches to maximise the value of cohort studies − Inventory of infrastructures − Report on practical, ethical and legal issues scoping exercise − Report on ethical guideline proposals − Report on new communication technologies used in cohort studies − Report on the prospects for new data collection technology

− Cohort and

− Academics − PIs of cohort studies − Coordinators of harmonisation initiatives − Research infrastructures

harmonisation initiatives landscape mapping − Review studies − Inventory of research infrastructures

One-way communication: Stakeholder to SYNCHROS

Monitor

One-way communication: SYNCHROS to stakeholder, there is no invitation to reply

− Peer-reviewed publications − Newsletters − Project website − Social media − SYNCHROS leaflets

− Dissemination and communication activities

− All stakeholder categories

Inform

(3) https://repository.synchros.eu/networks

Level of engagement

Approach

Who

Activities

Products

− Exploitation of project results − Further activities in research and development − Potential future collaborations

− Presentations at conferences − Meetings with stakeholders

One-way communication: SYNCHROS to stakeholder

− All stakeholder categories

Advocate

− PIs of cohort studies − Coordinators of harmonisation initiatives − Experts in methodological aspects cohort harmonisation

− Strategy brief on

Limited two-way engagement: SYNCHROS asks questions, stakeholders answer

− Stakeholder

harmonisation and integration methods

consultations

− Stakeholder survey

− Strategy brief on

Consult

practical, legal and ethical issues

− Experts in

ethical/legal aspects cohort harmonisation

− Scientific

− Strategy brief on

Limited two-way engagement: SYNCHROS asks feedback, stakeholders review Limited two-way engagement: SYNCHROS and stakeholders exchange ideas at an equal level

Advisory Board − Ethics Advisory Board − Key informants (methods, ethics/legal)

harmonisation and integration methods

− Interviews − Meetings

Review

− Strategy brief on

practical, legal and ethical issues

− Strategy brief on

harmonisation and integration methods

− Related H2020 projects

− Meetings

Exchange

− Strategy brief on

practical, legal and ethical issues

3.3.2. Evidence-informed policy-making In this phase of the project, the aim was to reach consensus about implementation. Pragmatic, moral and content reasons have been advocated for stakeholder involvement in this context (Van Tulder et al., 2004). Pragmatic because the views of heterogeneous stakeholders ensure a decision that is supported by a broad public, make stakeholders co responsible for decision-making, and prevent that decisions are rejected in a later stage;

moral because stakeholder participation is translated as a more democratic process that could strengthen the legitimacy of decisions and policies; and content because the heterogeneity of stakeholders implies confrontation of insights, interests and forms of knowledge that could incorporate elements such as equality, power and justice into the policy. The approaches for stakeholder engagement with emphasis on implementation are shown in Table 3.

Table 3. Stakeholder approaches during the phase of evidence-informed policy-making

Level of engagement

Approach

Who

Activities

Products

− SYNCHROS cohort repository − Report on

harmonisation and integration methods

− Cohort and

− Academics − PIs of cohort studies − Coordinators of harmonisation initiatives − Research infrastructures

− Report on ethical

harmonisation initiatives landscape mapping − Review studies − Inventory of research infrastructures

guideline proposals − Report on standards for improving future cohort data − Report on strategic agenda for better coordination of cohort globally − Report on the recommended sustainability model − Peer-reviewed publications − Newsletters − Project website − Social media − SYNCHROS leaflets − Exploitation of project results − Further activities in research and development − Potential future collaborations

One-way communication: Stakeholder to SYNCHROS

Monitor

One-way communication: SYNCHROS to stakeholder, there is no invitation to reply

− Dissemination and communication activities

− All stakeholder categories

Inform

− Presentations at conferences − Meetings with stakeholders

One-way communication: SYNCHROS to stakeholder

− All stakeholder categories

Advocate

Level of engagement

Approach

Who

Activities

Products

Limited two-way engagement:

− Stakeholder identification − Alignment of

setting and monitoring performance according to terms of contract

− European

− Meetings

Transact

Commission

expectations on project outcomes

− Report on stakeholder meetings − Report on stakeholder outcomes − Standards for improving future cohort data − Strategic agenda for better coordination of cohorts globally

