What Matters to Patients? A Mixed Method Study of the Importance and Consideration of Patient Demands

This study aims to: (i) understand how different types of demands expressed by patients are taken into account in the current delivery systems operated by Health Care Organisations (HCOs); (ii) explore the often overlooked content of specic non-clinical demands (i.e. demands related to interactions between disease treatments and everyday life).

HCOs must increasingly address non-clinical, lifestyle-related patient demands along the entire care pathway.
These non-clinical demands are considered important by oncology patients but are not adequately taken into account by HCOs.
There are 5 types of non-clinical demands that could be better addressed for oncology patients.
The incorporation of these demands in the healthcare structure could be inspired by examples from other service sectors.

Background
The increasing number of patients affected by chronic conditions leads healthcare organisations (HCOs), from primary care clinics to integrated health systems, to rede ne the process and scope of their delivery system. One key feature is to be more patient-centred, i.e. by giving organisational responses to all demands expressed by patients along the entire care pathway and their potential environment (e.g. hospitalisation, ambulatory care, nursing home and home) [1].
The incorporation of all patient demands into the healthcare delivery system poses a challenge.
Healthcare delivery systems are mainly designed around diseases and other clinical conditions [2][3][4].
Healthcare professionals traditionally translate patient demands into clinical needs based on their medical knowledge (one such example is personalized medicine where the clinical needs of patients, particularly in terms of therapy, are de ned by integrating the molecular and genetic characteristics of the patient) [5][6][7]. However, other "non-clinical demands" exist. For instance, the social determinants of a patient (e.g. social isolation or nancial barriers to care access) can lead to demands being addressed during management of the care pathway (e.g. transportation, home meal delivery services). The integration of these psychosocial needs is a current issue for healthcare delivery systems, particularly for vulnerable patients whose social situation impacts their clinical condition. This can lead to identify high need-high cost patients [8]. In addition, patients also express demands during the care process according to their preferences-i.e. ideas, expectations, values-, which Patient Centred Care (PCC) experts [9][10][11], have often claimed to be ignored in care delivery systems. In conclusion, patients express demands related to aspects of their lives that interact with the management of the disease. This can be help with administrative procedures or questions regarding adaptation of transport conditions, and a whole set of services generally poorly apprehended during care, which can be linked to the eld of hospitality management [12].The total lack of global consideration for these various types of demands can result in a fragmented healthcare system and ine cient use of meagre resources [3,13]. In order to initiate the design and implementation of a demand-driven organisational model, it is necessary to consider and study the content of all non-clinical demands and their association with clinical demands, an area that has hitherto been investigated to less of an extent [14,15].
The purpose of this article is twofold: (i) It initially aims to assess the importance of each demand expressed by patients and the extent to which these demands are taken into account in managing their care pathway. Two categories of demands are de ned in support of the analysis: clinical (C), and non-clinical (NC) demands and it is assumed that, as a general rule, greater consideration is given to clinical demands as opposed to nonclinical demands.
(ii) The article then seeks to investigate the content of speci c non-clinical demands in greater depth.
We use the concept of "demands" rather than "needs" or "unmet needs", two concepts that are often di cult to de ne, depending on the context in which they are applied and the disciplines involved [16]. The notion of "demands" departs from these concepts in two ways. Firstly, we concentrate on demands perceived as opposed to expressed, normative and comparative concepts of need [17]. Secondly, demands point to a lack of well-being which might (but need not) indicate an unmet need for care. It follows that the concept of demand does not involve an express need for care. Demands are only a starting point for looking at what patient personally perceive as di culties and complaints [18].
This research focuses on two Health Care Organisations (HCOs) specialising in oncology. Cancer is a signi cant example of chronic care, illustrating the need to consider various patient demands. Indeed, cancer patients who are increasingly exposed to chronic conditions are involved in care pathways, which include numerous return trips between home and the health care establishment. They are also experiencing stressful changes in their general health throughout the care continuum. This involves regular updates in the management of their care pathways and ongoing consideration of the interaction between clinical and nonclinical demands. In addition, the two selected HCOs have developed innovative patient paths based on remote patient monitoring systems, including nurse navigators (NNs). The principal role of NNs is to co-ordinate patient management activities. As pathway ambassadors, they are in direct contact with patients before, during and after periods of hospitalisation. They must, therefore, pay close attention to all expressed and non-expressed patient demands along the entire care pathway. In so doing, they provide interesting areas of exploration for understanding which demands patients can express along their care pathway and how these are taken into account in the HCO healthcare delivery system.
This study have been driven by an inductive perspective: How patient demands of all kinds are integrated in an organizational process? The rst step served to to understand and enhancethe importance of all demands (clinical and non-clinical), then see how these demands are incorporated into current modern health care delivery systems. We illustrate the usefulness of this analysis for delivering care and services at patient pathway level, providing additional insight into consideration of non-clinical demands in the patient-provider relationship during the clinical decision-making process [19,20] . The results of the rst step show the importance of considering non-clinical demands in HCO's. It prompts us to explore the content of these demands. What is the speci c nature of non-clinical demands that could be taken in consideration in the management of care pathways? The second step was a deep empirical identi cation and description of non-clinical demands. It revealed a variety of demands, including information about transportation from home to HCOs, athletic and sporting activities during a treatment phase, and compliance with religious values.

