Priorities and preferences for care of people with multiple chronic conditions

Abstract Background To guide the development of high‐quality care for people with multiple chronic conditions, partners of the European Joint Action CHRODIS developed the Integrated Multimorbidity Care Model. To assess its suitability for improving care for people with multimorbidity in the Netherlands, the model was piloted in a primary care setting with both patients and care providers. Aim This paper reports on the patient perspective, and aims to explore the priorities, underlying values and preferences for care of people with multimorbidity. Participants and methods Twenty persons with multimorbidity (selected from general practice registries) participated in a focus group or telephone interview. Subsequently, a questionnaire was completed by 863 persons with multimorbidity registered with 14 general practices. Qualitative data were thematically analysed and quantitative data by means of descriptive statistics. Results Frequently prioritized elements of care were the use of shared electronic health records, regular comprehensive assessments, self‐management support and shared decision making, and care coordination. Preferences for how these elements should be specifically addressed differed according to individual values (eg weighing safety against privacy) and needs (eg ways of coping with multimorbidity). Conclusion The JA‐CHRODIS Integrated Multimorbidity Care Model reflects the priorities and preferences for care of people with multimorbidity in the Netherlands, which supports its relevance to guide the development of person‐centred integrated care for people with multiple chronic conditions in the Netherlands. Patient contribution European patient experts contributed to the development and applicability assessment of the JA‐CHRODIS Integrated Multimorbidity Care Model; Dutch patients participated in focus groups, interviews and a survey.


| INTRODUC TI ON
After the wake-up call by the World Health Organization in 2002 that chronic conditions would be the biggest challenge for health systems worldwide in the new millennium, 1 chronic care today requires reform and innovation once again, now because of the growing populations with multiple chronic conditions. In some countries, the number of people with two or more chronic conditions already exceeds the number of people with one chronic condition. [2][3][4][5] Awareness has raised that the multifaceted nature of multimorbidity calls for alignment of services that go beyond (single) disease management programmes and single-disease chains of care, [6][7][8] necessitating intersectoral collaboration as well, particularly with social care and community services. Moreover, people with multimorbidity need person-centred care, which requires care professionals to take up new roles and develop new skills. 9-11

| Integrated Multimorbidity Care Model
Similar to the previous situation with the management of (single) chronic conditions, new care models are being developed to guide the development and implementation of person-centred integrated care for people with multiple chronic conditions. One of these models is the Integrated Multimorbidity Care Model (IMCM), which was developed by the European Joint Action CHRODIS based on previous integrated care models, [12][13][14] scientific evidence 15 and consultation of international experts. This resulted in agreement on 16 components of high-quality multimorbidity care, structured under five main components (Box 1). 16 Pilots in Italy, Lithuania and Spain that were conducted as part of the European Joint Action CHRODIS+ 17 show the potential of the model to improve care for people with multimorbidity in various settings and in countries with different health systems. 18

| Patient voice
Although patient representatives participated in the expert consultation, the patient voice in the development of the IMCM has been limited until now. Some previous studies reported priorities and preferences for care of people with multimorbidity. A survey in the United States among primary care patients to evaluate the implementation of a new team-based care model demonstrated that, although all patients considered continuity of care very important, people with multimorbidity attached even greater value to it. 19 A qualitative study in Canada and New Zealand showed that older participants with multimorbidity attached high value to good patientprovider communication (eg feeling heard; getting enough time; attention paid to their comprehensive needs and priorities), having a trusted care provider, knowledge to self-manage health and care, and smooth access to health and social services. 20 Another qualitative study, in the United States, confirmed that older people with multimorbidity value being heard and a person-centred communication and a single care provider coordinating their care and helping them prioritize competing demands. 21 Although these results support the relevance of certain components of the IMCM, empirical evidence from other, also European, countries is needed, since values and care preferences of people with multimorbidity may be context-and culture-sensitive. 22 To contribute to the empirical evidence of the model as a basis for multimorbidity care development and to further explore its relevance from the patient perspective, we conducted a mixed-methods study guided by the following research questions: BOX 1 Components of the integrated multimorbidity care model (see Palmer et al., 2018 for detailed description of components) These questions are also relevant considering that implementing person-centred integrated care in real-life care settings will often require a stepwise approach over a number of years. Knowing patients' priorities may be helpful to set goals and select a first set of interventions to start with.

