Open access Protocol Assessing the outcomes and experiences of care from the perspective of people living with chronic conditions, to support countries in developing people-­ centred policies and practices: study protocol of the International Survey of People Living with Chronic Conditions (PaRIS survey) Dolf de Boer,1 Michael van den Berg,2 Marta Ballester,3,4,5 Janika Bloemeke,6 Wienke Boerma,1 Katherine de Bienassis,2 Peter Groenewegen,1,7 Oliver Groene,6 Candan Kendir,2 Niek Klazinga,2 Ian Porter,8 Rosa Sunol,3,4,5 Laura Thomas,9 Jose Maria Valderas,8,10 Rachel Williams,9 Mieke Rijken ‍ ‍1,11 To cite: de Boer D, van den ABSTRACT Berg M, Ballester M, et al. Introduction In view of growing populations with chronic STRENGTHS AND LIMITATIONS OF THIS STUDY Assessing the outcomes and conditions, many countries are redesigning their health ⇒ To maximise the relevance and usability of the experiences of care from the International Survey of People Living with Chronic systems. However, little information is available about how perspective of people living Conditions (PaRIS survey), it has been developed in with chronic conditions, to health systems perform from the perspective of people living with chronic conditions. The Organisation for Economic Co-­ a collaborative process with many stakeholders in support countries in developing people-­centred policies and operation and Development (OECD) Member States therefore participating countries and at international level. practices: study protocol of the mandated the OECD to initiate the International Survey of ⇒ Translating and revising the survey questionnaires International Survey of People People Living with Chronic Conditions (PaRIS survey), which has been a process of carefully balancing local ap- Living with Chronic Conditions aims to provide insight in outcomes and experiences of plicability and cross-­country comparability. (PaRIS survey). BMJ Open care as reported by people living with chronic conditions. ⇒ To ensure optimal protection of personal data and 2022;12:e061424. doi:10.1136/ The PaRIS-­SUR consortium has been tasked by the OECD to avoid registration bias, persons with chronic condi- bmjopen-2022-061424 tions will be identified through self-­report of chronic support the development and implementation of the survey. ► Prepublication history for Methods and analysis As primary care services play conditions among service users aged 45 or old- this paper is available online. a pivotal role in the management of chronic conditions, er (assuming that many of them will have chronic To view these files, please visit the PaRIS survey will be implemented in the primary care conditions) who were selected from primary care the journal online (http://dx.doi.​ setting. Data will be collected with a survey among users of practices and gave informed consent to participate org/10.1136/bmjopen-2022-​ primary care services aged 45 years or older, of whom many in the survey. 061424). ⇒ Because we survey persons who have been in con- have chronic conditions. An additional survey is conducted tact with a primary care provider, real care expe- Received 31 January 2022 among their primary care providers. The nested study design riences are reported; this method is however less Accepted 31 August 2022 will allow analysis of the patient-­reported data in relation suitable to study access issues from the patient to characteristics of and care provided by primary care perspective. providers within and across countries. In 2022, the survey ⇒ The nested study design will allow analysis of the © Author(s) (or their will be tested in a Field Trial in participating countries. Data variation in patient-­ reported data in relation to employer(s)) 2022. Re-­use for cross-­country comparison will be collected by the Main the organisation of and care processes applied in permitted under CC BY-­NC. No Survey in 2023. commercial re-­use. See rights primary care practices and characteristics of the Ethics and dissemination Informed consent will be and permissions. Published by health system. obtained from primary care providers and service users. BMJ. National Project Managers search ethical approval of the For numbered affiliations see survey in their country, if required. Reporting by the OECD end of article. be disseminated through an international OECD flagship will focus on questions for international comparison. A report, conferences, scientific papers and policy briefs, to Correspondence to secured information technology platform will be developed inform strategies to improve care for people living with Prof. Mieke Rijken; for participants and stakeholders in countries to receive chronic conditions throughout the world. ​m.​rijken@​nivel.​nl feedback and answer their own questions. Findings will also de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 1 Open access INTRODUCTION the Organisation for Economic Co-­operation and Devel- Background and rationale opment (OECD) Member States mandated the OECD to Chronic conditions are a major health threat for individ- lead an effort to develop and implement an international uals, which often also negatively impact the quality of life survey to gain more insight in outcomes and experiences and functioning of these persons and their families in of care as reported by people living with chronic condi- various life roles (eg, as a partner, caregiver, employee) tions.20 This has led to the conceptualisation and devel- as well as their social and financial position.1–6 Besides opment of the International Survey of People Living with their impact on individuals and households, chronic Chronic Conditions (PaRIS survey). conditions pose a major challenge to health systems and societies. For example, the Medical Expenditure Objectives Panel Survey in the USA demonstrated that 60% of The PaRIS survey aims to help policymakers understand non-­institutionalised adults in the USA had at least one how their health system addresses the needs of people chronic condition—42% had more than one—and that living with chronic conditions. A core element of the these conditions were responsible for more than 80% approach of PaRIS is the international standardisation of of healthcare spending in the USA in 2014.7 Although instruments and procedures for sampling and data collec- the prevalence and management of chronic conditions tion, in order to facilitate international comparisons and vary across countries, a rapid growth of populations with cross-­ country learning. The ultimate goal is that the (multiple) chronic conditions and of related healthcare results of PaRIS will help countries to make their health expenditures have been reported from countries all over systems more responsive to people’s needs. the world.8 9 Moreover, chronic conditions affect coun- Inclusive development tries’ labour force and productivity substantially, irrespec- A key principle of the PaRIS survey is that it is devel- tive of continent and income.10 oped inclusively with stakeholders to ensure relevance, To improve the quality of care for people living with meaningfulness and the uptake of its results at all levels chronic