Setting
We developed our study in speci c HCOs, namely two cancer centres. HCO 1 is a major cancer centre. In the year of the study (2017), HCO 1 had 414 beds, 14,600 admissions, 6,133 consultations and 3,314 patients were included in the navigation programme. HCO 2 is a cancer centre including a teaching hospital in western France. In the year of the study, HCO 2 had 103 beds, 7,612 admissions and 2,867 consultations.

Research design
We adopted a mixed method comprising two phases. Based on exploratory research involving HCO1 and HCO2 to equal extent, we designed a quantitative approach (Phase 1) using a questionnaire to address the rst objective (how patient demands are taken into account by healthcare organisations). With regard to the second objective, namely the content of non-clinical demands, we adopted a qualitative analysis (semi-structured interviews) (Phase 2). We adopted a "demand driven" perspective as opposed to "supply driven" one. We started with the evaluation of the patient demands before de ning the organizational response, during both phases. In phase 1, HCO 1 is the main eld and HCO 2 the eld for con rmatory analysis, whereas in phase 2, HCO 1 is the unique eld of exploration. Phase 1. Quantitative phase: analysis of the importance and consideration of patient demands In phase 1, we assessed the importance of each demand and the extent to which patients perceived that their demand has been taken into account.

Questionnaire Design
Based on exploratory interviews and observations of interaction of oncology NNs with patients in both HCOs, and our experience as eld experts , we collectively identi ed 27 generic patient demands classi ed into two types, clinical and non-clinical. For each identi ed generic demand (except the one relating to "anticipated directive" that could not be evaluated), we formulated one question to evaluate the perceived importance and one question to evaluate how it has been taken into account in the HCO. Consequently, the questionnaire comprised 51 items that assess the importance and consideration of each demand based on an 8-item Likert scale regarding patient characteristics (age, household, income, professional status, SPC (social professional category), education, follow-up period, satisfaction with lifestyle), and ultimately, an open-ended question for potential additional comments (a full version of the questionnaire is appended). Then the questionnaire was pretested on two patients.

Sample
In order to boost the signi cance of the outcome of the questionnaire, which investigates importance and consideration regardless of condition, we decided to focus this study phase on a speci c population: breast cancer patients in a chronic situation. We applied the following inclusion criteria: (i) women aged 18 and over; residing in metropolitan France; (ii) life expectancy of over 6 months con rmed by health professionals, already monitored in the HCO; (iii) participant in one of the innovative follow-up programmes. The following exclusion criteria were applied: (i) non-French speaking patients; (ii) patients who have not yet received cancer treatment (just diagnosed with cancer).
Administration of the questionnaire On the main site (HCO 1), the selfadministered questionnaire was given to patients attending their consultation from 09/05/2017 to 23/05/2017, by professionals. after checking the inclusion criteria. A ballot box was provided for patients to submit their completed anonymous questionnaires.
On the second site (HCO 2), the questionnaires were given to day hospital, radiology and consultation personnel to be self-administered by patients under the same baseline conditions from 01/06/2017 to 30/09/2017.

