Co-ordinated care for people affected by rare diseases: the CONCORD mixed-methods study

What does care co-ordination mean?

The focus of this study is the co-ordination of care for people affected by rare diseases. A systematic review conducted in 2007 reported no single agreed definition of co-ordinated care and proposed the following working definition:

Care co-ordination is the deliberate organisation of patient care activities between two or more participants (including the patient) involved in a patient’s care to facilitate the appropriate delivery of health care services. Organising care involves the marshalling of personnel and other resources needed to carry out all required patient care activities and is often managed by the exchange of information among participants responsible for different aspects of care. 1

As part of our research, we sought to improve this definition in the context of rare diseases.

What is the problem being addressed by this research?

There are an estimated 6172 unique rare conditions. 2 Each rare disease affects fewer than 1 in 2000 of the population,3 but combined they affect a large number of people, approximately 3.5 million in the UK4 and 30 million in the European Union. 3 The problem being addressed by this research is the variation in how care is co-ordinated for people affected by rare diseases in the UK (and in many cases the complete lack of care co-ordination), depending on where they live and the disease they are affected by.

Rare diseases are often serious, chronic and complex in nature, affecting multiple systems of the body. As a result, several health-care professionals (HCPs) are often involved in patients’ care. Many people need access to a number of different NHS services to receive the care they need, including care by specialists and care nearer to home. Care by specialists may require travelling long distances and staying away from home, which can be inconvenient, costly and stressful. Care nearer to home may involve care by the local hospital or general practitioner (GP). Receiving care from a range of people, including specialists and local providers, can cause problems because co-ordination between the different professionals and services is often poor, care plans may not be in place or followed and some patients may have gaps in their care because they do not see the right professionals and, when they do, the information to facilitate appropriate care may not be to hand. Parents/carers of children with rare conditions often face a significant care burden, needing time off work to look after their children and take them to appointments. There can also be challenges in ensuring continuity of care when children transition from paediatric to adult services.

Why is this research needed?

There is evidence to suggest that care is poorly co-ordinated for people affected by rare diseases. In addition, improving care co-ordination for people affected by rare diseases has been raised as a major concern by policy-makers.

In January 2016, Rare Disease UK (London, UK) published results from a survey of more than 1200 people (patients and carers) affected by rare diseases, which found that information on test and procedure results and treatment was not shared effectively between services, meaning that patients may have received suboptimal treatment. 5 The survey also found that patients and families frequently had to attend multiple clinics and travel significant distances to reach them. For example, one in three respondents had to attend three or more clinics and 12% of respondents attended more than five different clinics. Respondents attended clinics monthly (23%), every 6–8 weeks (32%), quarterly (55%) or at least once a year (92%). 5

In addition, not only did patients have to frequently visit multiple clinics, but nearly half the survey respondents reported that they travelled for more than 1 hour to get to their furthest clinic, with 11% of respondents reporting that they had to travel for more than 3 hours. 5 The survey found that 81% of patients did not have a care co-ordinator or advisor, and a further 8% of patients were unsure whether or not they did. The survey also found that 40% of respondents did not know if there was a specialist centre for their condition. Of the patients who were aware of a specialist centre for their condition, only 66% used it. 5 These data illustrate the heavy burden that poor care co-ordination places on patients and families dealing with rare diseases, which could be improved by better co-ordination.

In September 2016, Genetic Alliance UK (London, UK) undertook a study to identify the hidden costs of rare diseases in the UK. 6 The aims of the study were to examine how services are co-ordinated for patients with rare diseases, what is known about the impact of the lack of co-ordinated care, what costs and outcomes are important to patients and families, and how these data might best be collected. The study involved interviews with patients, families and patient organisations. The main conclusions were that receiving co-ordinated care is important for patients with rare diseases, yet remains a challenge; the full costs and benefits associated with different models of care for patients with rare disease are unknown; patients and families face significant hidden costs, both financial and psychosocial, associated with the way their care is managed; and there are limitations associated with existing research and data sets for rare diseases.

The problem of poor co-ordination of care for patients with rare diseases has also been highlighted by the UK governments, although the evidence base is largely anecdotal. In 2013, the Department of Health and Social Care, Northern Ireland Executive, Scottish Government and the National Assembly for Wales published The UK Strategy for Rare Diseases,7 which said that it was essential to co-ordinate care for people with rare diseases. The strategy also stated that more needed to be done to improve co-ordination and that research was needed on how care for people with rare diseases should be co-ordinated. In the progress report from the All Party Parliamentary Group on rare, genetic and undiagnosed conditions it was noted that care continues to be badly co-ordinated. 8

More recently, the UK government, and patients and families, further highlighted the problem of co-ordinated care for people affected by rare diseases in The UK Rare Diseases Framework. 9 The UK Rare Diseases Framework9 restated that co-ordination of care was one of the top challenges facing people affected by rare diseases and better co-ordination was listed as one of the four top priorities. In addition, better co-ordination was also listed as one of the four major challenges facing the rare diseases community. In a ‘national conversation’ survey of 6293 members of the UK rare diseases community, conducted in 2019, co-ordination of care was identified as the top challenge by 16% of patients, 19% of families and carers, 11% of rare disease patient organisations and 18% of HCPs (Table 1). 9 We note that, although co-ordination of care was noted as a major challenge in its own right, any improvement in co-ordination is also likely to have a positive impact on the other challenges (e.g. by improving diagnosis, awareness and access).

Unfortunately, although there are indications that care needs to be better co-ordinated for people affected by rare diseases, there is not good evidence as to how this should be achieved. A 2013 report by Rare Disease UK10 provided anecdotal evidence of the benefits of having a named care co-ordinator and concluded that there was a strong case for investment in care co-ordinator posts, although quantitative evidence was lacking. Van Groenendael et al. 11 analysed the national service for an ultra-rare disease (Alström syndrome) and compared outcome and cost of the service with standard care. 11 Van Groenendael et al. 11 found that organised multidisciplinary ‘one-stop’ clinics achieved better outcomes than standard care, at similar costs. Indicative of the lack of evidence about how to improve care co-ordination for people affected by rare diseases, The UK Strategy for Rare Diseases7 called for further research in this area, in particular around how care for people with rare diseases is co-ordinated and how best it ought to be co-ordinated. Our study aimed to address these gaps.

Aims

The aims of this study were to use quantitative and qualitative research methods to investigate (1) if, and how, care of people with rare diseases is co-ordinated in the UK and (2) if, and how, patients and families affected by rare diseases, and HCPs who treat rare diseases, would like care to be co-ordinated.

Objectives

• To undertake a scoping review to identify what ‘co-ordinated care’ means, what the components of co-ordinated care are and to identify in what ways, and why, co-ordinated care for people with rare diseases might be similar to or different from co-ordinated care for people with other conditions.

• To understand if and how care of people with rare diseases is co-ordinated in the UK.

• To analyse preferences for different models of co-ordinated care by patients, families and HCPs.

• To develop a taxonomy describing how care for people with rare diseases could be co-ordinated.

• To calculate the costs of the models of co-ordinated care identified in the taxonomy.

• To work closely with patients and families throughout the project and disseminate findings widely.

Research question 1

What does ‘co-ordinated care’ mean, what are the components of co-ordinated care and in what ways, and why, may co-ordinated care for people with rare diseases be similar to or different from co-ordinated care for people with other conditions?

Research question 2

Is care for people with rare diseases in the UK co-ordinated and, if so, how?

Research question 3

What are the preferences of patients, families and HCPs in relation to how care for rare diseases is co-ordinated?

Research question 4

What are the different ways in which care for people with rare diseases might be co-ordinated?

Research question 5

How much do these options cost?

Our study was interested in exploring all spectrums of co-ordination (from a lack of co-ordination through to good co-ordination).

The study was funded by the National Institute for Health Research (NIHR) Health and Social Care Delivery Research programme from June 2018 to February 2021. The CONCORD (Co-ordinated Care Of Rare Diseases) study timeline is summarised in Figure 1. For RQ1, we undertook a scoping review (not only in rare diseases) that focused on care co-ordination across organisational boundaries and interventions employed to support and improve this. For RQs 2 and 3, we created a questionnaire-based survey of current experiences and costs, incorporating a discrete choice experiment (DCE) of preferences for co-ordination. In addition, we undertook an exploratory qualitative interview study to understand the impact of a lack of co-ordinated care on patients and carers, and preferences for co-ordination. For RQ4, we drew on the findings of the scoping review and also carried out interviews, focus groups and workshops with a range of stakeholders to develop a taxonomy of co-ordinated care for rare diseases. For RQ5, we reviewed the costs of different components of co-ordinated care.

FIGURE 1.

The CONCORD study flow chart. PPI, patient and public involvement; PPIAG, Patient and Public Involvement Advisory Group.

Structure of the report

This report is structured as follows.

Chapter 2 presents the overarching design of the study and provides an overview of the methods employed. (Detailed information on methods is presented within each of the findings chapters.)

Chapter 3 presents the methods and results of a scoping review to define co-ordinated care for people living with rare conditions. Chapter 3 extends previous research by providing an updated definition of co-ordinated care for rare conditions, and by identifying and categorising components of co-ordination according to their role within complex care processes.

Chapter 4 presents an exploratory qualitative interview study of patients affected by rare conditions and their carers, exploring how these groups are affected by whether or not care is co-ordinated, and the factors that might influence effective care co-ordination.

Chapter 5 presents the findings from a national cross-sectional survey of patients, parents/carers and HCPs about different aspects of care co-ordination for rare diseases, including the use of specialist centres, care co-ordinators and care plans.

Chapter 6 presents the findings of a DCE to evaluate preferences of patients, parents/carers and HCPs for characteristics of co-ordinated care.

Chapter 7 outlines the development and refinement of a taxonomy of care co-ordination for people living with rare conditions. The taxonomy outlines the six domains involved in co-ordinating care for rare conditions.

Chapter 8 presents selected models of co-ordinated care from the taxonomy and illustrative costs of different components of co-ordinated care.

Chapter 9 presents a discussion of our findings linked to our RQs and the implications for health services and research.

Setting

This study is concerned with how people with rare conditions are cared for across organisational settings in the UK, including the NHS, the social care sector and the third sector. The primary focus for this study was NHS care, but we were also interested in providers that are gatekeepers to social care provision and third-sector care, as significant elements of co-ordination specifically relate to the integration between health care and these other sectors. To identify as many different models of co-ordination as possible, no limits were set on the rare conditions included or where people lived in the UK.

Overview of approaches

We used the following methods in our research:

• a scoping review (not only in rare diseases) that focused on care co-ordination across organisational boundaries, and interventions employed to support and improve this

• an exploratory qualitative interview study to understand the impact of a lack of co-ordinated care on patients and carers

• a questionnaire-based survey of current experiences, incorporating a DCE of preferences for co-ordination

• interviews, focus groups and workshops with a range of stakeholders to develop a taxonomy of co-ordinated care for rare diseases

• a review of the costs of providing co-ordinated care.

There were numerous interdependencies between the different components of the study (see Figure 1). The scoping review provided the theoretical underpinnings for the taxonomy of co-ordinated care and informed the content of the survey, the DCE and what is known about the costs of co-ordinated care. The exploratory qualitative interviews informed the scoping review, the survey and the DCE. The survey and DCE helped to identify different models of care co-ordination, which were used to create the taxonomy. In addition, the survey and the DCE were also intended to provide data for the cost analysis of the different co-ordination models, which were, in turn, based on the taxonomy (which delineated the options to be costed).

Study participants

Study participants comprised patients (aged ≥ 18 years) affected by a rare condition, parents/carers (aged ≥ 18 years) of children or adults with rare conditions, HCPs (e.g. doctors, nurses and allied health professionals) involved in the care of people with rare conditions, national leads on specialist health-care commissioning, national patient groups and charities, and local providers and commissioners of co-ordinated care. More specifically, participants were involved in various elements of the research as follows:

• To find out if the scoping review findings applied to rare conditions and to support the design of the survey and DCE, we undertook three focus groups. The focus groups were as follows:

  • one virtual focus group with seven patients and carers affected by rare diseases

  • one face-to-face focus group with six patients and carers affected by rare diseases

  • one face-to-face focus group with four HCPs.

• To explore the impact of unco-ordinated care, and to support the design of the survey and DCE, we conducted interviews with 15 patients and carers affected by rare diseases [14 interviews were via telephone and one interview was via Skype™ (Microsoft Corporation, Redmond, WA, USA)].

• To ensure that the survey was worded appropriately and contained appropriate questions, we undertook a pilot study of the survey and DCE questionnaire with 24 patients, carers and HCPs. The study comprised four think-aloud interviews and 20 interviews that provided written or verbal feedback.

• For the survey and DCE, we obtained 1457 responses from 760 patients affected by rare diseases, 446 parents/carers and 251 HCPs.