conditions as well as to improve health outcomes of countries’ health systems. In view of this, the OECD and control healthcare expenditures, countries are rede- established a Taskforce to advise on the development of signing their health systems towards becoming more the survey during 2017. This Taskforce consisted of repre- people-­ centred and integrated.11 Primary care (PC) sentatives of patients and care providers, country repre- services play a pivotal role in these reforms, as in many sentatives and experts on population and health services countries PC providers are the first and main contact surveys and patient-­ reported indicators. The resulting with the health system for people living with chronic proposal,21 adopted by the OECD Health Committee in conditions in the community. Ideally, PC providers offer 2018, formed the basis for the current project, in which directly accessible care of a comprehensive and contin- the international survey is being developed with inputs uous nature. They could serve a coordination role across from all stakeholders and implemented in participating disciplines and sectors, taking a holistic, person-­centred countries. approach.12 13 Regarding chronic care management, PC providers could play a key role in all phases of the Governance citizen/patient journey, starting from educating and Figure 1 shows the governance structure of the PaRIS supporting people to improve their lifestyle, encouraging survey. The development of the PaRIS survey and its them to engage in programmatic screening, identifying implementation in countries is overseen by the Health high risks by (referral to) diagnostic procedures, treating Committee, a formal body in which all OECD countries chronic conditions or referring persons for treatment in specialist settings and by providing self-­management support, monitoring and follow-­up care along the entire care pathway.14–17 While countries face rising needs because of ageing and more medically complex populations, many struggle to assess how their health systems perform with regard to the management of chronic conditions. Outcome measures available today focus on mortality, and incidence and prevalence of disease, based on health services’ or Figure 1 Governance structure of the PaRIS survey. Four health insurers’ administrative data sets. Little informa- types of relations among the various actors exist: (1) Formal tion is available about how people-­centred health services responsibility for appointing country representatives and National Project Managers (indicated by black arrows); are from the perspective of users and to what extent (2) Formal reporting responsibility (indicated by blue these services provide (integrated) care that meets their arrows); (3) Technical support (indicated by red arrows); needs.18 Considering this critical information gap and (4) Advisory role (indicated by green arrows). OECD, the increasing recognition that patient-­reported experi- Organisation for Economic Co-­operation and Development; ences and outcomes play a pivotal role in making health PaRIS, International Survey of People Living with Chronic systems more people-­centred,19 the Health Ministers of Conditions; WP-­PaRIS, Working Party for PaRIS. 2 de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 Open access are represented. The OECD Working Party for PaRIS relevance, the results of PC service users living with (WP-­PaRIS) resides under the Health Committee, and chronic conditions will also be compared with the results consists of formal representatives of all countries partic- of PC service users without such conditions. As such, data ipating in the PaRIS initiative. These are usually officers analysis will be guided by the following main questions21 : from the Ministries of Health or governmental research 1. What are the patient-­reported outcomes of PC service institutes, who are assisted by experts in their own country users aged 45 and over with chronic conditions, com- or at the OECD, if needed (eg, legal or data management pared with those without chronic conditions, in the experts). Participation is open to all OECD members and areas of symptoms, physical, mental and social func- invited non-­OECD members. tioning, self-­reported health and health-­related quality To ensure that the patient perspective is fully consid- of life? How do these results vary across countries? ered, the OECD established the Patient Advisory Panel 2. What are the experiences of PC service users aged 45 consisting of representatives of national and interna- and over with chronic conditions, compared with those tional patient federations.22 This panel has been and will without chronic conditions, in the areas of access, com- be consulted at all major steps of the project. Further- prehensiveness, continuity, coordination, safety and more, the PaRIS-­SUR consortium is seeking advice in all people-­centredness of care, self-­management support, phases of the project from a worldwide community of trust and overall perceived quality of care? How do independent experts, the Technical Advisory Commu- these results vary across countries? nity, covering a wide range of expertise (eg, legal, health 3. How do patient-­reported outcomes and care experi- policy, health services, survey methodology, statistics).23 ences vary for PC service users aged 45 and over with At the national level, stakeholders and experts in the chronic conditions by background characteristics such participating countries are involved in the development of as age group, gender, education level, occupational the survey instruments and the national implementation status, household composition, health-­risk behaviours, plans. To this aim, each participating country appointed level of multimorbidity, disease status and confidence a National Project Manager, who works with a national in managing one’s own care? team and the PaRIS-­SUR consortium to coordinate the 4. How do key characteristics of PC practices relate to implementation of the survey in the country. The consor- the care experiences and outcomes of PC service users tium provides support to the National Project Managers aged 45 and over with chronic conditions? and ensures that the survey is being implemented consis- 5. How do characteristics of health systems and countries tently, allowing for cross-­country comparative analyses of relate to the care experiences and outcomes of PC ser- the survey data. vice users aged 45 and over with chronic conditions? Considering their key role, the OECD provided a task profile to help countries with appointing their METHODS AND ANALYSIS National Project Manager. Requested key competencies Study design mentioned in the profile were, among others, a trusted The PaRIS survey is designed for cross-­country compar- reputation in regards to survey implementation, chronic ison. The main instrument is a survey among PC service conditions, primary care and health policy; being accus- users aged 45 years or older to collect patient-­reported tomed with national approaches towards healthcare and data. An additional survey among their PC providers is applied research; access