Mirror meetings with frail older people and multidisciplinary primary care teams: Process and impact analysis

Abstract Objectives To analyse the process and impact of confronting multidisciplinary teams (MTs) in primary care with the experiences of frail older patients through mirror meetings (MMs), with the aim of supporting teams to organize care in a more patient‐oriented way. Methods Process and impact analyses were performed using a mixed‐method approach. MMs were held with 14 frail older patients and four MTs comprising 23 health‐care professionals (HCPs) in primary care in the Netherlands. Results Mirror meetings were feasible for frail older people living at home, although their recruitment was time‐consuming. Interaction between the patients was scarce, but they valued the opportunity to share their stories. HCPs preferred MMs overwritten reports about patient experiences. An impact analysis revealed four dominant professional areas for improvement: improve alignment with patient goals, improved communication with patients both orally and in writing, developing new pathways to connect with informal caregivers and an increased understanding that most HCPs are relative strangers to their patients. Conclusions Mirror meetings are a relatively simple and promising method for exploring the ways in which frail older patients experience care. Practice implications Given the right conditions, MMs could result in valuable processes to enable MTs to improve their working methods.

In 2015, the Organisation for Economic Co-operation and Development (OECD) emphasized the importance of integrating services from a patient's point of view rather than from a health-care provider's perspective. 8 Despite the fact that patient participation is high on the public agenda, studies have shown that multidisciplinary networks and teams often operate from a professional perspective, 9,10 while patient perspectives are rarely anchored in the design and working method of multidisciplinary teams (MTs). 11 Studies into interprofessional collaboration have largely focussed on the viewpoint of the professionals [12][13][14][15] or on more quantitative patient outcomes. [16][17][18] Researchers are hesitant to ask frail older people to participate in such studies due to their vulnerable state, 19 which means that little is known about the experiences of these patients regarding their multidisciplinary care teams. 20,21 In a previous study, we found that HCPs want to increase the consideration given to patient perspectives in the organization of their care, but are unsure how to realize this. [22][23][24] Many studies have investigated the quality and patient centeredness of care through the collection of survey data 25 or focus group interviews 26 about patient experiences, while others have researched the experience of patients invited to attend team meetings. 27 Besides some papers on so-called 'mirror meetings' (MMs; available in Dutch), we are not aware of any other publications in which patients have shared live feedback on care experiences with professionals. [28][29][30][31][32] The available literature suggests that learning from patients is increasingly important. The experiential expertise of patients seemed to offer a stimulating perspective on the provision of care, and the mirror meetings are an effective and powerful tool for generating learning points for health-care professionals and organizations from the patient's perspective. The open face-to-face confrontation appeals more to the individual HCP than survey results and other forms of indirect feedback. In order to achieve structural improvements in the provision of care from these learning points, a good follow-up process is considered necessary.

| Aim of the study
The objective of this study was to explore the feasibility of holding MMs with frail older patients and MTs in a primary care setting. Moreover, we were interested in the added value of MMs for MTs, as perceived by the HCPs. We aimed to detect potential adaptations that could be made to improve this method and organize multidisciplinary care for frail older people in a more patient-centred way. Issues that were addressed during the study were as follows: (1) the feasibility of holding MMs for frail older people; (2) the process of holding MMs involving older patients, HCPs and a moderator; and (3) the impact of MMs on the MTs.

| General methodology
In order to present MTs with the perspectives of elderly patients living independently in the community, a MM methodology was used. 29 The patients, seated in the 'inner circle', described their health-care experiences while HCPs, seated in the 'outer circle', listened but were not allowed to speak (Box and Figure 1). The stories told by the older people were the focus of the meetings. MMs can be seen as a method of collecting narrative data. 33

| Context of the study
This study was part of a project called PRECURO, which was focussed on gaining insights into the way in which multidisciplinary care (medical care, social services and community health services) in the Netherlands, in the region of Nijmegen, was organized for vulnerable older people living at home. Specific attention was paid to the experiences of older people, and the ways in which these experiences could be used to improve care. Data collection took place between February 2013 and June 2015, and consisted of interviews with frail older people, focus group meetings with HCPs, document analyses, observations of MT meetings, MMs and exchange meetings for MTs.

| Recruitment of participants
In the Netherlands, all patients are registered with a general practitioner (GP). GPs deal with more than 95% of all presented medical problems and arrange referrals to secondary care when needed.
Dutch GP services provide a comprehensive and patient-oriented approach with a high continuity of care. GPs also co-ordinate care for frail older patients with complex care needs. 34,35 From our network in the Nijmegen area of the Netherlands, we recruited four GP practices. We have approached these GP practices individually.
All four agreed to participate. During the selection of the practices,