• To develop the taxonomy, we planned to undertake up to 30 national and local stakeholder interviews. These interviews included national leads on specialist health-care commissioning, national patient groups and charities, local providers of co-ordinated care (including health care, social care and the voluntary sector) and local commissioners of co-ordinated care.

• To develop the taxonomy, we undertook 30 interviews with HCPs, charity representatives and commissioners, and four focus groups involving a total of 22 patients and carers affected by rare diseases.

• To refine the taxonomy, we conducted two workshops with 15 attendees each. Workshop participants included adult patients (aged ≥ 18 years) and carers of adult patients, carers of younger patients (aged < 18 years), care providers (including health care, social care and the voluntary sector) for adults with rare conditions, care providers (including health services, social services bridging health and social care and the voluntary sector) for children with rare conditions, and commissioners of co-ordinated care provision, including NHS England and local authorities.

Participants were accessed via patient and provider networks and organisations.

Scoping review

The scoping review was designed to help us understand what aspects of co-ordinated care could or should be provided for people with rare conditions, and help us build on what was already known about co-ordinated care in other contexts that might be used to enhance co-ordinated care for rare conditions. The scoping review had six stages. 12

Survey

We conducted a national survey to understand how care of people with rare conditions was co-ordinated in the UK. The questionnaire incorporated a DCE to quantify what aspects of care co-ordination participants preferred.

The content of the questionnaire was informed by 15 semistructured qualitative interviews with patients and carers to identify costs associated with living with rare conditions. These interviews were also used for the exploratory qualitative study to investigate the impact on patients and carers of having care that was not co-ordinated.

Survey participants were adult patients (aged ≥ 18 years) affected by a rare condition, parents/carers (aged ≥ 18 years) of children or adults with rare conditions and HCPs (e.g. doctors, nurses and allied health professionals) involved in the care of people with rare conditions. The target number of responses for each of these three groups was at least 300, with an overall target sample size of 1500. Participants were required to live in the UK, but there were no restrictions in terms of rare condition, demographic factors or geographical location within the UK.

Participants were accessed through patient networks and organisations, and through care providers and regional genetics services.

We produced a draft of the questionnaire based on the outputs of the in-depth interviews and scoping review (including the focus groups). This was reviewed by the PPIAG and amended accordingly. We then piloted the survey and made amendments according to feedback received. The survey was then finalised in discussion with the PPIAG.

The questionnaire covered a variety of topics, including experience of diagnosis, rare condition, availability/role of care co-ordinators, content and use of care plans, availability/role of specialist centres, use of health services and perceived impact of care co-ordination on quality of care.

A survey company generated online, electronic and hard-copy versions of the questionnaire ready for circulation. Most respondents completed the questionnaire via a weblink to the online questionnaire, which was made available on a dedicated website. Participants were also given options to complete the survey by mailed hard copy, electronically by e-mail or by telephone.

Analyses of the data were descriptive. The results of the categorical, ordinal and interval questions were reported as frequencies and percentages, or means and medians, with corresponding measures of spread [e.g. confidence intervals (CIs) or interquartile ranges].

For further details about the methods employed, see Chapter 5, Methods.

Discrete choice experiment

We undertook a DCE to investigate preferences for care co-ordination. 13 A DCE is a quantitative method used to elicit preferences from participants without directly asking them to state their preferred options. 14 The DCE formed one part of the survey questionnaire, eliciting preferences for the way in which care is co-ordinated for the three participant groups. A longlist of attributes was drawn from the scoping review and was shortened to six attributes based on feedback from the interviews, focus groups and the PPIAG. The levels of each of the attributes were based on feasible ranges derived from reviews of documentary evidence from the scoping review and feedback from the PPIAG and interviews. The DCE used a pairwise choice framework, describing combinations of levels and attributes of different models of co-ordinated care, including main effects only. We reduced the total number of feasible pairwise choice questions to 18, which were split into three blocks of six (i.e. each participant completed six choice questions). We also asked respondents to provide a simple ranking of the attributes according to importance.

The DCE data were analysed using conditional logit analyses. We selected this type of regression model given our focus on identifying which attributes significantly affect preferences, and which attributes are most and least important to respondents, conditional on the other attributes in the analysis. We tested for differences in preferences between responder groups. We calculated marginal rates of substitution (MRSs) with respect to costs, dividing the coefficient for each attribute by the coefficient for the cost attribute to calculate the ‘willingness to pay’ for each attribute. We also calculated the predicted probability that different combinations of the attribute levels would be selected, allowing us to rank different models of co-ordinated care in terms of their order of preference by the participants.

For further details about the methods employed, see Chapter 6, Methods.

Taxonomy of models of co-ordinated care

We developed and refined a taxonomy (classification) of different models describing how care for people with rare conditions could be co-ordinated. To do this, we conducted interviews and focus groups with stakeholders to derive a draft taxonomy. The sampling framework was designed to capture experience with different models of co-ordinated care. We aimed to conduct up to 30 interviews with national leads on specialist health-care commissioning, national patient groups and charities, and local providers and commissioners of co-ordinated care. We also conducted four focus groups with patients and carers. We then ran a series of workshops to discuss the draft taxonomy. We aimed to run up to five workshops, each with up to 20 attendees. Owing to COVID-19, we ended up amending the study to include two virtual workshops (with up to 15 attendees each), instead of five face-to-face workshops. To recruit for the interviews, focus groups and workshops, we used a range of methods, including e-mail invitation, social media, voluntary sector recruitment and recruitment via our partnerships with four NHS sites.

The interviews and focus groups used topic guides that focused on key aspects of care co-ordination, including use of specialist clinics, information-sharing between specialist and local services, transition from child to adult services, implications of co-ordination on clinic attendance and travel distances, and influential factors affecting the ability to provide co-ordinated care. The sessions were digitally recorded and professionally transcribed. Iterative and thematic analysis of all data were undertaken concurrently, accounting for outputs from the scoping review, the survey and the DCE. To develop the taxonomy, a combination of inductive and deductive thematic analysis was used (see Chapter 7 for more information).

The resulting draft taxonomy was tested in consensus-building workshops, which aimed to produce recommendations about the taxonomy (see Chapter 7 for more information). A final taxonomy was developed based on workshop feedback.

For further details about the methods employed, see Chapter 7, Methods.

Cost of co-ordinated care

We used the findings from the taxonomy to develop hypothetical models of co-ordinated care (see Chapter 8, Methods, for further details of the development process). These hypothetical models give an example of what co-ordinated care may need to look like in different situations. We aimed to calculate the costs of these models of co-ordinated care using data from both the national survey (see Chapter 5) and the workshops used to refine the taxonomy (see Chapter 7). Unfortunately, it was not possible to use either of these sources. In the case of the survey, most people did not experience co-ordinated care. In addition, it was not possible to attribute the hypothetical models that were developed to survey respondents. In the case of the workshops, the health service utilisation associated with each hypothetical model was unknown by workshop participants, primarily because this was likely to vary according to situation-specific factors. The result was that it was not possible to generate costs associated with each hypothetical model from the survey data or workshop data. Instead, we undertook a review of the costs of different components of co-ordinated care to illustrate indicative costs. For further details, see Chapter 8, Methods.

What was already known?

• Co-ordinating care may be beneficial to patients and carers, as it may ease treatment burden.

• Many terms and definitions have previously been used to refer to co-ordination of care (mostly for common chronic conditions).

What this chapter adds

• This chapter extends previous research by providing an updated definition of co-ordinated care for rare conditions.

• This chapter extends previous research by identifying and categorising components of co-ordination according to their role within complex care processes.

• This chapter highlights similarities and differences between co-ordination for common and rare conditions (i.e. many of the components apply to both common and rare conditions, but that there are additional components and context-specific issues that are relevant for rare conditions).

Stages of the scoping review

We followed these six stages to complete the scoping review.

Stage 5: collated, summarised and reported results

Narrative synthesis was used28 to develop definitions and identify and group components. A second researcher grouped 10% of components independently. To develop a definition of co-ordinated care, we coded individual definitions inductively. Codes were grouped by one researcher and used to develop a preliminary definition, which was reviewed and amended by the wider research team. Components were coded and grouped by one researcher. Examples of groups included ‘planning’, ‘methods of co-ordination’ and ‘approaches of co-ordination’. A second researcher double-coded 10% of components into groups. Disagreements were discussed and resolved. Groups of components were developed into themes and subthemes, and the number of reviews that reported each theme, subtheme and component was recorded. The themes were (1) care pathway (i.e. components that related to the care pathway), (2) approaches (i.e. components relating to care/co-ordination approaches), (3) support (i.e. components relating to support), (4) features (i.e. components relating to features of care) and (5) wider environment. Each theme had a number of subthemes that each contained multiple components. Once themes and subthemes of components had been developed, individual components were then reviewed and categorised into four types of components (Figure 2). The wider research team also reviewed and agreed on the categorisation of components.

FIGURE 2.

Categorisation of components of co-ordinated care. This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Review characteristics

We included 154 review papers22,29–181 (see Report Supplementary Material 2 for characteristics). Figure 3 outlines the review selection process. Common chronic conditions were reviewed in 139 reviews. Only three reviews focused on a single rare condition and 12 reviews focused on both rare and common chronic conditions.

FIGURE 3.

The study selection process (based on Moher et al. 182). This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Stakeholder consultation characteristics

Stakeholder consultation participant characteristics are shown in Table 3.

What does co-ordinated care mean for rare conditions?

Many terms and definitions were used to describe co-ordinated care (see Appendix 2 for the terms and definitions used). Stakeholder consultation findings indicated that terms and definitions were relevant for rare conditions, with some aspects emphasised [e.g. communication, expertise and multidisciplinary teams (MDTs)] or new aspects highlighted (e.g. the importance of care co-ordination being delivered equitably across geographical areas, individualisation, importance of the whole family and need to co-ordinate across a person’s whole lifetime).

From these findings, we developed a definition of co-ordinated care for rare conditions:

Co-ordination of care involves working together across multiple components and processes of care to enable everyone involved in a patient’s care (including a team of health care professionals, the patient and/or carer and their family) to avoid duplication and achieve shared outcomes, throughout a person’s whole life, across all parts of the health and care system, including: care from different health care services . . . care from different health care settings . . . care across multiple conditions or single conditions that affect multiple parts of the body, the movement from one service, or setting to another. Co-ordination of care should be family-centred, holistic (including a patient’s medical, psychosocial, educational and vocational needs), evidence-based, with equal access to co-ordinated care irrespective of diagnosis, patient circumstances and geographical location.

Reproduced from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/

What are the components of co-ordinated care for rare conditions?

We identified many components of care co-ordination. Figure 4 provides a summary of components identified through the review and stakeholder consultation findings. Throughout Results, we give examples of components and the percentage of reviews (n = 154) that reported each component (see Appendix 3). We also present findings from the stakeholder consultation (see Appendix 4). In this chapter, we briefly describe the components in relation to the review findings and stakeholder consultation findings. For a more in-depth analysis of these components, please refer to the published manuscript. 15

FIGURE 4.

Summary of components of care co-ordination (from review and stakeholder consultation findings). This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

This figure is adapted from Walton et al. 15 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Components indicating ‘what’ care pathway components need to be co-ordinated

Scoping review findings indicated that components relating to administration, assessment and diagnosis, planning, review and evaluation, feedback, follow-up care and technology were frequently reported (development of care plans, 61.7%; follow-up care, 52.6%; monitoring, 51.3%). Stakeholder consultation findings highlighted that although these components are important, patients and carers do not always receive these components in practice (e.g. a lack of care plans and regular reviews were identified):

I’m thinking of review and evaluation with the therapist and on in-community. So, for people with rare conditions, there’ll be a period of intervention, then all will go quiet, everything’s being managed well, and then some other problem will come up, but the way therapists work is that there’ll be a period of intervention, measured outcome, closed case, no contact kept until a crisis down the road and they come back, and when I was in that position, I always wanted to be able to keep that person under review [. . .] because you were then managing a situation before it became a crisis.

FG-HCP

Scoping review findings also highlighted many support components that need to be co-ordinated, including support for patients, carers and families (education/skills training for patients, 73.4%; self-management support, 54.6%) and support for HCPs (education, 32.5%; training, 31.2%). Stakeholder consultation findings highlighted that support for patients and families across a range of needs (including medical, psychological, practical, emotional and social) and from various people (including health-care providers, peers, schools and patient support groups) is important. Findings also highlighted that support for HCPs (to access specialist knowledge and to address fears and anxieties) is necessary when co-ordinating care for rare conditions. Despite the importance of support, patients reported a lack of support and information provision for rare and undiagnosed conditions.

Components indicating ‘how’ care can be co-ordinated

Five main groups of components that outline ‘how’ care can be co-ordinated were identified. There is potential overlap between some of these components, and they are not mutually exclusive.