to networks in the primary care included to collect data on the characteristics of and or other ambulatory care community; ample experience care they provide, in particular related to chronic care in planning and organising large-­scale surveys; familiarity management. The PaRIS survey has a nested design: PC with sampling, survey data collection, quality control service users are nested in PC practices, which are at their procedures and data management. It was emphasised turn nested in (national or regional) health systems in that the National Project Manager would need to build a countries (figure 2). This will allow analysis of the varia- multidisciplinary team to ensure that all expertise would tion in patient-­reported data in relation to characteristics be available to perform their tasks according to the inter- national standards adopted by the OECD. Main research questions The PaRIS survey seeks to address many questions that are relevant for people living with chronic conditions and their families, PC providers, policymakers and health authorities in countries. Reporting by the OECD will predominantly focus on the main questions for inter- national comparison in line with the overall goal of the PaRIS survey, that is, to inform countries on the patient-­ reported outcomes and care experiences of their citizens Figure 2 Multilevel design of the International Survey of living with chronic conditions compared with similar People Living with Chronic Conditions (PaRIS survey). PC, populations in other countries. To strengthen policy primary care. de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 3 Open access of and care provided by PC providers within and across Eligibility criteria participating countries. Since the PaRIS survey has a nested design, National The current project consists of a Field Trial (2022) to Project Managers should first sample PC practices in pilot the survey instruments and sampling, data collec- their country and subsequently, within the selected PC tion and analysis procedures, followed by data collec- practices, service users who meet the eligibility criteria. tion for cross-­country comparison with the Main Survey Eligible PC practices are those care practices that meet (February–October 2023). To monitor outcomes and the definition mentioned above. Eligibility criteria for PC experiences from the perspective of people living with service users are: (1) aged 45 years or older at the time chronic conditions, the PaRIS survey is expected to of sampling; (2) living in the community (in a private continue on an ongoing basis. The WP-­PaRIS will decide household, ie, not in a long-­term care facility, healthcare about its frequency in 2024, based on the results of or other residential institution); and (3) having had at the Main Survey in this project phase. Box 1 provides least one registered contact with a PC provider—either an overview of the countries that committed to its face-­to-­face, by telephone or online—for any medical or implementation. administrative reason, during the 6 months preceding the selection procedure. PC service users who cannot complete the question- Study setting naire independently are not excluded, as there is the The survey will be conducted in the primary care setting. possibility for them to be assisted by a family member or To identify this setting in each country, we defined PC friend, or from a helpdesk set up by the National Project practices as facilities or practices staffed by physicians who Manager. Absence of medically diagnosed chronic are licenced to serve the general population of a community disease(s) is neither an exclusion criterion for service and provide generalist care in an ambulatory (outpatient) users, for two main reasons. First, we want to collect setting. Generalist care is characterised by its compre- data also from people who do not have a chronic condi- hensiveness, continuity and coordination and is not tion, to compare their results with those of people living restricted to a specific category of patients. PC practices with chronic conditions. As such, the survey will provide are usually accessible without referral, and staffed with information on whether findings relate to care delivery medical doctors with a specialist training in general in general or to the management of chronic conditions practice, such as general practitioners or family physi- in particular. Second, we want to avoid using data from cians. Many PC practices are also staffed with other care patients’ health records, as this may create technical or professionals, such as nurses or allied healthcare profes- legislative barriers for participation in some countries. sionals, although single-­handed general practices also Recording of chronic conditions may also be different exist. within and across countries, which would affect compara- bility across PC practices and countries. Therefore, people with and without chronic conditions will be identified Box 1 Countries that committed to implement the based on their report of a number of chronic conditions International Survey of People Living with Chronic as proposed by the PaRIS Taskforce (the list of chronic Conditions (PaRIS survey) conditions to be included in the PaRIS survey consists of conditions proposed by the PaRIS Taskforce,21 with minor ⇒ Australia changes made during the development process of the ⇒ Belgium survey instruments: Alzheimer’s disease/other cause of ⇒ Canada dementia; arthritis/ongoing problem with back or joints; ⇒ Czech Republic breathing condition (eg, asthma, chronic obstructive ⇒ England (UK) pulmonary disease); cancer (diagnosis or treatment in ⇒ France the last 5 years); diabetes; chronic kidney disease; chronic ⇒ Greece liver disease; high blood pressure; cardiovascular/heart ⇒ Iceland condition; (ongoing) mental health condition (eg, ⇒ Israel ⇒ Italy anxiety, depression, schizophrenia); neurological condi- ⇒ Luxembourg tion (eg, epilepsy, migraine); another chronic condi- ⇒ The Netherlands tion).21 This list covers broadly defined areas of the most ⇒ Norway prevalent chronic conditions in OECD countries, and are ⇒ Portugal defined in a way that is easy to understand for laymen. ⇒ Romania There is an ‘other’ category for conditions that are not ⇒ Saudi Arabia on the list. By stating that participants should be 45 years ⇒ Slovenia or older, it is assumed that approximately 70% of the ⇒ Spain sampled PC service users will have at least one chronic ⇒ Switzerland condition. This assumption was based on available popu- ⇒ USA lation data7 combined with the hypothesis that restric- ⇒ Wales (UK) tion to individuals that visited the PC practice in the 4 de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 Open access preceding 6 months would yield a slightly higher propor- family physicians, may be used to identify (all) eligible PC tion. If there is evidence or a valid reason to believe that practices. Sampling guidelines drafted by the consortium this assumption will not hold in a particular country, the could help National Project Managers develop the most National Project Manager will adjust the sample size per appropriate