Statistical analysis
Descriptive analysis has been carried out for each HCO. Answers relating to importance and consideration were processed as quantitative scores (from 0 to 7 points). Analyses were carried out with R 3.4.3 software.

Sampling
In order to specify in depth the type of various non-clinical demands, which were identi ed and evaluated in the previous phase, we conducted semi-structured interviews with patients presenting all types of cancer in order to understand the common content of these demands under different conditions. We carried out this study in HCO 1, applying the same inclusion and exclusion criteria as in Phase 1, from January to March 2018 (KB) and from June to July 2019 (MW).

Data collection
Potential participants were recruited on site by KB and then by MW in the hall of HCO 1. After introducing the study, an appointment was made, mostly for the same day. Semi structured interviews were conducted in a dedicated room in HCO 1 (duration: 1hour on average). They were based on an interview grid listing the previous phase 1 non-clinical demands, with a focus on challenging aspects for the management of care pathways Discussions remained very open. At times it took a very narrative turn, as the patient started to describe his pathways from the beginning to the day of the interview. The interviewer asked for him/her to give precisions and examples concerning the different needs encountered at each step of the process.

Date analysis
The 36 interviews were recorded and transcribed in full. Data analysis followed an inductive process. This included data familiarisation, open coding and de nition of categories of demands. These categories emerged collectively as KB, MW, EM read the interviews. As Each of the three researchers started to code interviews, The codes were discussed during monthly meeting, allowing to distinguish ve categories of demands expressed by patients, and 18 types of needs per patient. Data collection ceased when no new categories were identi ed.
They were analysed by three researchers (KB, EM, and MW).

Phase 1. Analysis of Importance/Consideration
In HCO 1, the survey covering 26 demands was made available to 613 patients until the full circulation of the 550 printed questionnaires. 114 patients refused to answer or did not comply with the inclusion criteria and did not keep the questionnaire (patients undergoing screening, non-French speakers, etc.). Overall, 499 questionnaires were completed and handed in.
In HCO 2, 74 questionnaires out of 130 distributed to health professionals were completed and handed in: 44 for radiology, 16 for the day hospital and 14 for consultations.

Characteristics of respondents
Patients on both sites were similar in terms of average age (approximately 55 years of age) and size of household (number of adults, number of dependent children). Most of them were couples without CMU (Universal Health Insurance Cover). There were, however, signi cant differences between the two sites (as HCO 1 patients were generally monitored for longer), the level of quali cation (HCO 1 patients were slightly more quali ed), professional status (a large proportion of HCO 2 patients were on sick leave) and SPC (social-professional category) (more executives and intellectual professions in HCO 1) (see Appendix 2).

Comparison between importance and consideration perceived
As a general rule, considerable importance is given to all demands in both HCOs (average: 6.14). Six types of demands nevertheless still appear to be more important than others: adaptation of treatment in line with patient requirements (6.8), pain management (6.7), professional courtesy (6.7), patient information in terms of treatment choice and changes (6.7), respect for intimacy (6.6), the adaptation of paramedical care (6.6) and access to cutting-edge treatments (6.6). These response pro les are similar for both sites (Figures 1 and 2). One demand is considered less important on both sites: consideration of the patient's beliefs (4.7). However, it is also the demand that varies the most in terms of importance, depending on the patients concerned (see Figure 3).
In contrast, patients' opinions regarding the way in which these demands were taken into consideration during their treatment varies to greater extent. Demands which scored highly in this respect are: staff courtesy (6.5), the adaptation of treatment in line with patient requirements (6.3), respect for intimacy (6.3), the fact that professionals provide contact details (6.2) and pain management (6,1). HCO 2 patients mostly found that the adaptation of paramedical care (6.4) and timing adjustments in line with patient requirements (6.1) were given appropriate consideration.
Less consideration was deemed to be given to the following requirements: knowledge of tariff options (fees, procedures, accommodation) during hospital stays (3.0), access to legal assistance (3.0), access to administrative assistance (2.9) or social support (2.9), advance knowledge of levels of reimbursement (2.7) and advance availability of tariffs for therapeutic procedures (2.5). A certain variability in responses is noted, which is greater for non-clinical demands particularly when they involve very personal questions relating to care values (e.g. NC13, see Fig. 3).