Limited two-way engagement: discuss a specific issue

− Policy makers − Key

− Stakeholder dialogues

Negotiate

or range of issues with the objective of reaching consensus

stakeholders in the position to implement

4. Concluding remarks Over the course of the project, stakeholder engagement showed to be an essential mechanism to ensure the relevance of knowledge and outputs generated by SYNCHROS. The topics covered through the engagement activities, in the stakeholder consultations and dialogues, provided a comprehensive overview of the challenges and solutions when it comes to health-related cohort data harmonisation. Thanks to the active participation of many external stakeholders, these engagement activities provided an excellent source of knowledge, enabling a better understanding of the context for implementation and clarifying the current barriers and windows of opportunities. All engagement activities were a source of inspiration and strengthened the consortium’s aspiration for improved coordination of cohorts. The planned activities had each a tailored approach to ensure that the (level of) engagement best met the stakeholders’ needs. Based on the large number of entities involved in the different stakeholder categories, the SYNCHROS consortium is confident that all relevant stakeholder categories were offered the opportunity to become involved in the development of a sustainable European strategy for enhanced coordination and optimal exploitation of health-related cohort data.

References

Concannon TW, Meissner P, Grunbaum JA, McElwee N, Guise JM, Santa J, et al. A new taxonomy for stakeholder engagement in patient-centered outcomes research. Journal of General Internal Medicine 2010;27(8):985-991. https://doi. org/10.1007/s11606-012-2037-1 Forsythe LP, Ellis LE, Edmundson L, Sabharwal R, Rein A, Konopla K, Frank L. Patient and stakeholder engagement in the PCORI pilot projects: description and lessons learned. Journal of General Internal Medicine 2016;31(1):13-21. https://doi. org/10.1007/s11606-015-3450-z Goodman J, Korsunova A, Halme M. Our collaborative future: activities and roles of stakeholders in sustainability-oriented innovation. Business strategy and the environment 2017;26(6):731-753. https://doi.org/10.1002/bse.1941 Krstic, N. Stakeholder management from the business perspective. Megatrend Revija , 2014;11(2):165-182. https://doi.org/10.5937/megrev1402165k Martinez J, Wong C, Piersol CV, Bieber DC, Perry BL, Leland NE. Stakeholder engagement in research: a scoping review of current evaluation methods. Journal of Comparative Effectiveness Research 2019;8(15):1327-1341. Mielke J, Vermassen H, Ellenbeck S, Fernandez Milan B, Jaeger C. Stakeholder involvement in sustainability science - A critical view. Energy Research and Social Science 2016:17:71-81. https://doi.org/10.1016/j.erss.2016.04.001 Nyanchoka L, Tudur-Smith C, Porcher R, Hren D. (2020). Key stakeholders’ perspectives and experiences with defining, identifying and displaying gaps in health research: a qualitative study. BMJ Open 2020;10(11):1-8. https://doi.org/10.1136/ bmjopen-2020-039932 Pandi-Perumal SR, Akhter S, Zizi F, Jean-Louis G, Ramasubramanian C, Edward Freeman R, et al. Project stakeholder management in the clinical research environment: how to do it right. Frontiers in Psychiatry 2015;6:71. https://doi.org/10.3389/ fpsyt.2015.00071 Ovretveit J, Hempel S, Magnabosco JL, Mittman BS, Rubenstein LV, Ganz DA. Guidance for research-practice partnerships (R-PPs) and collaborative research. Journal of Health, Organisation and Management 2014;28(1):115-126. https://doi. org/10.1108/JHOM-08-2013-0164 Peters DH, Bhuiya A, Ghaffar A. (2017). Engaging stakeholders in implementation research: Lessons from the future health systems research programme experience. Health Research Policy and Systems 2017;15(Suppl 2):4-6. https://doi.org/10.1186/ s12961-017-0269-6

Project Management Institute. A guide to the Project Management Body of Knowledge (PMBOK® guide), 6th edition. Project Management Institute. Ray KN,Miller E. Strengthening stakeholder-engaged researchand researchon stakeholder engagement. Journal of Comparative Effectiveness Research 2017;64(4):375 389. https://doi.org/10.2217/cer-2016-0096 Slunge D, Drakenberg O, Ekbom A, Sahlin U. Stakeholder interaction in research processes - a Guide for Researchers and Research groups . University of Gothenburg, Gothenburg, 2017. https://doi.org/10.13140/RG.2.2.28518.22080 Van Tulder R, Kaptein M, van Mil E, Schilpzand R. The Strategic Stakeholder Dialogue . Rotterdam: Business-Society Management, Erasmus Universiteit Rotterdam, Schuttelaar & Partnerts, 2004.