sampling procedure for their country, which PC practice for the Field Trial as outlined below. should ensure that PC practices staffed by multiple PC professionals cannot be included in the sample more than Sample size once. The sampling procedure for PC practices should be Sample sizes for the Field Trial are based on statistical consistent with the principle of probability sampling as simulations carried out on a large data set of patient-­ much as possible. This entails that each PC practice in the reported care experiences from 34 countries partici- sampling frame should have a non-­zero and equal chance pating in the QUALIty and COsts of Primary Care in to be included in the sample. Europe (QUALICOPC) study.24 The simulations were Sampling of PC service users will be kept as simple as conducted with the data of four patient-­reported experi- possible by randomly drawing a fixed number of eligible ence scales (communication, access, comprehensiveness patients per PC practice. National Project Managers and continuity of care), each consisting of several items. design the sampling procedure for PC service users The main criterion for determining the sample size was together with the participating PC practices, respecting the reliability of country-­level estimates for each of the applicable legislation. four scales. Given the characteristics of the scales and the clustering within practices and countries, the value of this Availability of auxiliary data for survey weighting reliability depends largely on the number of practices per National Project Managers are asked to provide data on country and the number of patients per practice. The key characteristics of the population of PC practices and/ reliability of each scale was simulated for different combi- or of the sampled PC practices in their country, to allow nations of numbers of PC practices and patients per prac- calculation of weights to correct for possible response tice. The combination of 25 practices and 50 patients per bias. Similarly, they are asked to provide basic information practice resulted in highly reliable scales (>0.85). There- about the population of people aged 45 years or older who fore, we estimate that 25 PC practices per country and 50 have been in contact with a PC practice in the preceding eligible service users per PC practice (1250 eligible PC 6 months and/or the sampled PC service users. The feasi- service users per country) are required to participate in bility of survey weighting will depend on the availability the Field Trial. of such data. Accordingly, a survey weighting plan for the Depending on countries’ expected response rates— Main Survey will be developed based on which popula- if possible, based on previous survey studies among PC tion/sample data National Project Managers are able to providers and service users in the participating coun- retrieve in the Field Trial. tries—National Project Managers determine the sample sizes they need to achieve the required numbers of partic- Recruitment strategy ipants. It is evident that higher response rates are pref- To support the recruitment process, National Project erable above larger samples, but unrealistic expectations Managers develop a communication strategy tailored to of response rates should be avoided to ensure we will get country-­specific needs and preferences, local resources sufficient Field Trial data of each country to test the survey and communication channels, in consultation with instruments and data collection procedure for the Main national stakeholders and the consortium. To invite Survey. Response rates of 35% among PC practices and of selected PC practices, the mode of communication can 50% among PC service users have been set as targets for be either by email, a phone call or a letter first, with each country. Whether these targets are feasible will be follow-­ up contacts of any form. To invite selected PC concluded after the Field Trial. Based on the experiences service users, National Project Managers decide together and results of the Field Trial, sample sizes for the Main with the participating PC providers how their patients Survey will be set. will be approached. Regardless of the initial mode(s) of communication, all selected PC practices and service Sampling procedure users will receive written information about the survey at To achieve representative results, potential sampling some stage of the recruitment process and give informed frame(s) should be identified in each country and their consent. quality should be assessed based on estimated coverage and risk of bias. As the unit of data collection is a PC Questionnaire development facility (independent of size in terms of number of For the development of the PC service user and prac- patients or care professionals), complete and actual tice questionnaires and for supporting the plan of anal- national lists of these facilities are the ideal sampling ysis, a conceptual framework was developed through a frame of PC practices, or combined lists with a regional systematic, replicable, iterative and inclusive process.25 coverage. If such lists are not available or the quality is The framework identifies the following domains: poor, registries or member lists of key professional organ- patient-­reported outcomes (symptoms, functioning, self-­ isations in PC, such as colleges of general practitioners or reported health status, health-­ related quality of life); de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 5 Open access patient-­ reported experiences of care (access, compre- Translation and cognitive testing hensiveness, continuity, coordination, safety, people-­ The draft questionnaires have been translated in all centredness, self-­ management support, trust, overall national languages of the participating countries using perceived quality of care); service users’ health and care a stepwise team approach based on the TRAPD model capabilities; service users’ health behaviours (physical (Translation, Review, Adjudication, Pretest, Documenta- activity, diet, tobacco use, alcohol use); service users’ indi- tion).33 For reasons of efficiency and cross-­country compa- vidual and socio-­demographic characteristics; PC delivery rability, we advised National Project Managers to establish system (characteristics of the PC facility; characteristics of a collaboration of their national translation team with the the main PC professional); characteristics of the health translation agency cApStAn, which has ample experience system, policy and context. with this approach for international comparative surveys. Procedures for development of the questionnaires will However, they could also choose to apply the method- be described in full separately. Here we provide a brief ology with a completely national team, with only a final description of the development process and content of review by cApStAn. Countries that wish to have the ques- each questionnaire. tionnaire for PC service users available in other languages The questionnaire for PC service users aims to assess are encouraged to produce additional translations them- the primary outcomes of the