Components that contextualise co-ordination

Scoping review findings show that evidence-based practice (guideline-based treatment, 37%; evidence-based treatment protocols, 35.1%), individual differences and the wider health-care environment (access to care, 31.2%) may influence co-ordination. Stakeholder findings concurred with scoping review findings, but highlighted a lack of care pathways and defined standards for rare conditions and emphasised that treatments are not delivered consistently (where standards are available):

. . . it’s, having those clear pathways . . . and having something that people can, you know, work towards which is really clear . . . Even though it’s very rare, it’s, like, ‘OK, this is the process now’, and that’s really important . . . And I think that makes the systems work a lot better if there is something like that in place. I think when it’s wishy-washy or it’s not clear, or there is no clear, kind of, guidance or pathways, and because some of the situations are quite, you know, specialised, I think it’s difficult, it’s very difficult to manage.

FG-HCP

Participants reported having to travel to access care and that they were happy to do so if it meant that they received expert care. One of the key issues preventing patients with rare conditions from accessing care was a perceived limited availability of HCPs with expertise in their condition. To take limited expertise in each condition into account, findings also indicated the need to succession plan by training more experts.

Can the definitions and components of care co-ordination from common chronic conditions be applied to rare conditions?

Stakeholder consultation findings indicated that components identified through the scoping review were comprehensive and relevant for rare conditions. Although our findings highlighted that the components identified in the scoping review are relevant when co-ordinating care for people with rare conditions, we found that patients experienced a lack of co-ordination (e.g. having to attend multiple appointments on different days, gaps and delays in information-sharing and disagreements between professionals). There were particular concerns around emergencies and the necessity for patients to take control themselves to mitigate worries:

But why are we living in a world where we have to pick one or two or three of these? Why can’t it be all of them?

FG-PC1, patient

Certain factors may make it more difficult to co-ordinate care for rare conditions (e.g. difficulties in diagnosing rare conditions due to a lack of knowledge and ability to recognise symptoms and limited condition-specific expertise due to small numbers of patients).

Participants expressed views that some components were missing or may need to be emphasised for rare conditions. These included having someone to take responsibility for co-ordination, genome-based medicine/genetic screening, social support needs, counselling, and antenatal and bereavement care. In addition, participants expressed views that more focus should be given to undiagnosed patients and families for many of the components.

Key findings

Co-ordinated care for people affected by rare conditions requires working together across multiple components and processes of care to ensure that everyone involved achieves shared outcomes throughout a person’s whole life and different parts of the health and care system. Our definition encompasses the idea that for rare conditions co-ordinated care should be family centred, evidence based and equitable for all.

These findings suggest that many of the key components and issues for co-ordinated care apply to both rare and common chronic conditions. Stakeholder consultation findings revealed additional components and context-specific issues that are relevant in the context of rare conditions.

How findings relate to previous research

Our findings extend previous research1,117 by developing a definition of care co-ordination for rare conditions.

Previous research proposed that it is difficult to distinguish between aspects of care and co-ordination components. 17,183 Our findings extend previous research, as we have grouped components according to their roles, situating components within the wider health-care pathway and environment. Our four categories overlap with those outlined in previous research,1 but clarify how co-ordination components may be involved in complex care processes.

Our review highlights that little is currently known about co-ordination for rare conditions (as most of the reviews focused on common chronic conditions). Despite similarities in need, stakeholder findings suggest that co-ordination for rare conditions may be less consistent in practice,5,8 as many of the components identified were not delivered effectively or consistently for people with rare conditions, let alone co-ordinated. These differences may be attributed to complexities associated with rare conditions (e.g. that rare conditions affect multiple body parts, children, may be lifelong, need to be co-ordinated across multiple sectors and bring complexities of diagnosis due to limited expertise). These complexities suggest that more care co-ordination is needed in cases of greater system fragmentation, clinical complexity and decreased patient capacity. 1

Limitations

We used inclusive definitions for co-ordination and chronic conditions, and reviews covered a range of countries. We are unlikely to have captured every relevant review. To identify as many studies as possible, we conducted a comprehensive search, which included contacting experts and searching the reference lists of included reviews.

It is possible that the reviews that we included in this research may not have captured all evaluations of care co-ordination in practice. Therefore, publication bias may be present. However, we used many approaches to minimise this risk, including expert consultation and stakeholder consultations.

Components reported in published descriptions of interventions do not necessarily equate to the delivery of all components and, therefore, this review is limited to the components reported in the reviews. In addition, individual studies may be included in more than one of our included reviews. We based our analysis on the wording reported in the review papers and not individual studies.

This review focused on the identification of co-ordination components, rather than testing the effectiveness of co-ordination. The effectiveness of individual components, or combinations thereof, on relevant outcomes (e.g. reduced waiting times, better health-care outcomes and better experience) is not known.

Although the scoping review provides insight into the components that are involved in co-ordination, given the small number of stakeholders included in the scoping review, it was not always possible to provide further specifics regarding whether or not different components of co-ordination may be suitable for different conditions and situations (e.g. regarding the findings on who should take responsibility for co-ordination). However, this is something that is explored in our survey, DCE and taxonomy development work (see Chapters 5–7).

Implications

Our findings provide support for various international policy initiatives. 7,19,24 We identified many components within our review that are reported as necessary for co-ordination. 7 Our findings also show that delivering co-ordinated care is complex because there are many different options for co-ordinating care. Our findings emphasised the need for someone to take responsibility for co-ordination, as outlined in the NHS Implementation Plan,19 but highlight that there are many ways in which responsibility could be managed.

Researchers and clinicians could use the components to begin to develop and evaluate existing and new models of co-ordination for common and rare chronic conditions.

Future research

This review has highlighted different components of care co-ordination. Future research could test and evaluate the clinical effectiveness, implementation and cost-effectiveness of these components in practice. This could be achieved, for example, by identifying current use of these components and evaluating them empirically, either retrospectively or prospectively.

What was already known?

• Care co-ordination is considered important for patients with rare conditions.

• Research addressing the impact of care co-ordination on patients and carers with experience of rare diseases is limited.

What this chapter adds

• This chapter explores how care co-ordination (or lack of) has an impact on patients and carers.

• Unco-ordinated care results in delays and barriers to accessing care and places a burden on patients and carers, which, in turn, has negative effects on patients and carers in terms of physical health, financial and psychosocial impacts.

• Approaches to co-ordination that improve access to care and lessen the time and burden placed on patients and carers may be particularly beneficial.

Experiences of unco-ordinated care for patients with rare conditions

Participants described a range of experiences of unco-ordinated care (Figure 5, section A), which included the following:

• unco-ordinated appointments [i.e. unnecessary frequent appointments to see different professionals/services across different NHS settings, some of which were located far from home (i.e. at specialist centres), lack of choice about when or where their appointments took place and appointments with one professional at a time (with little evidence of medical and non-medical services being offered in the same clinic)]

• ineffective communication between professionals and between professionals and patients/carers [i.e. lack of communication/team approach across various care professionals (particularly between those in specialist centres and local teams), no point of contact to approach with queries, problems relating to information-sharing (particularly the timeliness of information-sharing) and limited use of care plans]

• patients and carers co-ordinating their own care (i.e. patients and carers undertaking a number of tasks, including chasing services, holding information and facilitating information-sharing, with many reporting being the main co-ordinator of care).

Experiences of unco-ordinated care varied across individuals and stages of the patient journey. For example, there were challenges associated with emergencies or acute episodes (with a lack of awareness locally and difficulties accessing timely treatment), establishing care and support pre and post diagnosis, following discharge from hospital (and receiving the appropriate care within the community) and transitioning from paediatric to adult services.

FIGURE 5.

Patients’ and carers’ experiences of unco-ordinated care and its impact. This figure is reproduced from Simpson et al. 184 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

This figure is reproduced from Simpson et al. 184 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

How unco-ordinated care impacts on patients and carers

Unco-ordinated care resulted in delays and barriers to accessing care and placed an additional burden on patients and carers (see Figure 5, section B). These delays and barriers, in turn, had several negative impacts for patients and carers (see Figure 5, section C).

Examples of co-ordinated care and approaches to reduce the negative impacts of imperfect care co-ordination on patients and carers

From transcripts, we extracted data on examples of good co-ordination (currently experienced by participants) and/or participants’ suggestions for how co-ordination could be improved (Table 4).

Key findings

The study identified two key consequences of unco-ordinated care on the patient and carer experience, that is delays and barriers in accessing care and additional time/burden. A range of impacts on patients and parents/carers were identified and categorised into three overarching themes: physical health, psychosocial impacts and financial impacts. The findings also suggest ways in which service users felt negative impacts of unco-ordinated care might be reduced, including with the support of a professional co-ordinator, using MDTs, care plans, technology and/or a point of contact to improve communication, and organising appointments to meet the needs of patients and families (by providing appointments locally or virtually, where possible, offering a range of services in one visit and scheduling appointments at a convenient time for the family).

The findings demonstrated that different levels of care co-ordination (e.g. the involvement of a care co-ordinator and how clinics/appointments are delivered) may be needed at different stages of the patient’s journey and/or to meet individual patient preferences. This highlights the need for co-ordination to take into account the characteristics and preferences of the patient when new services or support are required (e.g. post diagnosis or when transitioning to adult services) and when care is needed locally in non-specialist settings. There may also be key individual differences that affect the extent to which one might experience the consequences and impacts described here. Some impacts may be magnified for some. For example, co-ordinating care may be more burdensome for a single parent without a wider support network. Such differences also need to be taken into account when considering how care should be co-ordinated.

How do the findings relate to previous research?

Our data support previous findings from others on the burden of living with a rare and/or undiagnosed condition,186,187 including struggling to access appropriate expertise, support and information, financial strain (e.g. due to changes in employment) and psychosocial impacts (e.g. disruption to everyday life, living with uncertainty and stress). However, this study offers a novel contribution to the existing knowledge base. It has identified the specific costs (financial and otherwise) and potential benefits associated with how care is co-ordinated, rather than the more general consequences and needs associated with living with or caring for someone with a rare condition.

Previous research shows that patients come up against a number of challenges within the health-care system due to the rarity and complexity of their condition, including misdiagnoses and delays in diagnosis, a lack of information and support, and low availability of treatments. 189 This study suggests that unco-ordinated care also contributes to the challenges faced by patients with rare diseases, particularly around accessing care and the burden of managing the condition.

Although this chapter presents the impacts across three overarching themes (physical health, psychosocial and financial), it is important to note the inter-relations between them. This is supported by previous research,187 which found that parents faced financial challenges as a result of their caring role, which, in turn, was a major source of stress. Another recent study highlighted a number of factors affecting the mental health of rare disease patients and their carers, including trying to access services and support (e.g. financial and non-medical support). 190 Although it is helpful to categorise different impacts (e.g. for identifying measures for evaluation and further research), the complex interdependencies found in reality should be recognised.

The findings support recent research,186,191 which has found that patients and carers take on significant tasks in relation to care co-ordination and management. Having professional support to co-ordinate care could reduce the negative impact on patients and carers (and others). However, this support could take many different forms and further research and consultation is required to establish the extent to which responsibility should be transferred from patient to professional.

Limitations

The main limitation is that only 15 participants took part in what was an exploratory qualitative study. Although efforts were made to include a variety of individuals and experiences, the sample is not representative of all those affected by rare diseases. As an exploratory and qualitative study, the intention was not to recruit a representative sample of the rare disease population, but rather to focus on in-depth individual accounts that could support the development of data collection tools and interpretation of quantitative data in the CONCORD study. The data can be used to provide an insight into an under-researched area and to inform future research (e.g. on strategies to reduce the burdens described in this chapter). In addition, although positive examples of care co-ordination were shared, it was challenges and gaps in care co-ordination that were more commonly felt among participants.

In addition, there may be other factors (not included here) that have an impact on patient and carers experience of care co-ordination. There may be institutional barriers (e.g. the availability of resources), structural barriers (e.g. the limitations of information systems within NHS organisations) and/or cultural obstacles to change.

Future research

This study identified how care could be co-ordinated in a way that benefits patients with rare conditions, and their carers. Further work is required to develop models of care co-ordination (e.g. as considered in Chapter 7), assess their feasibility and then evaluate them in practice to determine if the potential benefits we have identified can be realised.

Our findings demonstrate the need for future research to collect the views of larger numbers of patients (and carers) and consider what might influence differences in preferences. This research should also be extended to gather the views of professionals (i.e. those delivering care and supporting patients). We consider this further in the following chapter.

What was already known?

• Few studies have examined if care for people living with rare diseases is co-ordinated.

• There are several ways in which care could be co-ordinated for people affected by rare diseases, including through the use of care co-ordinators, specialist centres and care plans.

