Integrating Patient‐reported Experience (PRE) in a multistage approach to study access to health services for women with chronic illness and migration experience

Abstract Background Patient‐reported Experience (PRE) is an emerging concept integrating patient perspectives and amplifying voices often marginalized in discussions surrounding health systems. However, it remains a challenge to use and integrate PREs when studying patient agency and access to quality services, particularly with data from multiple sources. In this article, using study materials from the Swiss MIWOCA project, we present and reflect upon a multistage PRE approach to study healthcare access. Methods The MIWOCA project, a study on healthcare access and quality among immigrant women with chronic illnesses living in Switzerland, provided data from multiple sources for the integration of PRE data. These sources included interviews with women (n = 48), two focus group discussions with women (n = 15), interviews with service providers (n = 12) and observations from stakeholder dialogues (n = 3). In addition, we utilized field notes, focus group illustration maps, patient vignettes and policy briefs to develop a multistage data linking model. PRE data served as starting themes and reference topics in each of the interlinked stages of knowledge production. Results Deploying PREs, we coherently linked the data from preceding stages and used them to inform subsequent stages. This, in turn, enabled us to identify, reflect and rectify factors limiting immigrant women's agency and access to quality services. Ultimately, the approach engaged patients as knowledge co‐producers for system‐level changes. This knowledge was transformed into a set of practice recommendations and a policy brief addressing ways to improve health systems to better serve immigrant women in Switzerland. Conclusions Building on PREs to systematically combine multiple data sources and engage patients continuously can improve our understanding of barriers in health systems. Beyond individual patient‐doctor encounters, a multistage PRE approach can identify structural problems and provide clues for resolving them at the systems level. The PREs approach presented may serve as an example and encourage more public health experts to consider PREs in future research and practice. Patient and Public Contribution Women with chronic illness and immigration experience contributed to interview‐guideline development, provided PREs in interviews, identified priority areas for health‐service change and actively participated in the development of practice recommendations.

level. The PREs approach presented may serve as an example and encourage more public health experts to consider PREs in future research and practice.
Patient and Public Contribution: Women with chronic illness and immigration experience contributed to interview-guideline development, provided PREs in interviews, identified priority areas for health-service change and actively participated in the development of practice recommendations. access to and quality of care. 1,2 PREs are particularly valuable when considering interactions of various dimensions of care that can include (but are not limited to) patient satisfaction, patient perception, patient preferences and patient engagement. 3 Although broad in their scope, most current PRE approaches focus on their potential for improving the individual healthcare experience in clinical settings, during or within a patient's encounters with services. 4,5 PREs are still in their early stages of conceptual development and remain a topic of wide discourse amongst health researchers and professionals. Many issues are currently on the research agenda, including the appropriateness and effectiveness of current methods in creating and implementing patient-reported experience measures (PREMs). [6][7][8] Currently, most studies exploring PREs/PREMs use quantitative approaches. This is particularly true for leading health systems that have implemented PREs/PREMs into service evaluations; primarily in the form of practice-specific questionnaires. 9,10 However, using questionnaires to capture PREs has its limitations, especially in that survey methods and data can be difficult to administer and interpret for healthcare staff, which acts as a significant barrier to their wider use. 11 Moreover, standardized quantitative methods are limited in addressing complex issues related to patients' utilization patterns.
This includes questions on how system features affect patients' perception, knowledge and ultimately their behaviours. 12,13 A particular challenge for PRE approaches is addressing social inequalities stemming from structural disadvantages. Many health system evaluations demonstrate that patients' access to high-quality healthcare services varies significantly by their social backgrounds (i.e., gender, race, migration status, socioeconomic status, etc.). A 2019 report on social inequalities in health systems found that across all OECD countries, Indigenous peoples and ethnic minorities were more likely to face socioeconomic disadvantages, language problems, cultural barriers and discrimination-all of which increase the likelihood of experiencing health disparities. 14 Qualitative PRE approaches may better capture these issues, particularly those of intersectionality, which is important to understand in patient groups experiencing multiple disadvantages. 15,16 Moreover, as these population groups also often face a higher burden of disease, their involvement in working towards improving services is even more crucial. 17 Given this background, the present paper demonstrates how a PRE approach can be used to link various data sources addressing healthcare system shortcomings. In doing so, we draw on data and insights from an empirical research study (Migrant Women's Health Care Needs for Chronic Illness Services in Switzerland [MIWOCA]) on immigrant women's healthcare needs and access to chronic illness care. 18 In the current paper, we illustrate how a multistage qualitative methodology can integrate patients' experiences in consecutive steps of the knowledge-production process and engage patients as coproducers of knowledge for improving care services.