PaRIS survey, that is, the selves or collaborate with other participating countries patient-­ reported outcomes and care experiences, as in this. The consortium will take responsibility for devel- well as PC service users’ characteristics, including demo- oping some translations for large minority groups, which graphics, chronic conditions, health and care capabili- will be offered to countries for use in the Main Survey. ties and health behaviours. A comprehensive approach The (translated) draft questionnaires are currently including a number of systematic reviews, engagement being tested by means of cognitive interviews with 10 PC with international stakeholders (Patient Advisory Panel, providers and 20 PC service users in each participating Technical Advisory Community and National Project country, preferably in two rounds: a first round with half Managers) was conducted for the identification of the total number of participants, resulting in recommen- eligible scales and items. Following a mapping exer- dations for either adaptation of the source question- cise onto the conceptual framework, four instruments naires or localisations, and a second round to check and for each domain were shortlisted using predefined confirm the modified questionnaires. The modified ques- criteria. The psychometric performance of the candi- tionnaires will then be piloted in the Field Trial. Based date instruments was assessed using the Evaluating the on the data collected with the Field Trial and the experi- Measurement of Patient-­Reported Outcomes (EMPRO) ences of the National Project Managers, final adjustments method.26 A subsequent modified Delphi procedure was may be made to the questionnaires before being used in implemented for selecting a core instrument for each the Main Survey. domain and additional relevant scales/items. Further consultations took place with the relevant stakeholders to confirm the suitability of the proposed questionnaire, Data collection methods whose feedback resulted in a number of iterations until a The two questionnaires have been developed for online final draft was agreed.27 The draft questionnaire consists administration; PC service users may also opt for a paper-­ of 120 items, including the PROMIS Scale V.1.2—Global and-­pencil version. Additional administration modes (eg, Health,28 WHO-­ 5 Well-­being Index,29 Person-­centred by telephone or face-­to-­face) may be offered in coun- Coordinated Care Experience Questionnaire30 31 and tries where response to online or postal questionnaires Porter-­Novelli Scale.32 is expected to be insufficient. National Project Managers The questionnaire for PC practices aims to collect infor- will set up a helpline for PC service users and providers mation about the characteristics of and services provided in their country, to refer to in case of questions about by the PC practice. The draft questionnaire has been the questionnaire. PC service users may also contact developed applying the same principles that guided the the helpdesk for help in completing the questionnaire, development of the questionnaire for PC service users, if they are unable to do this alone and informal help is including a literature review to identify existing ques- not available. A final question has been added to the PC tionnaires, mapping of questions to the domains of the service user questionnaire on whether the questionnaire conceptual framework, ranking and selection of items. was completed independently or with help from another The final draft PC practice questionnaire consists of 40 person. items to be completed by medical staff (eg, physician, National Project Managers will receive daily over- nurse) or non-­medical staff (eg, practice manager). It views from the international data management team covers the following topics: profile of the practice/facility that contain unique serial numbers that correspond to (eg, location and type, access and services) and key online responses to the survey. Only the organisation that elements of chronic care, including care planning, coor- performs data collection within a country will be able dination, follow-­up care and self-­management support. In to link these numbers to names and addresses for the addition, the questionnaire contains some questions on purpose of sending reminders. It is recommended that the impact of COVID-­19 outbreak on care delivery. non-­responders are reminded at least two times within 6 de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 Open access a period of 4 weeks after the invitation to complete the correlation coefficients will be calculated to assess the questionnaire. extent to which the scores of PC service users on the outcome variable are clustered in, respectively, the PC Data management practice they had been in contact with and the country (or Completed online questionnaires will be saved in the region) in which the PC practice is located. The country secured international database created by the consor- level (or regional level, if applicable) variance will be used tium. For countries that cannot or do not want to use to calculate predictions of the patient-­reported indicators this centralised approach, National Project Managers will (posterior means or empirical Bayes estimates) corrected create national level databases to store online survey data. for case-­mix variables for each country and/or region, For those countries the consortium provides guidance to including CIs. Case-­mix variables are variables that may avoid deviations that could threaten cross-­country compa- hamper a fair comparison of patient-­reported indicators rability of the data. between countries or regions. Potential case-­ mix vari- Completed paper-­ and-­ pencil questionnaires will be ables for analysis of the Field Trial data will be selected returned by the responders to their National Project based on theoretical notions and empirical evidence.35–40 Manager (or subcontractor), who is responsible for a Depending on the data analysis of the Field Trial, the secured storage of these questionnaires nationally, and case-­mix adjustment in the analyses of the Main Survey for data entry in a secured database. The consortium will data may be modified. provide instructions to ensure similar coding and equal Research questions 1 and 2 focus on outcomes and variable and value names as in the international database. care experiences of people living with chronic conditions In the national databases data collected by the question- compared with those without chronic conditions. This will naires will be encrypted and stored separately from any be specified in the models by including a dummy variable: potentially identifying participant data. chronic condition(s) present (vs absent). This allows for Identifying participant data will be securely saved in examination of variance in outcomes and experiences of each participating country as long as needed for data people living with and without chronic conditions. This collection (eg, sending reminders) and evaluation of the dummy variable will also be used in country-­level predic- representativeness of the participants. To