What this chapter adds

• We undertook a national survey involving 760 adult patients affected by rare diseases, 446 parents/carers of people affected by rare diseases and 251 HCPs who care for people affected by rare diseases to understand the extent to which, and how, care of people with rare conditions is co-ordinated in the UK.

• There is limited access to co-ordinated care for people affected by rare diseases. For example, only 12% of adult patients affected by a rare disease reported that they had a formal care co-ordinator, 32% reported that they attended a specialist centre and 10% reported that they had a care plan.

Survey instrument

A survey questionnaire was developed using data from three sources to identify elements of care co-ordination that should be explored. First, we identified themes from the scoping review of 154 reviews of co-ordinated care for rare and chronic conditions (see Chapter 3) to identify important components of co-ordinated care. Second, we ran three focus groups involving (1) patients aged ≥ 18 years affected by a rare condition, (2) parents/carers of children and adults affected by a rare condition and (3) HCPs involved in the treatment of rare conditions. One focus group was conducted virtually with four patients and three carers and two focus groups were conducted face to face (one with four HCPs and the other with two patients and four parents/carers). Third, we ran 15 one-to-one telephone or Skype interviews with seven patients and eight parents/carers. Using the findings from these activities, we identified three key areas of care co-ordination that mattered to patients and families: (1) access to care co-ordinators, (2) specialist centres and (3) care plans. We developed a first draft of the questionnaire to understand how care of people with rare conditions is co-ordinated in the UK, with specific reference to these items. The study PPIAG reviewed the draft questionnaire and modified the language and content, as necessary. The questionnaire was then piloted. Twenty respondents completed the questionnaire on their own using a draft information sheet to instruct them in the process. In addition, four ‘think-aloud’ interviews were undertaken with one patient, two parents and a HCP, who completed the draft questionnaire in the presence of a researcher. The questionnaire was modified according to the pilot data and the PPIAG performed a further review to the ensure that the language used was comprehensible and relevant to the intended focus of each question.

The survey instrument mainly contained close-ended questions with defined response categories. A smaller number of questions asked participants to provide qualitative or text information (e.g. the rare condition they were affected by) and an open-ended text box was provided for these questions. At the conclusion of each survey section, participants were also given the opportunity to provide any other comments they had on the section topic.

The first section of the survey was used to obtain consent and to determine participant eligibility. Participants were provided with a participant information sheet about the survey, which included the purpose of the survey, the organisations involved in conducting the study and assurances around anonymity and aggregation of data for reporting purposes. Participants were asked to click ‘next’, taking them to another webpage to access the survey, and were advised that by doing so they were consenting to participate in the survey. Participants were also told that they did not have to take part if they did not want to. As the survey was completed anonymously, after data were submitted by clicking the ‘submit’ button at the end of the survey it was not possible to withdraw individual respondents’ data.

The next section of the survey asked about experience of diagnosis/rare conditions. The main body of the survey followed, with three sections measuring the three key aspects of co-ordination of care identified from the preparatory research, namely experiences of care co-ordinators, specialist centres and care plans. Additional questions were then asked around use of health services. The next section included a DCE (see Chapter 6). The final section asked participants for information on socio-demographic factors. See Report Supplementary Material 1 for the patient version of the survey.

Survey sampling

Three groups of participants were eligible to complete the survey: (1) patients (aged ≥ 18 years) affected by a rare condition, (2) parents/carers (aged ≥ 18 years) of children or adults with rare conditions and (3) HCPs (e.g. doctors, nurses and allied health professionals) involved in the care of people with rare conditions. We aimed to recruit 300 participants for each group and to have an overall target sample size of 1500 participants. This was justified using two pieces of information. First, sample size calculations for surveys are possible based on population size, desired confidence level and maximum acceptable margin of error. Assuming a population size of upwards of 20,000 (predicted sample size remains close to constant for populations > 20,000), a margin of error of 3% and a confidence level of 95%, the required sample size is 1014 (calculated using SurveyMonkey^®; URL: www.surveymonkey.com/mp/sample-size-calculator; Momentive Inc., San Mateo, CA, USA). Second, our target figure of 1500 participants partly stemmed from another survey5 using a similar research design in the UK. A 2016 survey by Rare Disease UK5 achieved a sample size of 1213 participants.

There were no restrictions on participants in terms of the rare condition, demographic factors (other than age ≥ 18 years) or geographical location within the UK. We deliberately did not sample from specific rare diseases, nor limit the range of rare diseases we included, to identify as many different models of co-ordination as possible, and to include as broad a range of experiences and preferences with regard to care co-ordination as possible. A complete sample frame of all adults living with a rare condition in the UK does not exist. The total number of people living with a rare condition, their contact details and their sociodemographic characteristics, such as age, gender, highest education level and location of residence, are unknown. For these reasons, purposive snowball sampling was used for this study. We discussed routes to accessing patients and parents/carers with the PPIAG. Participants were accessed via patient and provider networks and organisations, including Rare Disease UK (which has more than 2000 registered supporters, including academics, clinicians, industry, individual members and patient organisations192), Genetic Alliance UK (a national alliance of organisations with a membership of more than 180 charities that support patients and families affected by genetic disorders193) and SWAN UK (a support network for families of children and young adults with undiagnosed genetic conditions in the UK run by Genetic Alliance UK194).

An independent survey company created an electronic version of the survey using a bespoke online platform. The survey was ‘live’ from August to December 2019. Potential participants were sent a weblink to the survey either by e-mail or via social media. The message containing the weblink also included an offer to send hard copies of the questionnaire by post or e-mail or to complete it verbally over the telephone with a researcher. We also recruited patients and parents/carers via six major care providers, where research co-ordinators at each site identified potential participants and asked participants if they were willing to participate in the study. If participants were willing to participate, they were provided with further details on how to do this, as described above. HCPs were recruited using the same routes described above for patients and parents/carers. In addition, we contacted the British Society of Genetic Medicine (London, UK) and its constituent organisations and special interest groups,195 and the NIHR Clinical Research Network: Genetics. 196 These organisations circulated details of the survey to their members via their electronic mailing lists. Participants had a 48-hour window where they were able to suspend completion of the questionnaire, if they needed to do so, and then to resume where they left off at a time that was convenient to them.

Analysis of data

All data handling was conducted in compliance with General Data Protection Regulation197 requirements and all responders agreed to their data being processed for research purposes. Responses where less than 20% of all data fields were completed were removed. All the quantitative results, except for those relating to the use of health services, were reported as frequencies and percentages using Microsoft Excel™ (Microsoft Corporation, Redmond, WA, USA) and Stata^® (StataCorp LP, College Station, TX, USA). For the use of services, we reported the mean values per patient, stratified by whether or not the patient had access to a care co-ordinator, a specialist centre and a care plan. We did not impute missing data. Responses were checked for any identifiable information entered in open-text boxes and redacted if necessary. All open-text data were exported by the survey company and formatted for analysis. Open-text responses specific to the topic of care co-ordinators, specialist centres and care plans were coded using inductive thematic analysis in Microsoft Excel. A list of all codes identified (per question) were summarised and were used to identify and develop cross-cutting themes and subthemes within the data. Below, we present the findings from our thematic analysis alongside the quantitative findings for the three main topics (i.e. care co-ordinators, specialist centres and care plans).

See Appendix 5 for a STROBE (STrengthening the Reporting of OBservational studies in Epidemiology) statement pertaining to the survey.

Responses and sample

In total, 1604 responses to the survey were received from 856 adult patients affected by rare diseases, 497 parents/carers of people affected by rare diseases and 251 HCPs who care for people affected by rare diseases. We excluded respondents who completed less than 20% of the data fields, which amounted to the exclusion of 96 (11%) patients and 51 (10%) carers. The final number of responses included for analysis was 1457 (adult patients affected by rare diseases, n = 760; parents/carers of people affected by rare diseases, n = 446; HCPs who care for people affected by rare diseases, n = 251). It was not possible to estimate a response rate for each group, as the survey was sent by multiple overlapping distribution routes using convenience sampling and snowball sampling techniques.

Among 760 (adult) patients with a rare condition, the modal age band was 45–54 years and more than 80% were female (Table 5). More than 95% of patients had been diagnosed with a rare disease (as opposed to being undiagnosed) and diagnoses had been confirmed by a genetic test in 30% of patients. Multiple body systems were affected, the most common being muscle, ligaments and joints (affecting 58% of the sample) and vision (affecting 57% of the sample). A total of 221 rare conditions were represented. The greatest number of responses for a single disease group was 101 (13%) for sarcoidosis. Most (57%) patients reported that they lived with a spouse or partner.

The modal age category of the 446 carers who responded was 35–44 years (35%), and in around two-thirds of cases the patient being cared for was aged < 18 years. More than 80% of carers who responded were female. Most of those cared for had been diagnosed with a rare disease (8% were undiagnosed) and in more than 60% of cases the diagnosis had been confirmed with a genetic test. A total of 259 rare conditions were represented in parents/carers, and the largest number of responses for a single disease group was 18 (4%) for Behçet’s syndrome. In more than 70% of cases the carer was the parent of the person affected by the rare condition, and in more than 90% of cases the carer lived with the they person cared for.

Around 90% of both patients and parents/carers were from the white ethnic group and both groups came from a range of educational backgrounds, with having a degree or higher degree being the modal education category. Respondents in both groups were from across the UK.

More than half of the 251 HCPs reported having specific clinical expertise in rare diseases, and they worked across a range of areas with patients with rare conditions. Around 30% of respondents were hospital doctors, 16% were nurses or midwives and 11% were allied health professionals. Respondents in this group worked across the UK.

Access to formal care co-ordinators

In the questionnaire, a formal care co-ordinator was defined as:

A professional with a recognised role in helping patients and carers manage a range of needs between different professionals or across care settings. They may be a full-time co-ordinator or may co-ordinate care as part of their main job, such as a GP.

Twelve per cent of patients reported that they had a formal care co-ordinator, and this figure was comparable to the proportion of parents/carers (14%) who reported that the person they cared for had a formal care co-ordinator (Table 6 and Figure 6). However, these figures were lower than for HCPs (35% of whom reported that the majority of patients had a formal care co-ordinator). Of the patients and parents/carers who reported having access to a formal care co-ordinator, 30–40% said that this was a person employed specifically for the care co-ordinator role. In addition, the care co-ordinator had another main role, which was usually a hospital doctor, a GP or a specialist nurse. Patients and parents/carers who reported having access to a formal care co-ordinator noted that a variety of roles were undertaken, the most common for both groups being liaising between professionals (75% of patients and 73% of parents/carers reported that their formal care co-ordinator carried out this role). Other common roles were scheduling appointments, liaising between HCPs and non-HCPs, updating the patient’s care plan and being the contact for emergency or acute episodes. When asked about the main factors that determine if someone with a rare condition has access to a formal care co-ordinator, the most common responses from HCPs were complexity of disease (49%) and the extent of the patient’s need for support (45%).

FIGURE 6.

Percentage of survey respondents with access to a formal care co-ordinator. Note that the figure reports the percentage of survey respondents replying ‘yes’, ‘no’ or ‘unsure’ to the question ‘Do you a have a formal care co-ordinator?’ for patients, ‘Does the person you care for have a formal care co-ordinator?’ for parents/carers and ‘Do the majority of your patients have a formal care co-ordinator?’ for HCPs.

All survey respondents were given a list of roles and asked if they would prefer each of these to be undertaken by the patient, carer or the formal care co-ordinator (Table 7). Although there was some variation in the magnitudes between groups, over half of patients, parents/carers and HCPs preferred appointments to be scheduled by the patient/carer. For all other roles (i.e. liaising between professionals, updating the care plan, ensuring availability of health records, co-ordinating transitions, arranging multidisciplinary clinics and arranging respite care), the majority of respondents in all three groups preferred these to be undertaken by the formal care co-ordinator.

Throughout the care co-ordinator section of the survey, respondents were asked to provide further information about their responses. In addition, at the end of the survey, respondents were asked if there was any other information they would like to provide that, specifically, related to care co-ordinators and, if so, to include this in a free-text box. The open-text qualitative information was summarised according to six overarching themes, which are described below. Issues relating to each of these themes were raised by all three respondent groups (i.e. patients, parents/carers and HCPs).

Access to specialist centres

In the survey, a specialist centre was defined as:

A centralised facility that enables patients to see a number of health-care professionals in one visit. Usually, the professionals at specialist centres will be experts in rare and undiagnosed conditions. Non-health-care professionals may also see patients at the same centre.