Results Phase 2. In-depth understanding of the content of the non-clinical demands
The patients interviewed during the third phase were mostly women (21/36) between 32 and 72 years of age. The most common type of cancer was gynaecological cancer (14/36) with most patients presenting breast cancer (12/14). A large number of patients received oral chemotherapy (24/36).
These interviews allowed us to classify non-clinical demands according to 5 categories determined by the nature of the demand in response to a speci c requirement (see Table 1) : 1. Demands related to lifestyle during the treatment period, 2. Demands related to alternative medicine requirements and improved wellbeing, excluding prescriptions issued by healthcare professionals, 3. Demands related to the organisational aspect of the treatment pathway 4. Demands for administrative and logistic assistance, 5.
Demands related to the use of new information and communication technologies (NTIC) and to telemedicine. NC demands selected for the questionnaire were included in n°1, n°3, n°4. N°2 and n°5 emerged from phase 2.

Discussion
Our study highlights two main points.
Firstly, it shows that non-clinical demands expressed by patients actually exist and in some cases are given less consideration than clinical demands in healthcare delivery systems. This nding is consistent with a large body of literature on patient experiences and PCC studying the quality of care from a patient perspective since 1985 [21,22]. They show the importance of developing tools that can assess the patient experience rather than his/her satisfaction, and highlight the Importance of considering needs of nonclinical nature (eg. information about waiting and process time) [23]. This study brings two added values in this context: (i) It proposes an empirical view of these non-clinical demands expressed by patients. Some of these demands may seem very similar to clinical demands such as the use of alternative medicines, but the observation reports nuances as they are self-managed by patients, without professional involvement.
Others represent aspects of everyday life with a chronic disease, a scenario frequently experienced by patients, and a far cry from patient care and clinical treatments issues. Moreover, it highlights the need to consider demands along the entire patient pathway. Some of these NC demands are largely beyond a patient-physician relationship such as transportation, or pet-sitting. Other NC demands emphasise the need to manage this physician-patient relationship through new and remote follow-up. In view of these various perspectives, the content of these NC demands gives fresh insight into the traditional concept of biopsychosocial needs. This latter, de ned as the motivation for achieving a satisfactory level of functioning as a human being [24], mainly focuses on care aspects and less on issues relating to everyday life in a disease scenario. More speci cally, these demands also consolidate the Patient-Centred care approach by starting from the patient experience of health care delivery schedules [25].
(ii) Our results also suggest a variation in importance and experience regarding each demand depending on the patient concerned. This applies to some NC demands in particular (NC3-4-12-17, Fig. 3). It stresses the need for customised answers according to the combination of C and NC demand expressed by each patient [2,13].
Secondly, and consequently, the need to consider all clinical and non-clinical demands via a customised approach questions the role of HCOs in delivering more patient-centred and demand-driven delivery systems. Furthermore, the content of these demands shows that they must be considered along the entire patient pathway, and not only in the provider-patient relationship.
This resonates with well-known approaches in other service sectors, such as hospitality management, de ning appropriate practices during the customer relationship in order to respond to their demands [26][27][28], and mass customisation [29,30] or personalisation to scale [31,32] proposing speci c answers for each customer at affordable cost. However, important challenges have to be overcome if such approaches are to be applied to healthcare. With speci c regard to the transfer of hospitality management techniques, the professional mandate must be extended beyond clinical practices. Although the structured response to customers' demands of all types is clearly aligned in the healthcare sector [33][34][35], these pioneering experiences remain isolated, calling for a review of the healthcare professionals' mandate. Some non-clinical demands are not taken into account appropriately. Logically, the distinction between those demands would seem to correspond to different healthcare priorities.
However, our results highlight nuances in terms of this assumption. Some of the non-clinical demands such as "courtesy", "intimacy" and "identi cation" appear to be taken into account in HCOs. These NC demands are not connected to clinical demands, but they imply the direct relationship with caregivers who generally believed these practices to be aligned with their clinical "mandate" [36,37]. The boundaries for considering such practices within the healthcare spectrum have already been studied [38,39], and resonate with the conventional debate between care and cure [40]. Modern activities highlight the need to develop practices more in line with these patient demands.
Development of a customised care model also requires consideration of segmentation analysis methods to incorporate clinical and non-clinical demands from a common as opposed to a primarily diseasebased perspective [2,41]. But less explicitly, it also requires answers to speci c organisational questions: from fabrication (i.e. ensuring that production meets each demand), to assembly (i.e. "modular" service structures capable of combining in a exible manner the multiple products and services required by the combination of demands), and distribution (i.e. co-ordination and integration of various assembly structures to ensure the timely delivery of a given service to the patient) [29,[42][43][44]. The fabrication of non-clinical demands can generate new services such as concierge services in HCOs [33,34,45]. Although some attempts have been made to incorporate speci c "fabrication" steps or to combine services in a "modular package" [46], there is no integrative system as far as we are aware. The transition from a supply-oriented delivery system to a demand-driven approach would require further investigations in this direction.
This study has some limitations.
1. To produce questionnaires that seem, according to our experience in HCO1, not too long, we have had to merge a certain number of situations into one main criterion (e.g. intimacy). This has led to a certain number of differences being overlooked. These differences were highlighted in conjunction with phase 2 patient interviews.
2. The study took place in two hospitals in France. This may not seem enough to assess the taking into consideration of patient demands how patient by healthcare organization. In fact, the perception patients have on how diverse demands (including demands concerning information about organizational characteristics) are taken into consideration depends more on healthcare teams in the numerous wards concerned than on the central administration of HCOs.
3. In the rst step, Non-clinical demands have been studied in in a speci c case, namely breast cancer. The focus made the questionnaire more signi cant. However, some non-clinical demands can be diseasespeci c, limiting then the scope of the results. Although it represents an exemplary case in which the patient potentially expresses numerous requirements, it is worthy of being investigated in other contexts. That is why we included patients suffering of all types of cancer in the second step of the study 4. Lastly, it may seem counter-intuitive to specify the content of the demands (qualitative phase) after the questionnaire. In our case, this is linked to the research strategy adopted, namely to address, rst, the consideration of different demands. The observation that numerous NC demands are not taken into account has led to a more in-depth study of their content. Our approach highlights the need to uctuate between qualitative and quantitative research from a patient perspective.