the latter end, tions to estimate the outcomes and care experiences of aggregated data on participant characteristics (eg, spread people living with chronic conditions compared with the of the responding PC practices and service users across outcomes and experiences of people without chronic the country) will be compiled, after which identifying conditions. participant data will be destroyed. Identifying partici- To answer research question 3, background character- pant data will never be transferred to the international istics of PC service users will be added to the null model; database. their fixed effects will be estimated separately and simul- Once the data collection is finished, the National taneously, as far as the data allow. Likewise, interaction Project Managers will transfer the pseudonymised micro- effects that are considered relevant (guided by hypoth- data collected by the questionnaires as well as the aggre- eses based on theory and empirical evidence) will be esti- gated data needed to assess the representativeness of the mated, as far as feasible. To answer research questions participants and to correct for response bias to the inter- 4 and 5, the models will be extended with independent national database. variables at the level of PC practices and countries, and their fixed effects will be estimated. Cross-­level interac- Statistical methods tion effects may also be considered. The data analysis The statistical analyses to answer the main research ques- plan for the Main Survey will be reviewed and consoli- tions will be conducted by the international consortium. dated after analysis of the Field Trial data. The plan of analysis includes a number of preparatory steps, such as the evaluation of the survey operations Patient and public involvement (eg, sampling and data collection), data cleaning, quality Patients’ priorities, experiences and preferences form the control and the evaluation of the psychometric perfor- basis of the PaRIS survey and were brought forward through mance of items and scales in the questionnaires. the participation of the European Patients’ Forum and the To answer the research questions, a series of multilevel Picker Institute in the PaRIS Taskforce,21 which advised the regression analyses34 will be conducted based on a three-­ OECD Secretariat on the design and the research questions level model. Units of analysis are individual PC service of the PaRIS survey. A patient advisory panel,22 composed users (level 1), who are nested within PC practices (level of representatives of international and national patient 2), which are nested in countries (level 3). Note that in federations, has been established to advise the OECD some countries where regional health systems exist there Secretariat and the PaRIS-­SUR consortium on all stages of will be an intermediate level. survey development and implementation. National Project For each of the questions, a null model will be defined Managers seek advice from patient and citizen representa- to estimate the variance of the outcome variable at stake tives in their country in developing national communica- specifying random intercept coefficients only, that is, esti- tion and engagement strategies. Patients and citizens from mated variance at level 1, level 2 and level 3. Intraclass many countries also contributed to the development of the de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 7 Open access conceptual framework and survey questionnaires through Reporting of study results their participation in focus groups, a Delphi study, and The consortium develops a secured information tech- cognitive interviews. nology platform that will allow stakeholders in partici- pating countries to receive feedback and answer their own questions, based on the data collected by the Main ETHICS AND DISSEMINATION Survey. Public reporting by the OECD will predominantly Research ethics approval focus on the five main research questions, for which the National Project Managers seek ethical approval of the consortium will provide the results of the data analysis. PaRIS survey, if this is required in their country. The In addition, the consortium will publish on issues that consortium has developed guiding materials to help are important from a scientific point of view, for instance them preparing the ethical review documents, including methodological questions (eg, effects of administration (1) information about the PaRIS survey and its ratio- modes on outcome variables in countries and correction nale, (2) elements of the study protocol, (3) information methods) and questions of a more explanatory nature about privacy protection and data security, (4) controller, (underlying causal factors and processes). processor and legal basis in different phases of data collec- tion and analysis, (5) draft invitation letter for PC service Protocol amendments users and (6) an overview of the PaRIS-­SUR consortium Based on the evaluation of the Field Trial, the study and the OECD, who commissioned the PaRIS survey to protocol will be modified for the Main Survey. These modi- the PaRIS-­SUR consortium. There was no overarching or fications will include the calculation of the sample sizes set coordinating ethics committee involved. for the Main Survey (based on insight in response rates in Consent of participants countries and analysis of the Field Trial data). Other modi- Informed consent will be obtained from PC providers fications may include final changes to the questionnaires and service users for participation in the survey. To this or data collection procedure (eg, number or nature of end, all invited PC providers and service users receive reminders), if considered necessary based on the experi- written information, addressing all issues that are neces- ences of the National Project Managers or analysis of the sary to be able to give informed consent. The consortium Field Trial data. As the Field Trial is intended to pilot both has prepared this information, which will be translated the survey instruments and implementation of the survey by the National Project Managers or their subcontractors. in the participating countries, any modifications that may PC providers and service users who want to participate improve the implementation of the Main Survey or the give informed consent by responding to a first question usefulness of its results for participating countries will be in the survey instrument. considered. Confidentiality The PaRIS survey will meet the highest regulatory stan- DISCUSSION dards for the legal and safe processing of sensitive personal The PaRIS survey is the first worldwide cross-­ country health data, including OECD Data Protection Rules41 and study implementing patient-­ reported indicators for any applicable national legislation, including the Euro- the purpose of policy development and monitoring of pean Union General Data Protection Regulation. chronic care management in PC settings. To ensure that Within countries, confidentiality is governed by the the survey will collect information that is useful for coun- National Project Manager according to national legisla- tries in