Thirty-nine per cent of patients reported that a specialist centre was available for their rare condition and 83% of these patients said that they attended the specialist centre. Therefore, 32% of patients attended a specialist centre for their condition (Table 8 and Figure 7). Among carers, the respective values were 37%, 88% and 33%. Sixty per cent of HCPs reported that a specialist centre was available for the majority of their patients with a rare condition. Specialist centres provided access to HCPs with specialist knowledge. More than half of the patients, parents/carers and HCPs reported that patients at the specialist centre saw doctors who were experts in rare or undiagnosed conditions, doctors who were experts in the aspects of health affected and specialist nurses. Other professionals were seen less frequently at the specialist centre. A wide range of services were provided by the specialist centre, with the most common being appointments with an expert in rare conditions and appointments to see different types of HCPs at the centre. When HCPs were asked the main reasons why patients with rare conditions might choose not to use specialist centres, the most common responses, each reported by more than 60% of respondents, related to the extent of travelling involved, including the distance, cost and physical difficulty of travelling.

FIGURE 7.

Percentage of survey respondents with access to a specialist centre. Note that the figure reports the percentage of survey respondents replying ‘yes’, ‘no’ or ‘unsure’ to the question ‘Is there a specialist centre available for you?’ for patients, ‘Is a specialist centre available for the person you care for?’ for parents/carers and ‘Is there a specialist centre available for the majority of your patients with rare conditions?’ for HCPs.

The free-text information about access to a specialist centre was summarised thematically into the six categories below. Issues relating to each of these themes were raised by all three respondent groups.

Access to care plans

In the questionnaire, a care plan was defined as:

A paper or electronic document which describes the health services and support that are needed and should be agreed between patients, carers and professionals. The care plan may be a single document or it may be part of another record which includes non-health services such as an Education, Health and Care Plan (EHCP).

Ten per cent of adult patients reported having a care plan relating to their rare condition, compared with 44% of parents/carers (Table 9 and Figure 8). Forty per cent of HCPs said that they use care plans to document care for patients with rare conditions. In terms of the responsibility for keeping the care plan up to date, patients reported that, most commonly, it was either themselves (27%) or a hospital doctor (27%) who was responsible for keeping the care plan up to date. Parents/carers, however, reported that, most commonly, it was their responsibility to keep the care plan up to date. HCPs advised that, most commonly, this responsibility was shared between professionals (25%). Sixty-four per cent of patients and 85% of parents/carers who had a care plan reported that they were involved in developing it. The most common items included in the care plan were general information about the patient, a medical summary and an assessment of current health needs. These items were also reported to be the most useful items to be included in the care plan by all three groups of respondents, along with a plan of care for emergency or acute episodes.

FIGURE 8.

Percentage of survey respondents with access to a care plan. Note that the figure reports the percentage of survey respondents replying ‘yes’, ‘no’ or ‘unsure’ to the question ‘Do you have a care plan relating to your rare condition?’ for patients, ‘Does the person you care for have a care plan relating to their rare condition?’ for parents/carers and ‘Do you use care plans as a means to document care for patients with rare conditions?’ for HCPs.

The open-text information responses about care plans were summarised thematically into the categories below. Issues relating to each of these themes were raised by all three respondent groups.

Access to combinations of care co-ordination

We showed that only 12% of adult patients affected by a rare disease reported that they had a formal care co-ordinator, 32% reported that they attended a specialist centre and 10% reported that they had a care plan. For parents/carers of patients affected by rare diseases, these figures were 14%, 33% and 44%, respectively. Table 10 shows access to different combinations of the three elements of care co-ordination covered by the survey. It shows that 54% of patients and 33% of parents/carers had access to neither a formal care co-ordinator, nor a care plan, nor a specialist centre (this was the modal combination). By contrast, only 2% of patients and 5% of parents/carers reported having access to all three elements.

Use of services

The mean use of health services per patient over a 12-month period is shown in Table 11. The findings are stratified by whether or not the patient had access to a care co-ordinator, a specialist centre and a care plan, and by whether the respondent was a patient or parent/carer. The ranking varied between groups, but the most common types of health service use contact across all groups were visits to the general practice, appointments for diagnostic and screening procedures (including blood tests and routine scans), outpatient appointments with doctors who are experts in the aspects of health affected (e.g. neurologist), outpatient appointments with other health professionals, outpatient appointments with experts in the patient’s condition or undiagnosed conditions and visits to the practice nurse. Contacts with patient support groups/charities were the most common contact. It was difficult to detect differences between groups, except that (adult) patients had more contacts with patient support groups/charities and more GP and practice nurse visits than parents/carers. Summing across all types of contact, the mean total number of contacts per patient ranged from 48 to 74 contacts per year, or from 1 to 1.5 contacts per week.

Perceived impact of care co-ordination on quality of care

Participants were asked to indicate how much they agreed or disagreed with the statements that having a care co-ordinator, attending a specialist centre or having a care plan improved the quality of care for people affected by rare conditions, regardless of their prior knowledge or experience of these factors. More than 70% of patients, parents/carers and HCPs agreed or strongly agreed that having a care co-ordinator would improve quality of care (Table 12), which is in sharp contrast to the numbers reporting access to this service (i.e. 12% of patients and 14% of parents/carers). More than 80% of patients, parents/carers and HCPs agreed or strongly agreed that having access to a specialist centre would improve quality of care, compared with 32% of patients and 33% of parents/carers actually having access to one. Similarly, more than 80% of participants in each group agreed or strongly agreed that having a care plan would improve quality of care, compared with 10% of patients and 44% of parents/carers actually having a care plan.

Key findings

We undertook a national survey, involving 760 adult patients affected by rare diseases, 446 parents/carers of people affected by rare diseases and 251 HCPs who care for people affected by rare diseases, to understand how care of people with rare conditions is co-ordinated in the UK. Only 12% of adult patients affected by a rare disease reported that they had a formal care co-ordinator, 32% reported that they attended a specialist centre and 10% reported that they had a care plan. For parents/carers of patients affected by a rare disease, these figures were 14%, 33% and 44%, respectively. Fifty-four per cent of patients and 33% of parents/carers had access to neither a formal care co-ordinator, nor a care plan, nor a specialist centre. By contrast, only 2% of patients and 5% of parents/carers reported having access to all three elements.

How the findings relate to previous research

There are few UK data to compare the current findings with. Nonetheless, the findings do support the findings of a limited number of previous studies,5,10 that is that levels of care co-ordination experienced by people affected by rare diseases in the UK are poor. For example, the 2016 survey by Rare Disease UK of more than 1200 people affected by rare conditions (including some carers) found that 12% of respondents had a care co-ordinator and 30% of respondents were aware of a specialist centre for their condition, with 66% of these respondents accessing it. 5

Limitations

It is not possible to determine the generalisability of the results from this study. There is a paucity of evidence regarding the total number of people living with a rare condition in the UK and their characteristics, such as gender, age distribution, ethnicity, level of education and socioeconomic status, are unknown. Respondents to this study were self-selected after receiving an e-mail or social media message about the survey, usually from a patient organisation that was on the distribution list of one or more of the study partners. We also recruited patients via NHS providers, but some patient groups may be under-represented in the survey respondents. For example, people who do not have links to one of the patient organisations and who do not use the NHS hospital services included in our study will be under-represented in the survey respondents. We offered several routes to accessing the survey (and not just via the online platform); however, there still remains a chance that people affected by rare diseases without a computer/e-mail address and those with lower electronic health literacy will be under-represented. It is assumed that the high proportion of females with a rare condition and female parents/carers in this study is over-representative. This may be because women are more likely to respond to surveys or because people were recruited to the study through patient organisations that women may be more likely to use. Sample bias may also exist in relation to the type of diseases that were represented among the study respondents. Some conditions may be over-represented in relation to the total sample size. Similarly, it is possible that more severe disorders, such as those that lead to major incapacity and/or early death, are under-represented. Despite this, our findings represent more than 200 rare diseases. In addition, as noted, it was not possible to estimate a response rate for each survey group. This was because the survey was sent via multiple overlapping distribution routes using convenience sampling and snowball sampling techniques, with multiple reminders and prompts to complete it. Therefore, although efforts were made to maximise the number of responses, it was not possible to know how many people received a request to participate in the survey. A survey undertaken by Rare Disease UK in 20165 used similar sampling processes and achieved a sample size of 1213, which suggests that our response rate was not especially low for undertaking research into people affected by are conditions.

Further research

This study provides new evidence on the extent of care co-ordination for people affected by rare conditions in the UK. We have provided evidence that only a small proportion of respondents have access to key aspects of care co-ordination. Further research would be beneficial to build on these findings and to identify new models of care co-ordination that could then be the subject of formal evaluation. Given the variation in preferences that was evident from the open-text feedback (e.g. around the benefits and limitations of different aspects of care co-ordination), further research would be valuable to gain more in-depth understanding of the diversity of preferences and the circumstances under which certain views dominate.

What was already known?

• Care for people affected by rare diseases is not well co-ordinated.

• There are several ways in which care could be co-ordinated for people affected by rare diseases, including using care co-ordinators, specialist centres, better communication and different types of support for patients, families and HCPs.

What this chapter adds

• We undertook a DCE to evaluate preferences for co-ordinated care by patients, parents/carers and HCPs.

• Participants preferred services where the cost of attending appointments was lower, electronic health records were immediately accessible to staff, the lead consultant was a medical expert in the patient’s specific medical condition, care was provided with the support of a care co-ordinator, a specialist centre was available and there was a documented emergency plan in place. There were some differences between the preferences of patients and parents/carers, and HCPs.

Overview of approach

Preferences were explored using a DCE. 13 In DCEs, respondents are typically presented with a series of questions that ask them to choose between two or more alternatives that describe a service in terms of a set of characteristics (attributes). This allows the attributes of a service that respondents prefer, as well as the trade-off they are willing to make between attributes, to be evaluated. These methods have been used to examine practitioner preferences for care co-ordination among older people. 198 DCE guidelines were followed for study design and analysis. 199 The DCE was one section of the survey described in Chapter 5.

Survey sampling

For details about the DCE sampling see Chapter 5, Survey sampling.

Discrete choice experiment questionnaire design

Respondents were asked to choose their preferred option of care from a series of pairwise choices. Each service was described by a combination of different levels of the attributes (see Table 14 for an example of a DCE ‘choice set’). An opt-out or ‘neither’ option was not included, as people are unlikely to choose none of the available options, given current levels of service provision. The number of potential combinations of attributes with four two-level attributes, one three-level attribute and one four-level attribute was 192 (2⁴ × 3¹ × 4¹). With two options to choose from in each choice question, this gave a possible 36,672 choices (192 × 191). To reduce the number of choices to a manageable number, a fractional design was applied, where the final design was selected using relative D-efficiency to maximise the balance and orthogonality of the design (this was undertaken using the –dcreate– command in Stata). Initially, the coefficient parameters were assumed to be zero. The choice set was reduced to 18 scenarios, which were split into three blocks of six, and a third of the respondents in each group were assigned to each block. It was felt, based on the feedback received, that six choice questions was a reasonable number to ask study participants. Nine versions of the DCE questionnaire were used (i.e. three for patients, three for parents/carers and three for HCPs). The questionnaire also included a question asking respondents to rank the six attributes according to their overall importance, from 1 (most important) to 6 (least important). Information on demographic, socioeconomic and rare disease-related experience was also collected. The questionnaire was piloted in 11 respondents (three patients, four carers and four HCPs) via three think-aloud interviews (with two carers and one HCP), with the other eight respondents providing written feedback. This resulted in minor improvements being made to the wording of the questionnaire.

Data analysis

Descriptive statistics for the characteristics of respondents were computed. Responses to the ranking questions were analysed graphically. The DCE data were analysed using alternative-specific conditional logit regression models, in which the outcome was service preference (i.e. service A or service B) and the variables in the equation were the individual attributes. A constant term was not included. Models were run for each group separately and differences in preferences between the groups were tested by comparing the coefficients for each group using chi-squared tests. Where the coefficients were not jointly different between groups, those groups were combined in subsequent analyses. The relative importance of each attribute was calculated as the difference in the coefficients between the best or most preferred level of each attribute and the worst or least preferred level of the same attribute. 200 We calculated MRSs with respect to the cost attribute (e.g. cost to patients and carers of attending all appointments during 1 year). This allows direct assessment of how much of one attribute participants are willing to trade for one unit of another attribute and, therefore, enables a comparison of different attributes on a common scale. Using the cost attribute as the denominator means that participants’ preferences and the trade-offs can be evaluated in terms of willingness to pay. The standard error of the MRS was calculated using the delta method. Findings from the regression analysis were used to calculate the predicted probabilities of choosing co-ordinated services compared with no co-ordination. No co-ordination was defined as cost to patients and carers of attending all appointments during 1 year were £1000; health records were not shared; the lead consultant was a medical expert in the area of the body primarily affected by the patient’s condition (e.g. neurologist); care was provided without the support of a care co-ordinator; a specialist centre was not available; and there was a documented emergency plan in place. In each co-ordination scenario, costs remained fixed at £1000 (i.e. co-ordination has no impact on costs) and the following potential characteristics of a co-ordinated service were amended individually and then jointly: electronic health records were immediately accessible to staff; the lead consultant was a medical expert in the patient’s specific condition; the patient/carer decided how they wished to be supported by the care co-ordinator; a specialist centre was available; and there was a documented emergency plan in place. We recalculated the predicted probabilities (1) assuming that no co-ordination was associated with high costs (£2000) and co-ordination was associated with low cost (£200) and (2) assuming that no co-ordination was associated with low costs and co-ordination was associated with high cost. All analyses were undertaken using Stata.