Box 1 What is a mirror meeting?
A mirror meeting is a meeting of a group of patients, under the guidance of an independent moderator, in which the central question is how the patients experience the care they receive. The care providers involved are present only as listeners. The aim of the meeting is to increase the patient orientation of the care by making care providers aware of the patient perspectives. The success of a mirror meeting depends on the involvement of the care providers and requires them to be committed and have an open attitude to want to learn from patient feedback. The care providers are thus given a 'mirror image' by their own patients. 29 heterogeneity was sought using the following criteria: geographical location, population served (deprived, commuter, city, village), years of experience with multidisciplinary elderly care and the scale of the GP practice setting. All participating GPs, including their most important stakeholders, aimed to organize integrated care for frail older people. GPs and their stakeholders together formed the MTs in our study (Table 1).
Three months before the MMs, all members of the four MTs For the PRECURO study, a total of 44 frail older people attending the four selected GP practices were interviewed (2.2.) about various subjects concerning the organization of their care. Potential frail older participants were selected by the GP and/or the practice nurse. 36 Purposive sampling 37 was used in terms of sex, age, living situation, degree of fragility and care needs to ensure a representative sampling of older people. Patients had to have been discussed in the MT meeting to be included. Those with severe cognitive impairments were excluded from participating. After obtaining written consent, the potential patients were approached. During the interviews, each patient was asked whether they wanted to participate in a MM. If the answer was positive, they were sent further details about the MM. Transport to the meeting was arranged if necessary. Patients received an information sheet explaining the purpose of the meeting and confirming their appointment. These representatives had experience with holding MMs with patients and professionals in various health-care settings. After each MM, the manual was assessed to determine whether it required any adjustments.

| Mirror meetings and process analysis
Each MM lasted 90 minutes. The MMs took place at a location in the community familiar to the patients. If desired by the patient, an informal caregiver (IC) could be present. All meetings were audiotaped and transcribed verbatim. A transcription protocol was made.
The names of the patients were replaced by consecutive numbers on the transcripts to ensure anonymity and confidentiality. All meetings were observed 38 by three occupational therapy bachelor students and captured in field notes. 39 A semi-structured topic list (Appendix S1), developed by the research team (SG and HS), was used to support all four meetings.
The content of the topic list was determined by the research questions, supplemented with themes that emerged in the interviews with the elderly patients (2.2.). Also, data were collected from focus group interviews with the four MTs, in which the HCPs discussed the organization of care and mutual cooperation. 40 In addition to this input, the HCPs were invited to submit topics for the The HCPs typically sat behind their patients so they could listen to them but could not directly look them in the eye, because this may have discomforted the patients. Professionals were allowed to ask questions at the end of the meeting to clarify what had been raised by the patients; however, they were not allowed to defend themselves or to enter into discussions with the patients (Box ).
The process analysis was carried out using the observations recorded in field notes and recordings and the results of the evaluation by patients and HCPs at the end of the MM.

| Measuring impact
The impact of the MMs on MTs and their individual members was investigated in various ways. The care and welfare providers were observed during the MMs. Immediately after the MMs, they completed an online questionnaire about their experiences during the meeting.
Two months after the MMs, the HCPs received a report about the MMs, which they discussed in their next MT meeting. After the discussion of the report, the principal investigator (SG) spoke with each GP (in one case also with the community nurse) to learn the results

Moderators
Observers of this discussion. This can be seen as a form of member checking. 41 In doing so, a number of open questions were asked, such as 'What did you like about the mirror meeting?' and 'What actions will be taken as a result of the mirror meeting?'. Six months after the MMs, the MT members received an online questionnaire to measure the impact of the MMs.

| Data analysis
Qualitative data were analysed using Atlas-ti version 7.1. Open coding and thematic analysis were applied. 42

| Participants
Initially, we recruited 37 frail older people. After they were provided with more detailed information about the MMs by telephone, 14 decided not to participate. All of the reasons provided for this concerned the state of their health (eg feeling too depressed, having too much difficulty walking, being admitted to a care facility and cognitive impairments). After a second telephone call to the remaining 23 people 1 week before the meeting, another six people decided not to participate due to their poor health. At the last minute, three patients cancelled due to ill health. A total of 14 elderly people therefore participated in the MMs (Table 2).
Of the 14 patients, 57% were male, and the majority were over 80 years of age (64%), used multiple drugs (71%) and had ≥5 chronic disorders (79%). Half of the patients brought an IC to the meeting, either for logistical purposes or for mental support during the meeting. The moderator pointed out that the elder person herself/himself should be the primary speaker. Only factual information was sometimes provided by the IC.

| Process analysis
In all meetings, we observed that both patients and HCPs listened

| Impact analysis
The response rate to the questionnaire immediately after the MMs  • Each patient tells his/her own story; • You can express both appreciation and criticism; • Respect each other's privacy; • HCPs listen, but do not comment; • Patients participate in the role of an experience expert and not as a patient; • Minutes and tape recordings are transcribed anonymously.

(B) Topic list
Although the topics addressed during the MMs were recognizable for the patients, they did not fit seamlessly with the subjects on the minds of the HCPs. Several professionals suggested that the topic list should be composed by, or in collaboration with, the MT, based on their goals.