assessing the performance and quality of care tion. Across country borders, only pseudonymised and/ from the perspective of people living with chronic condi- or de-­identified data will be transferred while the key tions, it is crucial that the survey is being implemented remains within country boarders. For PC service users, in the daily practice of nationally representative samples this key will be destroyed on completion of data collec- of PC providers. This requires nationwide coverage of tion. For practices, the key remains available to allow participating PC practices, including those that may feedback of aggregated results. be located in remote areas, have limited resources for chronic care management in terms of staff or equipment, Communication and engagement plan or less advanced information technology for care delivery To support the implementation of the PaRIS survey, and practice management. This has implications for the National Project Managers involve local associations of design of the survey and its implementation. For instance, citizens/patients and PC providers at an early stage in rigorous criteria for probability sampling of PC practices communicating about the survey with people living with or service users may not always be fully met and local chronic conditions and PC professionals. To this end, adaptations of the survey instruments will be needed to National Project Managers are developing a communica- ensure understanding and relevance in a local context tion and engagement plan in consultation with national but may also affect cross-­country comparability of the stakeholders. The consortium provides information data. To find an optimal balance between all needs, inter- materials that can be translated and adapted to the local ests and requirements, it is key that the development of context for use in countries. the PaRIS survey is a collaborative effort of all involved. 8 de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 Open access This is crucial to achieve the ultimate goal of PaRIS, which Commission to develop and implement the PaRIS survey. Authors LT, RW, OG and JB is to assess the performance and quality of PC from the work for a private company. perspective of people living with chronic conditions to Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to support countries in developing people-­centred policies the Methods section for further details. and practices. To date, the development process has been Patient consent for publication Not applicable. an experience that already provided various opportuni- Provenance and peer review Not commissioned; externally peer reviewed. ties to exchange experiences and learning from various stakeholder perspectives within and across countries. Open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which In 2022 Field Trial data will become available to further permits others to distribute, remix, adapt, build upon this work non-­commercially, improve the survey instruments and countries’ implemen- and license their derivative works on different terms, provided the original work is tation roadmaps. Lessons learnt will be used to optimise properly cited, appropriate credit is given, any changes made indicated, and the use measurements of experiences and outcomes in the Main is non-­commercial. See: http://creativecommons.org/licenses/by-nc/4.0/. Survey thereby providing countries with information they ORCID iD can use to make their health systems more responsive to Mieke Rijken http://orcid.org/0000-0001-6070-4091 peoples’ needs. Author affiliations 1 Nivel (Netherlands Institute for Health Services Research), Utrecht, The Netherlands REFERENCES 2 1 Alonso J, Ferrer M, Gandek B, et al. Health-­Related quality of life Organisation for Economic Co-­operation and Development, Paris, France associated with chronic conditions in eight countries: results from the 3 Fundacion Avedis Donabedian, Barcelona, Spain International quality of life assessment (IQOLA) project. Qual Life Res 4 Universitat Autònoma de Barcelona, Barcelona, Spain 2004;13:283–98. 5 Red de Investigación en Servicios de Salud en Enfermedades Crónicas, Barcelona, 2 Verhaak PFM, Heijmans MJWM, Peters L, et al. Chronic disease and Spain mental disorder. Soc Sci Med 2005;60:789–97. 6 3 Scharn M, Oude Hengel K, Boot CRL, et al. Influence of chronic Optimedis AG, Hamburg, Germany diseases on societal participation in paid work, volunteering and 7 Department of Sociology / Department of Human Geography, Utrecht University, informal caregiving in Europe: a 12-­year follow-­up study. J Epidemiol Utrecht, The Netherlands Community Health 2019;73:136–41. 8 Health Services and Policy Research Group, University of Exeter Medical School, 4 OECD/European Union. Chapter 1. The labour market impacts of ill-­ Exeter, UK health. In: Health at a glance: Europe 2016. state of health in the EU 9 cycle, 2016. Ipsos MORI UK Ltd, London, UK 5 Rijken M, Spreeuwenberg P, Schippers J, et al. The importance 10 Department of Family Medicine, National University Health System, Singapore of illness duration, age at diagnosis and the year of diagnosis for 11 Department of Health and Social Care Management, University of Eastern Finland, labour participation chances of people with chronic illness: results Kuopio, Finland of a nationwide panel-­study in the Netherlands. BMC Public Health 2013;13:803. 6 Rijken M, Groenewegen PP. Money does not bring well-­being, but Twitter Candan Kendir @candankendir it does help! the relationship between financial resources and life satisfaction of the chronically ill mediated by social deprivation and Acknowledgements We thank all experts, including (representatives of) loneliness. J Community Appl Soc Psychol 2008;18:39–53. patients, citizens and care professionals, who contributed to the design, 7 Buttorff C, Ruder T, Bauman M. Multiple chronic conditions in the development and/or implementation of the PaRIS survey, either by their United States. RAND Corporation, 2017. participation in the PaRIS Taskforce, Patient Advisory Panel or Technical Advisory 8 World Health Organization. Noncommunicable diseases. Factsheet, Community and/or by their participation in the international focus groups, the 2021. Available: https://www.who.int/news-room/fact-sheets/detail/​ Delphi study or the cognitive interviews in their country and/or their contribution noncommunicable-diseases to the development of national communication and engagement strategies. 