Responses and sample

In total, 996 responses to the DCE section of the survey were received, 528 responses from patients, 280 responses from carers and 188 responses from HCPs. Descriptive statistics of respondents who completed the DCE are in Table 15. These descriptive statistics are similar to the descriptive statistics for the survey respondents overall (see Chapter 5, Responses and sample).

Simple attribute ranking

The responses to the ranking question posed after the DCE questions were examined (Figure 9). Ninety-seven per cent of patients and carers and 99% of HCPs provided full responses to this question. Attributes were ranked by likelihood of being selected as the most important factor. Clinical expertise and access to a specialist centre were ranked highly by each group, and the role of the care co-ordinator was consistently ranked to be the least important factor. Attributes were ranked in the same order for each group.

FIGURE 9.

Ranking of attributes by group. (a) Patients (n = 512 respondents); (b) parents/carers (n = 271 respondents); and (c) HCPs (n = 186 respondents).

Regression analysis

There was no substantive difference in the preferences for the attributes between patients and parents/carers. The findings were quantitatively very similar (i.e. there was no statistically significant difference between the coefficients) and qualitatively identical (in terms of the sign and statistical significance of the coefficients) between the two groups, and so we reran the analyses for both groups combined (Table 16). Individuals in all groups preferred services with better co-ordination where the cost of attending appointments was lower; electronic health records were immediately accessible to staff; the lead consultant was a medical expert in the patient’s specific medical condition; care was provided with the support of a care co-ordinator; a specialist centre was available; and there was a documented emergency plan in place. There were some differences between the preferences of patients and parents/carers and the preferences of HCPs. In terms of care co-ordinators, HCPs preferred that care was entirely co-ordinated on behalf of the patient by a care co-ordinator, whereas patients and parents/carers preferred that they decided how they wished to be supported by the care co-ordinator. In terms of emergency plans, all three groups preferred there to be a documented emergency plan in place, but the preferences of HCPs for this was stronger than for patients and parents/carers.

Relative importance of the attributes

For patients and carers, access to a specialist centre was the attribute valued most highly, followed by clinical expertise, access to health records, the cost of attending all appointments during 1 year and having a documented emergency plan. The role of care co-ordinator was the attribute least valued. For HCPs, access to health records was valued most highly, followed by having a documented emergency plan, access to a specialist centre, the cost of attending all appointments during 1 year and clinical expertise. The role of care co-ordinator was valued least highly. These findings are preferred to the simple attribute ranking, as they account for the levels of the attributes.

Marginal rates of substitution

As an indication of their strength of preference, and the value they put on each attribute, patients and parents/carers were willing to pay £2509 for access to a specialist centre, £2470 for a consultant who was a medical expert in the patient’s condition, £2442 for electronic health records that were immediately accessible to staff, £1367 for a documented emergency plan and £1306 for the support of a care co-ordinator where the patient/carer decided how they wished to be supported (see Table 16). HCPs were willing to pay £1864 for electronic health records that were immediately accessible to staff, £1832 for a documented emergency plan, £1802 for patient access to a specialist centre, £1252 for a consultant who was a medical expert in the patient’s condition and £1131 for a care co-ordinator who entirely co-ordinated care on behalf of the patient. These MRS values reflect the relative importance of the attributes.

Predicted probabilities

The probability that respondents would choose a service with different types of care co-ordination compared with no co-ordination is shown in Figure 10. We defined ‘no co-ordination’ as a service where health records are not shared, the lead consultant is a medical expert in the area of the body primarily affected by the patient’s condition (e.g. neurologist), care is provided without the support of a care co-ordinator, a specialist centre is not available and there is not a documented emergency plan in place. Compared with this option, we found that respondents had a higher probability of choosing a service that had any of the individual attributes of co-ordination. For patients and parents/carers, the probabilities ranged from 0.60 to 0.67, depending on which individual attribute was selected, with the attributes ranked in terms of their predicted probability in the same order as the relative importance (see Figure 10a). If a service achieved all of the attributes of co-ordination, then the probability that patients and carers would prefer to use that service was 0.94. For HCPs, the probabilities for each individual attribute ranged from 0.59 to 0.66. For a service that achieved all of the attributes of co-ordination the probability was 0.96 (see Figure 10b). When co-ordination reduced (increased) costs compared with no co-ordination the probability that respondents would choose a service with the different types of care co-ordination increased (see Appendix 6).

FIGURE 10.

Predicted probabilities of choosing co-ordinated services. (a) Patients and parents/carers combined; and (b) HCPs. Note that for ‘no co-ordination’, health records are not shared, the lead consultant is a medical expert in the area of the body primarily affected by the patient’s condition (e.g. neurologist), care is provided without the support of a care co-ordinator, a specialist centre is not available and there is not a documented emergency plan in place. For ‘full co-ordination’, electronic health records are immediately accessible to staff, the lead consultant is a medical expert in the patient’s specific condition, the patient/carer decides how they wish to be supported by the care co-ordinator (patients/carers) or care is entirely co-ordinated by a care co-ordinator (HCPs), a specialist centre is available and there is a documented emergency plan in place. All other co-ordination scenarios are as for ‘no co-ordination’ except for the attribute indicated. In all scenarios the cost to patients and carers of attending all health-care appointments during 1 year is held constant at £1000. Scenarios are ordered from left to right in ascending order of magnitude of the predicted probability of choosing the co-ordination service (note that the ordering is different for patients and carers combined and for HCPs).

Key findings

Patients, parents/carers and HCPs all preferred services where the cost of attending appointments was lower, electronic health records were immediately accessible to staff, the lead consultant was a medical expert in the patient’s specific medical condition, care was provided with the support of a care co-ordinator, a specialist centre was available and there was a documented emergency plan in place. Preferences were found to be consistent with better co-ordination of care, although there were some differences between the preferences of patients and parents/carers, and HCPs. The probability that participants would choose a service with all the elements of co-ordination studied in place was high.

How the findings relate to previous research

There are several studies that have explored how people affected by rare diseases would like their care to be co-ordinated, although these studies tend to focus on single options, such as care co-ordinators10 or specialist centres. 5 We are not aware of any studies that have compared between multiple aspects and, to the best of our knowledge, none have used a DCE-based approach.

Limitations

Several limitations are acknowledged. DCEs elicit hypothetical choices and, therefore, might lack external validity if individuals do not make the same choices in real-life situations. Some aspects of the DCE might be difficult for respondents to understand, such as the forced choices between services, probabilities and clinical concepts. The representativeness of the samples used might be limited by the recruitment strategies, yielding potential sampling bias (e.g. there was a high proportion of female patients and parents/carers). The modal education category was those who were educated to degree level or higher, and it is unclear if costs would have been the least important attribute if, for example, the sample was, on average, less well educated. Although the overall sample size was large, we obtained fewer responses from parents/carers and HCPs than targeted. There might be other components of co-ordinated care that are important, but were not included in the present analysis. Unfortunately, the number of attributes that can be included in a DCE is limited by the number of data that participants can process. The nature of our piloting work meant that we were unable to produce initial estimates of the model coefficients, which could have been used to inform the final study design. Initially, the coefficient parameters were assumed to be zero. Preferences might vary by subgroups within our study groups (e.g. parents of children affected by rare diseases vs. carers of adults with a rare disease), but sample size considerations make subgroup analyses problematic.

Further research

This study provides new evidence on the elements of care co-ordination that matter to people affected by rare diseases. Further research would be beneficial to develop different models based on people’s preferences, as described in this study, describing how care for people with rare conditions could be co-ordinated. These models could then be the focus of further formal evaluation. Further research would also be helpful to understand the reasons for the differences in preferences between patients and parents/carers on the one hand and for HCPs on the other.

What was already known?

• Improving care co-ordination for people with rare conditions may help to reduce the burden associated with co-ordinating their own care.

• Care co-ordination has many components and, therefore, there are many different ways of co-ordinating care for rare conditions.

• Taxonomies can be used to help us to organise complex phenomena, such as care co-ordination.

What this chapter adds

• We have developed a taxonomy of care co-ordination for rare conditions, which outlines the six domains involved in co-ordinating care for rare conditions. The six domains that we identified were (1) ways of organising care, (2) ways of organising teams, (3) responsibilities, (4) how often care appointments and co-ordination take place, (5) access to records and (6) mode.

• We have outlined qualifying factors related to each of the taxonomy domains. The qualifying factors provide insight into different participants’ preferences, benefits and challenges of different models of care co-ordination, factors that influence co-ordination, and barriers to and facilitators of co-ordination. These findings provide information on which models of co-ordination may be suitable in different situations.

Design

Our study used qualitative methods (i.e. interviews, focus groups and workshops).

Using a qualitative approach allows for a more in-depth understanding of complex phenomena. 206,207 In addition, the use of qualitative methods allows for the direct involvement of those with most experience in the phenomena being studied and classified, such as patients, HCPs and carers. This is particularly important in health-care service research in which patients, carers and HCPs are the key stakeholders. 208 By understanding patients’, carers’ and HCPs’ views on the organisation of care co-ordination for rare diseases, we could improve health-care services and optimise the patient experience, therefore, reducing burden. 209 It has also been proposed that qualitative studies are well suited to explore new concepts. 207 As co-ordination of care is a relatively new field, using qualitative methods will offer a rich perspective on care and stakeholders’ preferences.

This research was conducted in a two-stage process. First, interviews and focus groups were conducted to develop an initial taxonomy. Interviews and focus groups were felt to be appropriate methods for exploring and gathering in-depth perspectives on stakeholders’ experiences of co-ordination and the different models of care co-ordination that currently exist, together with preferences for potential new models of co-ordination. Second, workshops were conducted to refine the proposed taxonomy. Workshops were felt to be appropriate for helping to gather consensus around whether or not the taxonomy was appropriate and to develop recommendations to improve the taxonomy. Figure 11 outlines the two stages.

FIGURE 11.

Overview of the two stages involved in developing the taxonomy.

Measures

To gather data to inform the development of the taxonomy, two topic guides (one for interviews and one for focus groups) were developed and used to collect data (see Report Supplementary Material 1). Questions focused on stakeholders’ experiences of co-ordinated care, implications of co-ordinated care, preferences for key aspects of care co-ordination (including preferred way of co-ordinating care, format, access, frequency, location, information-sharing and transition), benefits and challenges, and factors that help and get in the way of co-ordination. Feedback on the topic guide was sought from the PPIAG prior to data collection.

To gather data to refine the taxonomy, one topic guide for both workshops was developed and used to collect data (see Report Supplementary Material 1). The topic guide was based around the six categories identified in the taxonomy and prompted participants about whether or not they had feedback on the category (e.g. if we had missed anything and whether or not findings seemed appropriate based on participant experiences), the appropriateness of options in the light of the COVID-19 pandemic and asked for recommendations to improve the category.

Procedure

Participants were recruited using a range of methods, including e-mail invitation, social media, via the charity sector and through our partnerships with four NHS sites. Potential participants were asked to contact the study researcher via e-mail or telephone.

To ensure that a range of participants with different experiences were recruited, potential participants were asked to provide responses to eligibility questions when registering their interest. Participants were sent these eligibility questions by e-mail. For HCPs, these eligibility questions included questions regarding their occupation, specialty and geographical region. For patients and carers, these eligibility questions included questions regarding age, ethnicity, geographical region, role, whether or not they receive co-ordinated care (i.e. a specialist service and who co-ordinates) and whether or not they have a diagnosis. The researcher checked that participants met the eligibility criteria for the study.

Selected individuals were asked to complete consent forms (one for the researcher and one for the participant) prior to taking part in the interviews, focus groups or workshops. Participants who took part virtually or via telephone were asked to return written consent forms in advance. Participants were informed that their data would be kept confidential, fully anonymised and that they could withdraw at any time without providing a reason. Focus group participants were informed that any data collected up until the point of withdrawal would be kept because of difficulties removing individual participants from focus group data. We took steps to ensure that quotations from the participant who withdrew from the study were not included in publications. These steps included removing withdrawn quotations from the analysis spreadsheet.

To gather data to inform the development of a taxonomy, interviews with HCPs, charity representatives and commissioners, and focus groups with patients and carers, were conducted.