| ME THODS
A mixed-methods study was conducted to make optimal use of the benefits of both qualitative and quantitative research methods. To assess patients' priorities for multimorbidity care (research question 1), we considered a qualitative study design most appropriate as it allowed us to collect and analyse rich unstructured data, also because we aimed to identify the values underlying participants' priorities.
Therefore, we organized face-to-face focus groups and telephone interviews with people with multimorbidity. To assess patients' preferences for the concrete implementation of the prioritized components of care (research question 2), we conducted a survey among a larger group of persons with multimorbidity using a structured questionnaire. In this way, we were able to examine patients' preferences for a number of care interventions that could actually be implemented by Dutch general practices and their local partners.

| Participants
Participants were recruited from the National Panel of people with Chronic illness or Disability (NPCD). 23,24 In this panel study, about 1500 to 2000 adults with chronic conditions, randomly selected from general practices throughout the Netherlands, participate each year in surveys to monitor developments in chronic care and social participation. Occasionally, participants are invited for focus groups or interviews to collect more in-depth information.
In February 2018, we sent an invitation letter explaining the purpose and methods of the study to 314 panel members (aged 18+) who were randomly selected from the total group of 631 panel members eligible for this study, that is diagnosed with at least two chronic diseases and not having participated in a focus group or (telephone) interview over the last year. Though we preferred focus groups above telephone interviews, expecting that the interactive nature of focus groups would encourage participants to also reflect on other views than their own, we also offered the option for a telephone interview to reduce barriers for participation (particularly because of the possibly poor functional health status of the invited persons) to be as inclusive as possible.
Eight persons indicated their willingness to participate in a faceto-face focus group and 37 in a telephone interview (total 45; 14%).
As two persons were not able to participate in a focus group on the dates we could offer, six actually participated in a focus group. One group consisted of two participants and the other of four. Of the 37 persons who expressed their interest in a telephone interview, we randomly selected persons, and conducted and analysed interviews until we felt that data saturation had been reached. This was after 14 interviews (in addition to the focus groups). In addition to the information letter they received, we further explained the study verbally before the start of the focus group or interview and answered questions; all participants signed informed consent.

| Data collection
Each focus group was moderated by two researchers; the telephone interviews were conducted by three interviewers. To guide the focus groups, we developed ten statements ('For me it is important that …'; see Table 1), based on the five main components of the IMCM. 16 Each statement was printed on an A3 poster and hang (in a ran- For the telephone interviews, participants received a set of ten cards (each containing one of the ten statements; randomly ordered) by post some days before the interview with a letter asking them to read the statements beforehand and select three cards that best reflected what they valued most in caring for their multiple chronic conditions. During the interview, the interviewer invited the interviewee to explain his/her choice for the statements. Dependent on how much time this had taken, interviewees were also encouraged to comment on the statements they had not prioritized.
The focus groups and telephone interviews were audiotaped, with permission of the participants. The audiotapes were transcribed verbatim by an independent transcription service; all identifying information was removed from the transcripts.

| Data analysis
Descriptive statistics of participants' characteristics and selected statements were calculated. The transcripts were thematically analysed 25 ; the components of the IMCM were included as predefined codes for all transcripts (transcripts of the focus groups and telephone interviews were analysed together), but new codes were added for phrases that were not covered by the predefined codes. Only phrases in the transcripts that related to the concept of quality of care were coded. Initial coding of all transcripts was done by one researcher and checked by a second researcher. Differences in interpretation of (un-) coded phrases were discussed with a third researcher, if necessary to decide on initial coding. Subsequently, higher-order themes were identified based on these codes. As we were interested in participants' values behind their prioritization of certain aspects of multimorbidity care, which are usually not explicitly expressed, we examined codes including participants' explanations of why they prioritized certain aspects of care. The preliminary identified themes were reviewed by re-reading all relevant phrases of the transcripts and checked for alternative explanations. The final themes included both semantic and latent themes.

| Participants
Participants were selected from the clinical information systems of fourteen general practices that participated in a quality improve- The questionnaire contained seven questions (see Table 2) developed by the authors, to assess the preferences of the respondents regarding those components of the IMCM that had been prioritized most frequently by the participants in the focus groups and telephone interviews. Respondents were allowed to choose more than one answering option per question. Furthermore, they could explain their answers in an open text box. Additional data collected by the questionnaire were age, gender and health-related quality of life as assessed with the first item of the RAND-36 28 : In general, how would you say your health is? Response options are as follows: excellent, very good, good, moderate and poor.

| Data analysis
Univariate statistics were computed to describe the demographic characteristics of the participants and to assess their preferences for care.