Conclusions
Despite these limitations, as far as we know, this is the rst empirical study that has examined the consideration of various patient demands in healthcare delivery systems. Our typology of ve non-clinical demands is also an initial step to providing a better understanding of various patient preferences and demands beyond the clinical sphere. Given our focus on oncology, we encourage researchers to examine this issue in relation to other chronic conditions such as diabetes or cardiovascular diseases. The nascent and heterogeneous state of knowledge in this area is also challenging in order to make summative ndings at this point based on this single research project. However, our study highlights the fact that invisible demands should be identi ed more clearly and integrated in a segmentation analysis of patient needs and individualised responses in order to develop new demand-driven and patient-centred approaches. The data sets used and analysed in this study are available from the corresponding author on reasonable request.

Competing interests
The authors have no competing interests to declare.

Funding
This work was supported by the "Fondation de l'Avenir". The funder had no role in the study; Neither in the writing of the manuscript; or the decision to submit this manuscript for publication.
Authors' contribution MW and EM designed the study. MW EM and BL designed the survey with the help of MXA. BL provided the statistical analysis. MW, EM and KB designed the qualitative inquiry and analysed qualitative data.
MW and KB performed the qualitative inquiry. MW drafted the article with the contribution of EM and BL. All authors approved the nal manuscript.
HCO1 Importance and consideration of demands perceived by HCO 1 patients. CF. List of the 26 demands, appendix 1. Score allocated to each demand. Cf. List of the 26 demands appendix 1