One researcher conducted interviews either by telephone or face to face, depending on participants’ preferences. The interviews lasted approximately 1 hour (range 44–74 minutes). Two researchers (HW and AS) conducted the four focus groups (one researcher facilitated and one researcher took notes). 210 A third researcher (EH) observed one of the focus groups. Two focus groups were conducted face to face (one in London and one in Birmingham) and two focus groups were conducted virtually using Skype. Focus groups were up to 3 hours in length, including a break (range 149–154 minutes). Interviews and focus groups were digitally recorded using an encrypted dictaphone (with consent from participants) and professionally transcribed. Transcripts were checked for accuracy and fully anonymised (including names and places). Data were stored in the University College London Data Safe Haven (a secure electronic environment, certified to ISO27001 information security standard, conforming to the NHS Information Governance Toolkit) and coded using NVivo 12 (QSR International, Warrington, UK).

To refine the taxonomy, workshops were conducted virtually. Workshop participants were sent a brief 15-minute video prior to the workshop, which outlined the findings of the taxonomy (including each domain and the options within each domain). The presentation also covered qualifier findings, including preferences, barriers/facilitators, factors influencing co-ordination and benefits/challenges of different options. During the workshops, participants were given an introduction to the workshop before being split into three breakout groups. Each breakout group had one facilitator (HW, EH or AIGR) and one note taker (JJ, SM or AH). During the breakout groups, facilitators encouraged the small groups to discuss each of the six domains in the taxonomy. Facilitators prompted participants about whether or not they had feedback on the category (i.e. whether or not anything had been missed and whether or not findings seemed appropriate based on participant experiences), the appropriateness of options in the light of the COVID-19 pandemic and asked for recommendations to improve the category. After the breakout groups, participants reconvened in the main group and each group provided feedback on their discussions. Workshops were recorded using an encrypted dictaphone. Notes were checked for thoroughness and summarised prior to being sent to a graphic facilitator (New Possibilities, Birmingham, UK) to create a graphical representation of the findings.

Analysis

Thematic analysis was used to analyse interview and focus group data. It is recommended that inductive and deductive data analysis approaches are used to develop taxonomies. 204 Therefore, to generate codes, a combination of inductive and deductive coding was used.

To develop an initial coding frame, six interview transcripts were coded inductively by two researchers (HW and AS). From this, a coding framework was developed and agreed. The coding framework included codes on aspects of care co-ordination (e.g. types, who is involved, mode, information-sharing, where, frequency, transition and methods of access) and qualifier codes (e.g. preferences, benefits/challenges, barriers/facilitators and factors influencing co-ordination). The coding framework was used to code all interview and focus group transcripts (HW). A second researcher (AS) also coded 20% of the data (for six interviews and one focus group transcript). Coding was discussed and discrepancies were resolved and agreed.

Next, findings were grouped into themes and subthemes using Braun and Clarke’s211 thematic analysis methodology, supplemented by Neale’s188 iterative categorisation process, which is designed to support thematic analysis. Given the large number of data, this was carried out in two stages: (1) development of themes and subthemes for the data on aspects of co-ordination (to develop initial taxonomy options) and (2) development of themes and subthemes for the data on qualifying codes (to develop models). The following five themes were developed: (1) ways of organising care, (2) ways of organising teams, (3) responsibilities for co-ordination, (4) access to co-ordination and (5) mode of co-ordination. Themes and subthemes were discussed by co-authors and used to develop a taxonomy. Our taxonomy included different options, ranging from lack of co-ordination through to strategies to improve co-ordination.

To develop the proposed taxonomy, Nickerson et al. ’s212 six stages of taxonomy development were followed:

1. Identify the meta-characteristic that will inform the choice of characteristics in the taxonomy.

2. Identify ending conditions (i.e. requirements that the taxonomy needs to meet to be finalised).

3. Choose approach.

4. Identify a subset of objects to classify, using findings from the interviews and focus groups.

5. Identify common characteristics (i.e. similarities and differences will be identified to identify common characteristics and discriminatory characteristics for co-ordinated care).

6. Group the characteristics using a manual or graphical process. 203

Table 17 outlines how we applied these six steps.

To refine the taxonomy, the workshop notes were analysed by one researcher using inductive thematic analysis. Workshop notes were coded and grouped into themes relating to stakeholders’ experiences of the model of co-ordination, benefits and challenges of the model of co-ordination, factors influencing co-ordination, missing aspects and the impact of COVID-19. Feedback on aspects that were missing in the taxonomy were used to refine and finalise the taxonomy. Findings were discussed with the wider research team.

Participant characteristics

Seventy-nine participants took part in this study (including patients, carers, HCPs, commissioners and charity representatives). Data from 52 participants informed the development of the taxonomy (interviews, n = 30; focus group participants, n = 22) and data from 27 workshop participants informed the refinement of the taxonomy (patient/carer workshop, n = 12; professional workshop, n = 15). Two of the interview participants also took part in a workshop. Demographic characteristics are shown in Table 18.

Overview of care co-ordination taxonomy

Our taxonomy of care co-ordination consists of six domains: (1) ways of organising care, (2) ways of organising the team, (3) responsibility for co-ordination, (4) how often appointments and co-ordination take place, (5) access and (6) mode (Table 19). Each domain has different options for co-ordinating care (labelled ‘characteristics’). ‘Objects’, which outline the different ways of co-ordinating care for each option are also highlighted, for example, quotations illustrating each option within the taxonomy (Table 20). A summary of the domains and options are highlighted in Figure 12. The examples given in this taxonomy refer to those identified through interviews and focus groups (and validated within the workshops). Many of the examples highlight examples of real-world practice; however, some of the examples are hypothesised and relate to potential new ways of co-ordinating care.

FIGURE 12.

A summary of the taxonomy domains (numbered) and care co-ordination options (bullets).

The taxonomy outlined in this chapter is the final version that has been refined from workshop feedback. Findings from the workshops highlighted key aspects that needed to be clarified within the taxonomy, including the need to emphasise that care is not just medical (i.e. it also includes social and educational aspects of care) and that care is lifelong. Workshop findings highlighted the need to separate out collaborations that include patients/carers from collaborations between professionals, the need for third-sector involvement in collaboration (where appropriate), the need to emphasise the role that charities and patients/carers play in care co-ordination, a hybrid model of frequency and the need to clarify aspects of the mode domain. We amended the taxonomy in line with this feedback. See Report Supplementary Material 3 for visual representations of the workshop findings.

‘Qualifiers’ refers to information relating to people’s preferences, the benefits and challenges of different ways of co-ordinating care, factors influencing co-ordination, and barriers to and facilitators of underpinning co-ordination. A summary of qualifying information, relating to preferences (see Appendix 7), benefits and challenges (see Appendix 8) and factors influencing co-ordination (see Appendix 9), identified through the interviews and focus groups is shown in Table 21.

Our interview and focus group findings also identified many barriers and facilitators underpinning these domains of care co-ordination. Barriers and facilitators fit within the following five themes: (1) ability, (2) attitudes, (3) opportunity, (4) resources and (5) environment. We grouped barriers and facilitators into patient factors (i.e. ability, attitude and opportunity), provider factors (i.e. ability, attitudes and opportunity), health-care environment factors (i.e. resources, environment and attitudes) and societal factors (i.e. resources and attitudes) (see Appendix 10).

Taxonomy domains and qualifiers

Ways of organising care

Our findings highlighted different ways of organising care (see Table 19). These ways ranged from local care provision where all care is delivered locally through to care being delivered in national centres that serve all patients in the country with a particular rare condition. There are also some ‘hybrid’ options, which combine both specialist and local care (e.g. outreach clinics that are delivered by clinicians from specialist centres but in a patient’s local area). See Table 19 for further examples and Table 20 for sample quotations.

Workshop findings indicated that participants experienced a change in services due to COVID-19 (e.g. reduced access to specialists and limited capacity for local services).

Figure 13 provides a summary of the different ways of organising care.

FIGURE 13.

Ways of organising care (visual representation of taxonomy domain 1).

In terms of qualifying factors (see Table 21), findings indicated that participants from all stakeholder groups may prefer nationally commissioned services and hybrid models. Hybrid models include specialist centres co-ordinating care and local services delivering care, outreach clinics and provision of support for local providers:

Hmm . . . well obviously ideally close to home but I think the majority of our patients, if they feel they’re going to be getting a good service and a specialist service, they are willing to travel to a specialist centre.

Interviewee, HCP

Both of these options have benefits and challenges. For example, nationally centralised locations, such as specialist centres, might improve co-ordination and increase access to expertise. However, these services are not available for all conditions and may not cover all aspects of care that the patient needs. For hybrid options, people pointed out benefits, such as reducing travel and the ability to provide education to local health-care providers.

Many factors were perceived to influence the type of centralisation, including patient factors (e.g. age, ability to travel and rare condition), health-care environment factors (e.g. availability of resources, such as funding issues and availability of experts and models of co-ordination), environmental factors (e.g. ease of access and suitability of the environment, and relationships between care teams, such as specialist and local teams) and societal factors (e.g. funding and availability of service specifications and policies). For example, the patient’s condition was perceived to influence how care is organised in a number of ways, including the nature of the condition (e.g. the complexity of the condition, whether or not the condition affects multiple body systems, the number of disciplines involved in a patient’s care and need for co-ordination across a whole spectrum of care and not just acute medical situations). Participants also felt that specialist services work only if the condition has a discrete phenotype and if services know exactly who a patient will need to see. Conditions that do not fit into a clinical group or are difficult to define may not be well placed to be cared for within a specialist service. In addition, conditions that are more stable may require less co-ordination (e.g. may just require a point of contact within a specialist centre). Where the patient lives also determines how care should be co-ordinated. Findings indicated that patients and families may fit into three groups: (1) those who live far away from the specialist centre but can travel, (2) those who live far away from the specialist centre but cannot travel and (3) those who live close to the specialist centre. Different models of co-ordination may be needed for these different types of families. For example, those who live far away from the specialist centre may require visits to specialist centres to be minimised and care to be delivered locally or online.

Ways of organising teams

Our findings highlighted different ways of organising teams (see Table 19). Options ranged from little collaboration (e.g. not having a MDT) to high levels of collaboration (e.g. all HCPs working together to provide or discuss care in a condition-specific clinic or MDT meeting). Other options included some HCPs working together (e.g. in joint clinics). See Table 20 for example quotations.

Workshop findings highlighted that COVID-19 may have offered new opportunities for collaboration, such as the ability for local team members to dial into MDT meetings. Figure 14 provides a summary of the different ways of organising teams.

FIGURE 14.

Ways of organising teams (visual representation of taxonomy domain 2).

In terms of qualifying factors (see Table 21), findings outlined preferences for condition-specific clinics or joint clinics, meetings and some transition methods to support patients. Each of these options have benefits and challenges. For example, condition-specific clinics and MDTs allow teams to figure out who patients need to see, ensure that all those involved in a person’s care receive the same messages and may reduce travel. However, these options are difficult to organise and may be tiring for patients:

. . . we do support this idea of multidisciplinary team clinics, and then that those MDTs develop good lines of communication with GPs and other providers. That seems to work the best, and we think that there’s some evidence that patients do better when they’re under the care of those sorts of clinics.

Interviewee, charity representative and HCP

Many factors were perceived to influence the type of collaboration, including patient factors (e.g. age and condition), provider factors (e.g. knowledge, understanding and whether or not the team has multidisciplinary expertise) and health-care environment factors (e.g. resources and availability of collaboration models, such as joint clinics, MDT clinics and orientation visits, and availability of experts). Patient factors included age (e.g. clinics varying for adults and children) and the patient’s condition. For example, the nature of the condition influences collaboration, as the type of clinic used depends on how multisystemic the condition is and how many disciplines it involves. Carousel clinics (where the patient sees different HCPs one after another) or MDT clinics may be suitable for only those conditions that affect multiple body systems, and MDTs may only work if there is a discrete phenotype with clarity over which professionals need to be seen.

Responsibilities

Our findings highlighted the different types of responsibility involved in co-ordinating care for rare conditions (see Table 19), including administrative, formal and supportive roles. Administrative support included help with organising appointments and having a point of contact. Many different professionals were identified as doing, or having potential to do, these roles (e.g. administrators, co-ordinators, rare disease charities, and a combination of patients and administrators). Formal co-ordination responsibilities were identified across three roles: (1) those conducted by a co-ordinator (i.e. an administrative, general or clinical co-ordinator), (2) those conducted by a clinical lead (e.g. overseeing care) and (3) those conducted by a GP (e.g. being a point of contact, co-ordinating care, referrals and signposting). Findings outlined different options for who the co-ordinator could be, including those in doctor-equivalent roles, nurse or allied health professional roles, non-medical roles and social care sector roles. Supportive roles were also identified, including those conducted by patients and carers (e.g. in a direct role as co-ordinators, or involvement in education, MDTs and information provision) and those conducted by charities (e.g. direct roles in co-ordination, providing support for co-ordination and advocating on the patient’s behalf). Figure 15 provides a summary of the different types of responsibility. See Table 20 for sample quotations.