| Participants
The group consisted of thirteen women and seven men. The mean age was 68.2 (SD: 12.3; range: 40-89). Many participants reported three or more chronic conditions, both somatic (eg diabetes, ischaemic heart disease, cancer, arthritis, asthma, COPD, multiple sclerosis, thyroid disorder) and mental conditions (anxiety disorder, ADHD, depression). Twelve persons (60%) evaluated their general health as moderate; five as good (25%); and the other three as excellent (5%), very good (5%) or poor (5%). Below, we describe the results of the qualitative analysis referring to the most frequently prioritized components.

| Shared electronic health records
Three quarters of the participants wanted to have all their health data in one electronic record shared by multiple care providers.
Participants were less unanimous about which parts of their data to share with whom. Some participants would favour to share as much information as possible with as many of their care providers: 'Just pressing a single button and they have all your data. Isn't that wonderful?' When asked which care providers should get access to their medical information, these people not only mentioned their GP, pharmacist and other medical specialists, but also their physiotherapist, dietician, dentist and many more. Some felt that social care and community services should also have direct access to their health record(s). Others mentioned conditions and restrictions for sharing their data. These people wished to be asked for permission each time their data were shared with or transferred to another care provider; they mentioned restrictions regarding the type or amount of data to be shared or regarding the type of care providers data could be shared with.
The different attitudes of participants towards sharing their health data seem to be related to how they value and balance continuity and safety of care against privacy. Participants who strongly favoured unlimited exchange of their data emphasized that this would improve the quality and efficiency of their consultations with care providers and decrease the risk of inappropriate or even harmful care interventions because of care providers missing essential information. Some of them also mentioned they did not want to be responsible for accurate information transfer between care providers; others just did not like to tell their story over and over again. There were also participants who did not favour unlimited exchange of their health data, however. These people seemed to be more inclined to protect their privacy. One

| Care preferences regarding prioritized IMCM components
Participants were asked to indicate their preferences for care regarding those IMCM components that had been prioritized most frequently by participants in the qualitative study. The results are presented in Table 2; below, we describe the most significant findings.

| Shared electronic health records
Forty per cent of all respondents reported that physicians should be able to share their medical information with all health-care providers involved they had contact with. Another 20% indicated that their medical data could also be shared with organizations/persons responsible for the assessment of their needs for support provided by the municipality, such as domestic help. At the same time, more than a quarter of the respondents indicated that the exchange of data among care providers should be limited to what is really necessary.

| Regular comprehensive assessments
More than a third of the respondents indicated they want a comprehensive assessment of their health and functioning when they feel this is necessary. These people explained that they know their own bodies best and that they want to determine themselves when a comprehensive assessment is needed. A similar proportion mentioned that they (also) want such an assessment if their GP considers this necessary. Arguments of this group are that the GP knows what is going on, has the expertise and that they trust the GP's judgement.
Almost a quarter of all respondents preferred to have a comprehensive assessment of their health and functioning every year. These people explained they consider this important from a preventive point of view, that they want to be reassured or that they consider it necessary because of their age.
In terms of comprehensiveness, about half of all respondents preferred an assessment to cover their physical and mental health and their functioning at home and in daily life. As an explanation of their preference for a broad assessment, respondents indicated that they consider it important that 'the whole picture' is seen and that everything influences each other. On the other hand, a substantial group (27%) indicated a preference for an assessment limited to their physical health.

| Self-management support and shared decision making
The vast majority of the respondents (70%) preferred to receive advice and support to self-manage their health from their GP. More than a quarter (28%) also indicated they wished to get such advice and support from another medical specialist. There was less enthusiasm for self-management support provided by a nurse working in general practice.

| Care coordination
Most respondents indicated a preference for organizing the care and support they need themselves (60%). There was also a substantial proportion (29%) who prefer their GP (or nurse practitioner/practice nurse) to take the role of care coordinator. These people explained that they are in good contact with their GP, that it feels familiar, that the GP knows them well and that GP and practice nurse have the necessary expertise and short lines of communication. Respondents who chose the 'other' option (9%) predominantly indicated that they do not need any care or support; a few mentioned they had no idea who could organize their care or support.