| METHODS
This is a sub-study of MIWOCA, short for Migrant Women's Health Care Needs for Chronic Illness Services in Switzerland (SNF NRP74 2017-2020), a larger research project in which we researched access to and quality of healthcare service among women in Switzerland with chronic illness and migration experience. 18,19 MIWOCA included migrant and native-born women with chronic illnesses across diverse cultural and social backgrounds as well as care providers and other relevant stakeholders. Women participated in interviews and focus group discussions (FGDs). Subsequently, they were invited to participate in a series of stakeholder dialogues convened to develop policy recommendations. Detailed information about the study population, sampling strategies and data collection has been described previously. 18 In the current paper, we present a methodological approach to linking data sources based on the empirical fieldwork conducted during the MIWOCA project. We used data from multiple sources: observations and field notes from regular project meetings and three stakeholder dialogues, an analysis of project documents (such as minutes, reports, notes, patient vignettes, MIWOCA evidence and policy briefs), as well as a re-analysis of qualitative semi-structured interviews with women with chronic illnesses and health/social service providers (n = 48; n = 12) and two FGDs (n = 15). A survey among study participants provided auxiliary data on participants' perception and assessment of the multistage approach.
To develop our data linking model, we analysed field notes, documents and interview and focus group data, the latter ones had been analysed using the framework method, 20 supported by Atlas.ti software. We applied the following steps: transcription, familiarization with the data, coding and categorizing, identifying themes, developing a working analytical framework, applying an analytical framework, charting the data into the framework matrix and interpreting the data. We used both inductive and deductive approaches to help create an analytical framework. Throughout the data analysis process, we held regular peer debriefings among coauthors to discuss and refine findings and to interpret the meanings of data.

| RESULTS
In the following, we present (1) the multistage approach from MIWOCA that integrates PRE data, and thus patient involvement, at each step. We also present (2) selected examples of specific PREs on access to and quality of healthcare services. We will use those concrete examples to demonstrate how PREs were continuously reflected and considered throughout the multistage methodology to ultimately generate recommendations for improving access to and quality of healthcare services for this patient population. In addition to that, we present (3) participants' perceptions of the multistage PRE approach developed here. We used findings from data analysis in different formats: PREs from Stage 1 were used as substantive findings that were analysed and reported (e.g., identified barriers to accessing particular healthcare services [lack of system knowledge, stigma, etc.]). PREs from Stages 1 F I G U R E 1 Data flow in MIWOCA. Six data streams linking four different sources. In streams a-c, PREs directly informed the next step in data production and analysis. In streams d-f, PREs merged with other data to inform the next step in data production. MIWOCA, Migrant Women's Health Care Needs for Chronic Illness Services in Switzerland; PRE, Patient-reported Experience. and 3 were also used to develop complementary formats, namely 'vignettes' and 'focus group illustration maps'. 21 These formats aimed to represent 'thick descriptions' 22 of patients' experiences and were presented and discussed as part of the stakeholder dialogues in Stage 4. Each stage combined additional and different kinds of data. As a consequence, the increased richness and complexity of this data facilitated data triangulation and advanced analyses. Likewise, findings from each stage served to inform the content and focus of the next stage. For example, parts of the analysis of interviews with patients were used to help create the interview guide for interviews with providers. In principle, insights from PREs were carried forward to each mode of data production and analysis. In integrating data from all stages, we identified barriers and resources to accessing care (e.g., women's unfamiliarity with the complex healthcare system; informal social support networks). These were confirmed when merging the different sources, including other patients, care providers and thirdparty stakeholders (e.g., insurance experts). Through applying this model, we were able to gain insights on (1)