(C) Moderator
The HCPs were not very enthusiastic about the moderator. They felt that the questions asked were too general: 'Patients sometimes did not answer because they were presented with too much information'. Also, the moderator should have had more of an understanding of the local situation and the patients: 'I did not find the moderator suitable because she had no connection with the interviewees'.

| Insights and action
After the MMs and the discussion of the MM report in the MT meeting, the teams formulated the following points of action based on the experiences the patients shared with them: • To involve ICs in the care of their vulnerable relatives or friends, and find new ways to connect with them; • To visit the patient, prior to an MT meeting, to explain what the meeting is about and discuss the goals of the patient. After the MT meeting, the patient should then be informed about the outcomes of the meeting; • To provide better written information to the patient about 'who is who' (pictures of HCPs) and about the care plan.

| Summary
Mirror meetings with frail older people were found to be fea-

| Comparison with existing literature
As indicated in the introduction, little information was found on the subject of our study; however, we did find a book by Bijker, 43  Our experiences with the recruitment of patients were similar to those described in the existing literature, such as the challenges of travelling to the research site, the dropout rate throughout the study due to the deterioration of patient health, and the challenges of involving patients with cognitive and psychiatric conditions. 36,44 Many studies have elaborated on the importance of involving ICs for frail older people. [45][46][47] Our study endorses these findings.

| Strengths and limitations
As far as we could establish, the MM methodology has not been written about in international literature before, despite being a relatively simple and elegant way to involve the voice of the patient in the process of organising care with MTs.
Methodological triangulation was used in both data collection (observations, audio-taping and transcription, questionnaires, interviews) and the corresponding data analysis. Member checking was applied by sending the report of the MMs to the teams and reflect on the outcomes of the discussion of the report in the MT. This way of working makes qualitative research more robust.
The MM methodology used with this group was not piloted due to time pressures and limited resources. We did not perform further research into the experiences of the patients or the consequences for the care organization, although our material is suitable for this.
If time had allowed us to do so, we would certainly have done this.
The experiment was performed with four teams in the Netherlands; however, we believe our results are also relevant for HCPs in other settings and countries because frail elderly people living independently and receiving multidisciplinary care are fairly universal: MTs throughout the Western World are striving for ways to hear the patient's voice, which could be achieved using MMs.

| Conclusions
The

| Practice implications
With this research, we aimed to deliver usable information for HCPs

ACK N OWLED G EM ENTS
We would like to thank all patients, informal caregivers and health-and social care professionals who participated in our study for their openness towards us and for the time they invested in data collection. We would like to thank Joni Scholte and Anne Kuijpers for the preparation and organization of the mirror meetings and their support with data collection and analysis. We would like to thank Nancy Weijers and Nathalie Koopman from Zorgbelang Gelderland (a regional patient organization) for sharing their expertise in the field of mirror meetings. We are also grateful to them for moderating the mirror meetings and writing the reports. We would like to thank Bob van Bergen, Petra Vosman and Hanne Wilger, student occupational therapy at the HAN University of Applied Sciences, for their contribution to data collection and analysis. We would like to thank Sarah Jose of Radboud in'to Languages for her comments on our English writing.

AUTH O R S' CO NTR I B UTI O N S
SG contributed to the conception and design of the study, contributed to acquire funding for the study, collected, analysed and interpreted data, and drafted the manuscript. GM contributed to the design of the study, supported data interpretation, critically revised the manuscript and gave final approval of the version to be submitted. HS conceived and designed the study, acquired the funding for the study, supported data interpretation, critically revised the manuscript and gave final approval of the version to be submitted.

CO N S E NT FO R PU B LI C ATI O N
All patients gave written consent for the use of anonymized quotations from the mirror meetings in presentations and publications arising from the research.

I N FO R M E D CO N S E NT A N D PATI E NT D E TA I L S
I confirm all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details of the story. Onderzoek]). Therefore, it was not necessary for the CMO region Arnhem-Nijmegen or another recognized review committee to make a positive assessment. The participants signed an informed consent form to be included in this study.

DATA AVA I L A B I L I T Y
Data may be available on reasonable request from the principal investigator (SG). Mirror meeting data are not publicly available as they contain information that could compromise research participant's privacy and consent. The mirror meeting manual is momentarily only available in Dutch. On request from the principal investigator, an English version can be provided.

Box 3 Additional findings: Practical issues of mirror meetings
• The maximum duration of a MM for both patients and HCPs is 45 minutes.
• The minimum distance between the inner and outer circle is two metres.
• Set a quiet environment without many distractions and ensure a smooth and organized approach. Provide enough food and beverages, a pleasant temperature, a warm welcome, etc.
• Be aware that you are dealing with frail patients. They may cancel at any time, even on the day of the meeting.
• Pay attention to audibility, including hearing aids and dental prostheses, and consider microphones.
• The moderator should be someone who knows (of) the community, uses factual examples, probably meets the patients beforehand, but is independent of the (organization of the) professionals involved.
• Some patients prefer to see the faces of their HCPs. Make an informed choice about where to position patients and carers.