Methodological considerations for optimising the exploitation of cohort data in Europe Albert Sanchez-Niubo, Anastassja Sialm, Ellen Vorstenbosch, Maria Panagiotopoulou, Sergei Gorianin, Yannick Marcon, Desirée Gutierrez, Teresa Torres, Laura Rico-Uribe, Ángel Rodriguez-Laso, Daniel Morillo, José-Luis Ayuso, Juan R. González and Jivko Stoyanov

1. Introduction 2. Cohort data harmonisation and integration 2.1. Strategies 2.1.1. Prospective harmonisation

2.1.2. Ex ante retrospective harmonisation 2.1.3. Ex post retrospective harmonisation 2.2. Analytical methods 2.3. Interoperability 3. Main methodological challenges 4. Best practices, solutions and implications for implementation References

Albert Sánchez-Niubò Department of Social Psychology and Quantitative Psychology. Faculty of Psychology, University of Barcelona (UB), Barcelona, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain ✉ asanchezniubo@ub.edu Anastassja Sialm Disability Policy & Implementation Research Unit, Swiss Paraplegic Research (SPF), Notwil, Switzerland Ellen Vorstenbosch Research, Innovation & Teaching Unit, Parc Sanitari Sant Joan de Déu, Sant Boi de Llobregat, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain Maria Panagiotopoulou • Sergei Gorianin European Clinical Research Infrastructure Network (ECRIN), Paris, France Yannick Marcon Epigeny, St Ouen, France Desirée Gutiérrez-Marín Research, Innovation & Teaching Unit, Parc Sanitari Sant Joan de Déu (PSSJD), Sant Boi de Llobregat, Spain Teresa Torres-Moral Dermatology Department, Melanoma Unit, August Pi i Sunyer Biomedical Research Institute (IDIBAPS), Barcelona, Spain Laura Alejandra Rico-Uribe School of Health Sciences, Universidad Internacional de La Rioja, Logroño, Spain Ángel Rodríguez-Laso CIBERFES (Network-based Biomedical Research Consortium, area of Frailty and Healthy Ageing), Instituto de Salud Carlos III, Spanish Ministry of Science and Innovation, Spain Daniel Morillo-Cuadrado • José-Luis Ayuso-Mateos Department of Psychiatry, Universidad Autónoma de Madrid (UAM), Madrid, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain Juan R González Barcelona Institute for Global Health (ISGlobal), Barcelona, Spain Jivko Stoyanov SCI Population Biobanking and Translational Medicine Group, Swiss Paraplegic Research, Notwil, Switzerland Institute for Social and Preventive Medicine, University of Bern, Bern, Switzerland

1. Introduction SYNCHROS (SYNergies for Cohorts in Health: integrating the Role of all Stakeholders) is an EU Horizon 2020 project that aims to identify the methodological, practical, legal and ethical barriers and opportunities in cohort research. The main aim of SYNCHROS is to formulate a sustainable European strategy for the next generation of integrated cohorts. In so doing, SYNCHROS supports developments for a stratified and personalised medicine approach and facilitates health policy. In relation to the methodological domain of the SYNCHROS project, it is essential to identify the methodological problems faced by cohort researchers as well as provide solutions and recommendations for research practice. A definition of cohort study has to be agreed upon prior to describing and contextualising the key issues of data harmonisation, integration and optimisation, including the main methodological challenges, best practices, and implications for implementation. A cohort study refers to a specific longitudinal type of study design. It usually involves a group of people who share a common characteristic, such as living in a specific location or experiencing a defining event in a selected period (such as birth), and performs cross sectional repeated assessments of the studied outcomes at pre-defined intervals over time. A cohort study can be either retrospective or prospective. In the retrospective case, the study relates to data collected in the past (e.g. medical records). In the prospective case, the cohort study relies on the collection of new data. SYNCHROS created a repository (https://repository.synchros.eu) to share key information on initiatives that harmonise and/or integrate cohort data. The SYNCHROS repository initiatives include information about harmonisation strategies and the different types of analytical approaches to produce integrated results between different cohorts; details about these strategies and approaches are presented in this chapter. 2. Cohort data harmonisation and integration First of all, it is necessary to clarify the differences between the concepts of harmonising and integrating data from different cohort studies. Harmonisation can be understood as practices that improve the comparability of variables from different studies and thus reduce the heterogeneity across studies (Maelstrom Research). Integration, on the other hand, is understood as the act or process of combining or pooling similar data from different studies into a unified whole. Therefore, harmonisation of data from different cohorts is a first necessary step, which allows these data to be integrated into the same dataset, and thus increase their value and usefulness for specific research purposes (Lesko et al., 2018).