| PREs identified from interviews with patients
PREs often addressed women's knowledge of the health system's available health services and how to access those services, health insurance options and issues of how to mobilize social support. Those issues were identified as key elements for a patient's ability to make choices and use health services effectively thus, facilitating their agency.
Here, the patient agency is defined as the ability of patients to influence and contribute to the decision-making process behind their care. 23 Through PREs, women interviewed provided rich context information yielding a better understanding of issues such as limited health knowledge among certain population subgroups or lack of financial resources needed to utilize the healthcare services. In regard to the former, women described how their perspectives relate to both sides of the provider-patient relationship, specifically in the context of receiving and understanding medical information. On the one hand, patients critically evaluate the quality of information providers offer. On the other hand, they also reflect on their positions as receivers of this information. Patients highlight issues such as the lack of information, lack of clarity of the information or lack of details.
For example, one woman notes her disapproval: It's the 8th day now, but how aware am I [about the medication]. How much information was given to me?
They didn't provide information about anything; this is used for these diseases; this is used for these people … do you really want it 100% or not … they didn't say that.
Apart from direct interaction between medical service providers and patients and its impact on knowledge of medication, the women

| PREs applied in the providers' interviews
To facilitate the utilization of different data sources and further explore those issues around information deficits, language problems and access to care, researchers included the following questions in the interview guide for interviews with providers (Stage 2): (1) Do you think the current healthcare system works equally well for (i) women compared to men with chronic illnesses?
(ii) For Swiss women compared to foreign-born women?
(iii) For women with lower versus higher educational attainment?
(2) Are there any consequences of those differences?
Providers echoed not only that immigrant women with chronic illnesses have less information on how the health system operates, but also that this negatively impacts the ways in which they access care. For example, one provider stated:

| PREs applied in the FGDs
Researchers synthesized and analysed initial findings from Stages 1 and 2 in a short report and presented this to immigrant women (2) more support and counselling on the day-to-day struggles of managing a chronic illness, (3) and early access to user-friendly information about health insurance, particularly policies and contracts.

| Study participants' perspectives on the multistage approach
Study participants varied in their perception of the multistage approach to integrating PREs. Participating women tended to express a great level of enthusiasm for engagement as well as gratitude 'for being really heard' as expressed by one woman. At the same time, some women discontinued participation or voiced disappointment about the scope of the project over time. They had hoped to receive individually tailored recommendations for their own health and context, concrete advice concerning their own situation rather than to contribute to a system-level discussion and recommendations.
Participating service providers stated that it is crucial but nevertheless not very common in their contexts to involve people concerned from the beginning and in all phases of the research. participants as well as beneficiaries of trusted health systems'. 25 Moreover, the use of PREs in advancing people-centred care will allow institutions to identify problems with the delivery of care, implement new changes and interventions based on patient feedback and might promote the transparency and accountability of healthcare providers. 26 As researchers in the current study integrated perspectives from various stakeholders, this did not come without limitations.
Although patient engagement was present in every stage of the knowledge-building process, patients' personal participation declined while moving from identification of care deficits to analysing the data and drafting recommendations in stakeholder dialogues. In addition, women who were rather fluent in German tended to be overrepresented when it comes to continuous participation, despite translation services offered at meetings and dialogues. At the same time, the women who continuously participated in the MIWOCA study had been in close interaction with other patients through the group discussions and perceived themselves as representatives, particularly for the ones who might not have felt ready to speak up. Additionally, while stakeholder dialogues included representatives from a wide variety of sectors, future studies might consider including patients' families and caregivers as they can provide unique perspectives on relevant topics.

| CONCLUSION
Integrating patients' experiences proved a useful approach in the current system-oriented research on access to health services. Applying a multiple-stages process that featured PRE data allowed us to successfully link different sources and formats of data. This integrative PRE approach nurtured a process of co-production allowing to engage chronic disease patients in each step of the research process, from the identification of problems to the development of recommendations to mitigate them. While the focus was on women with a migration background in Switzerland, the methodology presented may facilitate future PRE studies reaching beyond the themes and contexts addressed here. The method presented may thus serve as an example and encourage more public health experts to consider a PRE approach for patients' involvement in health systems research.

AUTHOR CONTRIBUTIONS
Thomas Abel defined the main research question, developed the structure of the paper, wrote major parts of the text and lead the author team. Lidya Tadesse conducted the literature search and its integration into the text, contributed to creating a focus for the argument and wrote major parts of the text. Annika Frahsa contributed to the development of the argument, analysed and interpreted data sources and wrote major parts of the methodology description. Sibel Sakarya helped in developing the focus and the structure of the paper. All authors contributed to the revision of the manuscript. All authors read and approved the final manuscript.