| Discussion of main findings
Qualitative analysis revealed that multimorbid patients' attitudes towards shared electronic health records depend on how much value they attach to either continuity of care and safety or their privacy.
This was confirmed by the survey results, demonstrating substantial support for shared electronic health records and exchange of medical data among care providers, but also caution to which care providers should get access to which data.
Regular comprehensive assessments were felt important from a preventive point of view, but also as a way of being reassured about one's health status. These values may reflect different coping styles: proactive, problem-focused coping versus reactive, emotionfocused coping. 29,30 In the latter, regular assessments may be useful to reduce emotional discomfort, whereas in the former they may provide starting points for proactive (self-)management. Both ways of coping could co-exist among people with multimorbidity, 31 but not necessarily. This may also explain why individual care plans, which are being used, amongst other things, to support patients' self-management, 32 did not get an equally high priority as regular comprehensive assessments.
Participants in the qualitative study who gave high priority to self-management support valued being in control of their health and care, but not all for the same reasons. Some pointed to the importance of self-management to maintain independent functioning, whereas keeping control was also induced by bad experiences with and a lack of trust in care providers.
The survey showed little interest in self-management support from primary care nurses and group-based or online self-management training courses. Most participants preferred to receive self-management support from their GP as part of their regular consultations, which is at odds with current policy to give nurses a greater role in supporting self-management of primary care patients with chronic conditions. 33,34 In this respect, it is also important to note that contact with social care and community services did not get high priority. Unfamiliarity, patient expectations (expecting medical solutions) and uncertainty about the benefits of these types of care may play a role. 35  Nevertheless, about a third of the participants in the survey preferred a professional to take this role, most often a GP or primary care nurse. These people believed the GP or nurse to have the expertise and also the shortest lines of communication with other care providers, while also having a good overview of their needs.

| Strengths and limitations
As in all qualitative studies, we had a limited number of participants in the focus groups and telephone interviews. Their age and gender distributions resembled those of the participants in the survey and of other multimorbid patients registered in Dutch general practices. 44 However, their self-rated health was worse (compared with the survey participants), which might explain their motivation to being heard. Moreover, one can expect people willing to participate in focus groups or interviews to be relatively empowered. As such, these people may also be better capable to organize their care.
Qualitative analysis of the focus group and interview data did not reveal elements of care that were felt important by people with multimorbidity that were not covered by the IMCM. However, it is important to note that our focus groups and interviews were guided by this model, which may have narrowed participants' frame of mind in thinking about important elements of multimorbidity care. Also, the fact that we provided only ten statements to the participants should be taken into account. Their specific formulation may have impacted participants' weighing of components in determining their priorities.
We therefore believe the exact ranking of prioritized components in with three or more chronic conditions in the Netherlands. 45

| Recommendations for future research
As mentioned earlier, values and care preferences of people with multimorbidity are likely to be context-and culture-sensitive. 22 That is why we could not rely on insights from previous studies conducted in the United States and other non-European countries.
However, this also means that the results we found regarding patients' care preferences in the context of Dutch primary care cannot be generalized to other settings and countries. We therefore recommend to conduct similar studies in other countries and settings as well. Moreover, we advise local care providers to explore the care preferences of their own patient populations with (multiple) chronic conditions, as practice populations may differ (eg age, ethnicity and socio-economic situation), which not only impact the population needs, but probably also impact the values and preferences for care. These insights could then be used to develop tailored integrated care at a local scale. Further research of patients' values and preferences for care may also focus on the role of patients' health beliefs and perceived health to strengthen the evidence for developing effective care interventions for people with (multiple) chronic conditions.

| CON CLUS ION
The JA-CHRODIS Integrated Multimorbidity Care Model covers elements of care that are considered of great importance by Dutch primary care patients with multimorbidity. As such, the model can be considered suitable to develop high-quality care for people with multimorbidity in the Netherlands, as has already been done in other European countries. The IMCM shows which elements of care need to be addressed in caring for people with multiple chronic conditions rather than specifying how this should be done, as this will depend on local resources and individual patient preferences.

ACK N OWLED G EM ENTS
The authors thank all people with multimorbidity who participated in the focus groups, telephone interviews or the survey for their valuable contributions to this study. Many thanks also to the GPs and practice staff of De ZOED Schalkwijk / De ZOED Boerhaavewijk and all other participating practices for their efforts in collecting the survey data, and, above all, their willingness and motivation to improve multimorbidity care. The authors also wish to thank the partners of the Joint Actions CHRODIS and CHRODIS+ for the inspiring collaboration over the years to take care for people with multimorbidity in Europe to a higher level, and thank in particular Dr Graziano Onder for his dedicated leadership.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R CO NTR I B UTI O N S
MR designed the study, and was involved in the data collection as a moderator of the two focus groups and as an interviewer; she designed and implemented the patient survey, contributed to the qualitative and quantitative data analysis and wrote the first draft of the paper. RS contributed to the qualitative data analysis and reviewed the draft paper. CL was involved in the data collection as an interviewer, and contributed to writing and reviewing of the draft paper. MJLB was involved in the data collection as a second moderator of one of the focus groups and as an interviewer, and reviewed the draft paper. JK was involved in interpretation of the data and reviewed the draft paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.