The reported benefits of integrating various types, levels and sources of data in such a way that they can be made compatible and comparable, and thus increasing their value and usefulness for specific research purposes, include: y y Increased sample size and, consequently, improved statistical power. y y Improved generalisability of results. y y Increased ability to investigate effect heterogeneity due to more diversity among participants. y y Answering novel research questions that cannot be answered by individual cohorts. y y Helping to ensure the validity of comparative research and reproducibility. y y Encouraging more efficient use of existing data, time and resources. Because of all these benefits, more and more international research initiatives have incorporated data harmonisation into their design to investigate and/or identify risk or protector factors for health outcomes (Harris et al., 2012; Magnin et al., 2020). Some of these initiatives aim to highlight the value of health data through large-scale analysis and to ensure the development of harmonised measures and standardised information infrastructures. The ability to effectively harmonise data from different cohort studies facilitates the rapid extraction of new scientific knowledge on the emergence and progression of diseases and conditions and the classification of their respective phenotypes. This approach has the ambition to address the unmet needs raised by many pathological conditions, including the identification and the following validation of new biomarkers, phenotypic and genotypic stratification of patients to different treatments, as well as screening and selection of novel therapy targets (Zhang et al., 2016; Kouroru et al., 2018). 2.1. Strategies Before focusing on particular and case-specific harmonisation methods, one has to consider the different available harmonisation strategies, which can be derived from the literature. The use of one strategy or another will depend on the stage and condition of the available cohort data rather than on an arbitrary researchers' choice. Three types of harmonisation strategies can be differentiated (Granda et al., 2016): prospective harmonisation, ex ante retrospective harmonisation and ex post retrospective harmonisation. 2.1.1. Prospective harmonisation In this strategy, studies share the same study design, questionnaires and instruments for collecting measures (e.g. biological, psychological and/or social). Some adaptations may occur for individual data collection sites, but the goal is to maintain comparability y y Providing opportunities for collaborative and multicentre research. y y Bringing together expert knowledge across disciplinary boundaries.

from the beginning of the process. This is typically used in multicentre studies where researchers from different centres come together to decide upon the same study design, questionnaires and/or instruments for collecting measures. The advantages are clear to see, as the process of harmonisation is established in the beginning and the integration of the data into one combined dataset can be easily achieved. Nevertheless, in prospective harmonisation difficulties may be encountered in the management of the different participating centres, and there may still be differences between study designs and data collection for cultural reasons, or unequal diligence on the part of the research teams. Involvement and support by funding agencies is likely to be essential to the success of the prospective harmonisation approach (Chandler et al., 2015). Funding agencies may be best suited to initiate and coordinate data harmonisation initiatives because of their comprehensive knowledge of research in particular areas, potential to leverage additional resources, ability to encourage collaboration among researchers, and unique perspective on goals that extend beyond individual research projects. 2.1.2. Ex ante retrospective harmonisation This strategy combines data from cohort studies that were not specifically designed to be comparable, but through the use of standard collection tools and operating procedures, their data can be integrated. Different standards are available for collecting and storing data, such as the Clinical Data Interchange Standard Consortium (CDISC) (https://www.cdisc.org) standards for clinical trials, the Global Alliance for Genomics and Health (GA4GH) (https://www.ga4gh.org) standards for genomic data, the Observational Medical Outcomes Partnership (OMOP) standards established within the Observational Health Data Sciences and Informatics (OHDSI) (https://www.ohdsi.org) initiative for observational studies etc. Nevertheless, in practice, the standards in cohort research are rather set by specific cohort studies that are recognised for their innovation and impact in a certain research context. Such studies provide an example for relating future studies in terms of the surveys and measuring instruments to be used (e.g. the Health and Retirement Survey, https://g2aging.org/). It should be noted that the rapid development of new technologies (e.g. array-based gene expression surpassed by RNA seq) (Wijmenga and Zhernakova, 2018) can make standardisation difficult even within cohorts, and under such circumstances, tailor-made methods should be developed to harmonise data produced by different technologies (Rudy and Valafar, 2011; Borisov et al., 2017). As far as possible, it may be useful to create transformation standards or links between different standards on the same measured variables in order to have available equivalent scores for various scales that measure the same health construct. 2.1.3. Ex post retrospective harmonisation Ex post retrospective harmonisation also combines data from cohort studies that were not specifically designed to be comparable but, even though no standard formats or protocols have been used, variables can be assessed and edited to achieve commonality