9 Violan C, Foguet-­Boreu Q, Flores-­Mateo G, et al. Prevalence, determinants and patterns of multimorbidity in primary care: We also thank all National Project Managers for their help in improving the a systematic review of observational studies. PLoS One survey instruments and adapting them to the local context, and for their 2014;9:e102149. dedicated efforts to pilot the PaRIS survey and prepare its implementation in the 10 Chaker L, Falla A, van der Lee SJ, et al. The global impact of participating countries. non-­communicable diseases on macro-­economic productivity: a systematic review. Eur J Epidemiol 2015;30:357–95. Contributors MvdB and NK designed the PaRIS survey in collaboration with 11 World Health Organization. Who global strategy on people-­centred other OECD staff, the Working Party for PaRIS and in consultation with the PaRIS and integrated health services. interim report. Geneva: WHO, 2015. Taskforce. WB and JMV participated in the PaRIS Taskforce. DdB, MB, WB, PG, 12 Starfield B. Is primary care essential? Lancet 1994;344:1129–33. OG, RS, LT, JMV, RW and MR drafted the initial project plan based on the Terms of 13 Kringos DS, Boerma WGW, Hutchinson A, et al. The breadth of Reference set by the OECD and the previous work conducted by the OECD together primary care: a systematic literature review of its core dimensions. with the PaRIS Taskforce. MvdB, CK, KdB and NK reviewed this initial plan and BMC Health Serv Res 2010;10:65. contributed to the definite project plan, including the project planning, development 14 Schäfer WLA, Boerma WGW, Spreeuwenberg P, et al. Two decades of change in European general practice service profiles: conditions of the PaRIS survey conceptual framework and survey instruments. DdB, MvdB, associated with the developments in 28 countries between 1993 and MB, JB, WB, KdB, PG, OG, CK, NK, IP, RS, LT, JMV, RW and MR prepared the 2012. Scand J Prim Health Care 2016;34:97–110. implementation of the PaRIS survey together with the National Project Managers, 15 Starfield B, Shi L, Macinko J. Contribution of primary care to health supervised by the Working Party for PaRIS. DdB and MR wrote the draft manuscript; systems and health. Milbank Q 2005;83:457–502. MvdB, MB, JB, WB, KdB, PG, OG, CK, NK, IP, RS, LT, JMV and RW critically reviewed 16 Valentijn PP, Schepman SM, Opheij W, et al. Understanding the draft manuscript and provided additional inputs. All authors (DdB, MvdB, MB, integrated care: a comprehensive conceptual framework based JB, WB, KdB, PG, OG, CK, NK, IP, RS, LT, JMV, RW and MR) read and approved the on the integrative functions of primary care. Int J Integr Care final manuscript for publication. 2013;13:e010. 17 Ricci-­Cabello I, Stevens S, Dalton ARH, et al. Identifying Funding Funding for the PaRIS survey was provided by the Organisation for Primary Care Pathways from Quality of Care to Outcomes and Economic Co-­operation and Development (OECD), participating countries and the Satisfaction Using Structural Equation Modeling. Health Serv Res European Commission. 2018;53:430–49. 18 Organisation for Economic Co-­operation and Development. Competing interests The authors’ institutes received funding from OECD Recommendations to OECD ministers of health from the high level Member States, other countries participating in the PaRIS survey and the European reflection group on the future of health statistics, 2017. Available: de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424 9 Open access www.oecd.org/health/health-systems/Recommendations-from-high-​ 31 Lloyd H, Fosh B, Whalley B, et al. Validation of the person-­centred level-reflection-group-on-the-future-of-health-statistics.pdf coordinated care experience questionnaire (P3CEQ). Int J Quality in 19 Black N, Burke L, Forrest CB, et al. Patient-­Reported outcomes: Health Care 2019;31:506–12. pathways to better health, better services, and better societies. Qual 32 Maibach EW, Weber D, Massett H, et al. Understanding consumers' Life Res 2016;25:1103–12. health information preferences: development and validation of a brief 20 OECD Health Ministers. The next generation of health reforms: screening instrument. J Health Commun 2006;11:717–36. ministerial statement, 2017. Available: www.oecd.org/health/​ 33 Harkness J, van de Vijver FJR, Johnson TP. Questionnaire design in ministerial/ministerial-statement-2017.pdf comparative research. In: Harkness J, van de Vijver FJR, Mohler PP, 21 Organisation for Economic Co-­operation and Development. Patient-­ eds. Cross-­Cultural survey methods. Hoboken, NJ: Wiley & Sons, Reported indicators survey (Paris): measuring what matters. public 2003: 19–34. version of the revised proposal submitted to the health Committee. 34 Leyland AH, Groenewegen PP. Multilevel modelling for public health OECD, 2018. 22 The PaRIS patient Advisory panel. Available: https://www.oecd.org/​ and health services research. health in context. Switzerland: Springer health/paris/PaRIS-Patient-Advisory-Panel.pdf Open, 2020. 23 The Paris technical Advisory community. Available: https://www.​ 35 Iezzoni LI. Risk adjustment for performance measurement. In: Smith oecd.org/health/paris/PaRIS-Technical-Advisory-Community.pdf C, Mossialos E, Leatherman S, et al, eds. Performance measurement 24 Schäfer WLA, Boerma WGW, Kringos DS, et al. QUALICOPC, a for health system improvement: experiences, challenges and multi-­country study evaluating quality, costs and equity in primary prospects. Elsevier, 2009. care. BMC Fam Pract 2011;12:115. 36 de Boer D, van der Hoek L, Rademakers J, et al. Do effects of 25 Porter I, Rijken M, Groene O. The International survey of people common case-­mix adjusters on patient experiences vary across living with chronic conditions (Paris survey): development of the patient groups? BMC Health Serv Res 2017;17:768. conceptual framework. Qual Life Res 2021;30:S48. 37 Hatfield LA, Zaslavsky AM. Implications of variation in the 26 Valderas JM, Ferrer M, Mendívil J, et al. Development of EMPRO: a relationships between beneficiary characteristics and Medicare tool for the standardized assessment of patient-­reported outcome advantage CAHPS measures. Health Serv Res 2017;52:1310–29. measures. Value Health 2008;11:700–8. 38 Damman OC, de Boer D, Hendriks M, et al. Differences between 27 Porter I, Rijken M, Groene O. The International survey of people living family practices in the associations of patient characteristics with with chronic conditions (Paris survey): development of the patient health care experiences. Med Care Res Rev 2011;68:725–39. questionnaire. Qual Life Res 2021;30:S42. 39 Johnson ML, Rodriguez HP, Solorio MR. Case-­Mix adjustment and 28 Hays RD, Bjorner JB, Revicki DA, et al. Development of physical and the comparison of community health center performance on patient mental health summary scores from the patient-­reported outcomes experience measures. Health Serv Res 2010;45:670–90. measurement information system (PROMIS) global items. Qual Life Res 2009;18:873–80. 40 Paddison C, Elliott M, Parker R, et al. Should measures of patient 29 Topp CW, Østergaard SD, Søndergaard S, et al. The WHO-­5 experience in primary care be adjusted for case mix? Evidence well-­being index: a systematic review of the literature. Psychother from the English general practice patient survey. BMJ Qual Saf Psychosom 2015;84:167–76. 2012;21:634–40. 30 Sugavanam T, Fosh B, Close J, et al. Codesigning a measure 41 Recommendation of the Council concerning guidelines governing of person-­centred coordinated care to capture the experience the protection of privacy and Transborder flows of personal data. of the patient: the development of the P3CEQ. J Patient Exp Available: https://legalinstruments.oecd.org/en/instruments/OECD-​ 2018;5:201–11. LEGAL-0188 10 de Boer D, et al. BMJ Open 2022;12:e061424. doi:10.1136/bmjopen-2022-061424