through data processing procedures. The main technical challenges found in this harmonisation strategy are the data quality issues that may be found across different studies, such as data incompleteness, potential misclassification of data, or ambiguous terminologies (Fortier et al., 2017). The efforts of this strategy are focused on being able to standardise the data of each cohort study retrospectively and to create bridges to harmonisation of the variables with minimum possible bias. The main challenges to the ex post retrospective approach are (Dubrow and Tomescu-Dubrow, 2016; Fortier et al., 2017): y y To find a balance between precision and number of harmonised variables across studies, in order to ensure correct inferences. y y To find ways to measure the quality of harmonisation. y y To define the kind of documentation of the involved studies required to ensure maximum efficiency in the harmonisation process. y y To disseminate the harmonisation process so that it is reproducible. y y To overcome a minimum investment in resources in terms of time and money, which is usually sought. 2.2. Analytical methods Harmonisation reduces the heterogeneity of comparable data across studies. We refer to reducing because heterogeneities can hardly be eliminated even in prospective harmonisation strategies. Therefore, harmonised data must be analysed to obtain integrated results, considering possible heterogeneities to correctly interpret differences in assessing an exposure or intervention effect of interest (Burke et al., 2017). Common heterogeneity sources, amongst others, are: y y Measurement characteristics such as levels of precision of each study. y y Different covariates collected. y y Differences between population samples. Therefore, proper modelling of the data should control for the specificities of the studies. Studies can be combined at the level of individual participant data (one-stage or pooled analysis) or at the level of their results or inferences (two-stage or meta-analysis). To conduct a pooled analysis, data from the different studies must necessarily meet two main requirements: the data must be available to the same researchers across the different studies, and they can be properly harmonised. The requirement that data can be harmonised depends very much on the homogeneity of the same data across studies, with the main challenge of inferences being as correct as possible. To do this, a y y Unmeasured confounders that vary by centre. y y Stage of the trials or conditions of the cohorts.

balance must be found between accuracy and validity among the different variables in each study in order to construct the new common variable. Most prominently, pooled analyses adjust for the centre using either fixed or random effects on the outcome. Fixed effect models are preferred when having large sample sizes per study and the different cohorts provide samples regarding the same target population. On the other hand, random effect models are preferred when there are some studies with a small number of subjects and when there is much difference between the studied populations. The two‐stage approach first analyses the individual participant data at each centre separately and then combines the results across studies. For this second stage, this approach is also called “aggregated data meta-analysis”. The aim of the first stage is to produce estimations of the statistics of interest with its variance in each study. In the second stage, the coefficient estimates obtained in the first stage are combined. Again, they can be considered either fixed or random across studies. When the effect of interest is considered fixed, the estimates from the previous analyses are considered as observations of normal variables, and the most common method to estimate the single effect across studies is to consider the average of the individual effects weighted by the inverse of their variances. Alternatively, when the effect of interest is considered random, the estimate of the effect across centres is again a weighted average of these individual effects but, in this case, the weights are usually given by the inverse of the sum between variances at each centre and the between study-variance, a measure of the between-study heterogeneity. The main advantage of the two-stage meta-analysis is that it accounts for between-study heterogeneity. The one-staged pooled analysis can also account for it when necessary, but its main advantage is its higher statistical power, which derives from the reduction of parameters to estimate and the potential to better modelling the combined data. Other important differences include: y y The two-stage approach assumes that the estimates distribute normally with known variances while the one-stage approach actually models the individual participant data and does not rely on this assumption. y y The two-stage approach is problematic for studies on rare outcomes, with low number of participants or unbalanced subgroups. Different weighting averages can be used to correct for these issues. y y The two-stage approach accounts for the clustering of patients by trial. The one stage approach needs to explicitly account for the clustering by stratifying the analysis by trial or assuming a random-effect for the intercept across trials. y y The one-stage approach may overlook that the effect of adjustment for covariates could be different between studies. This is automatically done in the two-stage approach but needs to be explicitly defined in one-stage analyses.