FIGURE 15.

Types of responsibilities for co-ordination (visual representation of taxonomy domain 3).

In terms of qualifying factors (see Table 21), findings indicated that participants would prefer a point of contact to answer queries, a co-ordinator (e.g. a nurse or allied health professional), a clinical lead, support from their GP and support from charities. Workshop findings highlighted the importance of charities in care co-ordination and the importance of patients and carers who are often co-ordinating their own care:

I agree. I don’t think it’s difficult. I think you can have a . . . I think you need a named consultant as the overall co-ordinator . . . But then maybe the person you have face to face with, the person who is a co-ordinator or . . . Often I do think the senior nurse is really good.

Focus group participant, patient

Each of these options have benefits and challenges. For example, benefits of co-ordinators include helping build relationships between patients and the team, and supporting patients. However, co-ordinators need time and a dedicated role. These roles may not always exist currently. People felt that clinical leads provide expertise, holistic care and facilitate collaboration between professionals. People felt that GPs were lacking time, sometimes motivation and clear methods to refer patients to services.

Many factors were perceived to influence who takes responsibility, including patient factors (e.g. diagnosis, age of patient, condition and the individual’s needs and preferences), provider factors (e.g. knowledge, support and education, understanding of the health-care system, interest and motivation, and time and availability of a team to work with), health-care environment factors (e.g. resources, such as availability of co-ordinator roles) and societal factors (e.g. availability of patient organisations, stigma and willingness to change). The patient factor that was discussed most frequently was the patient’s individual needs and preferences. For example, patient choice on who sees their records, which HCPs they see, who co-ordinates their care and the extent to which the patient/carer are involved in co-ordination and meetings. In addition, individual patient needs influence who is involved in co-ordination (e.g. the need for co-ordinated care and who is involved to be tailored and take individual family needs and ambitions into account). For example, a national care co-ordinator model that takes the person’s individual needs into account to determine how much contact they have with their co-ordinator or the level of co-ordination. An additional factor relating to individual needs was the patient’s ability to self-manage and co-ordinate their own care. Some patients may be able to co-ordinate their own care, but others would struggle and, therefore, need a co-ordinator.

How often care appointments and co-ordination take place

Our findings highlighted different time periods for care appointments and co-ordination activities. Options included regular appointments and on-demand appointments (see Table 19). Workshop findings highlighted the need for a hybrid category that combines both regular care (at a minimum) with on-demand support. See Table 20 for example quotations.

Workshop findings highlighted that COVID-19 may have provided opportunities for on-demand appointments for those with stable conditions (as long as safety nets are in place).

See Figure 16 for a summary of this domain.

FIGURE 16.

Different options for how often care appointments and co-ordination appointments take place (visual representation of taxonomy domain 4).

In terms of qualifying factors (see Table 21), there was less agreement regarding preferences for different time periods, with some participants preferring on-demand appointments for care and/or co-ordination (as this enables them to access care when needed and not waste providers’ time) and other participants preferring regular appointments (e.g. to receive check-ups and updates on care).

Many factors were perceived to influence frequency, including patient factors (e.g. diagnosis, age, ability to travel, condition, stability or progression associated with the condition, phase, severity, individual needs and time since treatment), provider factors (e.g. time and knowledge) and health-care environment factors (e.g. availability of job roles, guidelines, time within job roles and funding):

I guess it depends on the condition and how much things are changing, and whether it is a life-limiting condition, because if it is a life-limiting condition there is probably more things that are changing more rapidly. So, I think it has to be condition specific, so I guess you would be guided by what the experts think is appropriate.

Interviewee, charity representative and HCP

Access

Our findings highlighted different types of access to records. Options ranged from full to restricted access to records for patients and providers (see Table 19). Workshop participants highlighted that full access to records with a summary of important details may be helpful. See Figure 17 for a summary of this domain. See Table 20 for sample quotations.

FIGURE 17.

Options for access to records (visual representation of taxonomy domain 5).

In terms of qualifying factors (see Table 21), patients and HCPs having access to records was seen as important throughout the interviews and focus groups, but it was less clear what people would prefer in terms of full or filtered access. For example, for HCPs, our findings show that it is important for HCPs to have access to information and records. However, the extent to which HCPs can access information and records was not as clear. Some participants felt that any HCP should be able to access the records. Other participants felt that access to records should be limited (e.g. to only necessary information, such as current and relevant information/information necessary for each discipline). Reasons for this tended to differ across patients, carers and HCPs. For example, some patients and carers felt that they would not want all of their HCPs to have access to all aspects of their records (e.g. those parts that are irrelevant to the current condition or situation/aspects of childhood records that are no longer relevant) and that they would want control over who has access. Some HCPs spoke about how access to complete records can also be overwhelming and that it may be necessary to filter information by relevance:

I mean, I personally wouldn’t mind it shared with anyone. I’d rather the more people.

Yeah I’m the same.

I’m the same. I mean, with my daughter, I was, like, ‘Yay. If you want to look at this and you want to use this to help her . . . if this can make you more informed, if this can connect you to my daughter, please do ahead and do it’, you know.

Factors that were perceived to influence access included patient factors (e.g. diagnosis and consent), health-care environment factors (e.g. resources, environmental factors and attitudes) and societal factors (e.g. funding).

Mode of communication

Our findings highlighted different modes, including modes for information-sharing, care delivery and/or co-ordination and communication (see Table 19). Figure 18 provides a summary of this domain. See Table 20 for sample quotations.

FIGURE 18.

Different options for mode of co-ordination (visual representation of taxonomy domain 6).

For information-sharing, our participants described many different modes, including digital methods, written methods, verbal methods or a lack of information-sharing. In terms of qualifying factors (see Table 21), our participants preferred digital methods (e.g. online portals, records, mobile applications, e-mails and databases) and written methods (e.g. care plans, letters, written agreements of responsibility, patient-held records and condition-specific passports). Digital methods were seen to provide easier and quicker access to information, but were limited by information technology failures and were difficult to keep up to date. Written methods were seen to keep everyone up to date and ensure accountability, but were considered to have the potential to get lost or delayed:

I think that in a totally ideal world – see, it’s pie in the sky, but in a totally ideal world, if all of the NHS had electronic patient records that were all on the same system and could be shared automatically between units then, you know, we’d be able to see things more nationally.

Interviewee, charity representative and HCP

In terms of care and co-ordination appointments, our participants described many different modes, including face to face, digital, telephone and a combination of methods. Workshop findings highlighted that COVID-19 has accelerated the shift from traditional methods of delivery of care (i.e. face to face) to digital and telephone delivery of care for people living with rare conditions. In terms of qualifying factors (see Table 21), it was less clear what people preferred and each mode had benefits and challenges. We found that digital appointments may reduce travel and may be suitable for reviews and updates, but these appointments cannot fully replace face-to-face appointments. Using a combination of methods was felt to keep everyone updated, reduce travel, save time and money, and ensure that everybody involved is informed/in agreement and has the same information.

For communication, participants described many different modes, including face to face, digital and telephone methods. In terms of qualifying factors (see Table 21), participants’ preferred different modes and each mode had many different benefits and challenges. Face-to-face methods were perceived to reduce misunderstandings and help to agree plans, but were limited by availability. Digital methods were seen as good for reducing time and agreeing solutions. Telephone methods were considered suitable for answering patient queries.

Perceived factors influencing mode of information-sharing and care delivery included patient factors (e.g. age, condition and individual needs) and health-care environment factors (e.g. access to technology).

Key findings

We have developed a taxonomy of care co-ordination for rare conditions. We identified the following six domains of care co-ordination: (1) ways of organising care (i.e. local, national or a hybrid), (2) ways of organising the team (i.e. high collaboration, some collaboration or low collaboration), (3) responsibility for co-ordination (i.e. administrative support, formal roles and responsibilities, supportive roles and no responsibility), (4) how often appointments and co-ordination take place (i.e. regular appointments, on-demand appointments or a hybrid), (5) access (i.e. full or filtered access to records) and (6) mode of information-sharing, care co-ordination/delivery and communication.

Our findings highlighted various stakeholder preferences, benefits/challenges and factors influencing co-ordination for different options within each of the six domains. These findings indicate that different models of care co-ordination for rare conditions may be appropriate in different situations.

How findings relate to previous research

These findings extend knowledge on care co-ordination for rare conditions. National policy documents and previous research has highlighted the importance of care co-ordination. 8,15,19 However, findings indicate that little is known about co-ordination for rare conditions15 (see Chapter 3). Previous research has shown that co-ordination for rare and common chronic conditions has many components,1,15,117,184 but care co-ordination had not been formally categorised. In addition, although previous taxonomies have been developed for other complex health concepts, such as integrated care213 and the burden of treatment for patients with chronic conditions,209 to the best of our knowledge, there have been no taxonomies that have focused specifically on care co-ordination for chronic or rare conditions. The taxonomy presented in this chapter extends previous research by formalising care co-ordination for rare conditions into six domains (each with different options).

In addition, this research offers insight into participants’ preferences, the benefits and challenges of different models of co-ordination, factors influencing co-ordination, and the barriers to and facilitators of co-ordination in general. These findings extend previous knowledge by identifying possible situations in which different models of co-ordination may be appropriate. For example, previous research has highlighted that some aspects of care co-ordination may be necessary for rare conditions (e.g. care co-ordinators and specialist centres). 5,10,11 However, to the best of our knowledge, there has been little research on the benefits and challenges of each model and how they work in practice. This research extends this knowledge by outlining the benefits and challenges associated with each model.

Given that there are many different rare conditions, it was not known whether or not one taxonomy would be applicable across different rare conditions. These findings highlight that although different conditions have different characteristics and challenges, it is possible to develop a taxonomy that covers a range of rare conditions. Previous research1 has indicated that more care co-ordination is needed in complex situations (e.g. limited patient capacity and clinical complexity). Our findings concur with this and highlight a range of factors that need to be considered when choosing how to co-ordinate care, including patient factors, provider factors, environmental factors and societal factors. Examples of patient factors included severity and complexity of condition, where patients live, the patient’s ability to travel and the patient’s ability to co-ordinate care. Therefore, findings indicate that a ‘one size fits all’ approach to care co-ordination is not appropriate and that we should develop models of care co-ordination that take into account a range of individual, organisational and societal factors, rather than developing different models of co-ordination for each individual condition. Models can then be tailored to individual situations.

Our findings highlighted three main options for organising care. Findings extend previous research11,214–216 by demonstrating that participants from all stakeholder groups indicated a strong preference for nationally commissioned services and hybrid models (including hub and spokes models, network models and outreach models) because of the associated benefits (e.g. increased co-ordination, access to expertise and reducing travel). This supports previous research, which highlights the potential benefits of specialist services,11 hub and spokes models214,215 and outreach models216 for different health conditions. However, for rare conditions, findings indicated that these models may not be appropriate in all situations and, in some situations, patients may prefer local care (e.g. if they are unable to travel or do not live near to a specialist centre). In addition, specialist services may not be appropriate for every condition (e.g. for conditions that do not have discrete phenotypes). These findings highlight that different models of co-ordination are needed for different types of families (e.g. those who live near to specialist centres, those who live far away but can travel and those who live far away but cannot travel).

Our findings on the organisation of teams for rare conditions supports previous research that indicates the importance of collaboration and MDTs for rare conditions and other conditions. For example, previous research has highlighted negative implications for patients associated with co-ordinating their own care, including repeating information to different HCPs. 184,186 Findings also support previous research that has indicated a need to join up care appointments from different disciplines and hospitals into one appointment (e.g. condition-specific clinics) to facilitate co-ordination. 15,184 However, findings indicate that collaboration does not always happen in practice and that improvements in collaboration/joined-up working are needed.

Our findings extend previous research by highlighting the different types of responsibility needed to co-ordinate care for rare conditions. Previous research has indicated the importance of care co-ordinators. 5,10,15,184 However, our findings extend previous research by highlighting the importance of the many different roles needed to co-ordinate care. These roles include administrative support, co-ordinators, clinical leads, GPs and charities. Patients and carers currently play large roles in care co-ordination; however, we found that patient involvement in co-ordination was not always appropriate if patients were unable to, or did not want to, co-ordinate their own care. This finding is consistent with previous research that has indicated the negative impact co-ordinating care can have on patients and families184,186 and treatment burden more generally. 205,209 These findings indicate that different models of co-ordination are needed to take into account those who are able and want to co-ordinate their care and those who cannot. For example, the level/type of co-ordinator offered should vary depending on complexity and the patients’ ability to co-ordinate their own care. However, findings indicate that these roles do not always exist in practice and that further resources are needed (e.g. specific co-ordinator roles and training pathways).

Clinical guidelines for rare conditions (where available) outline how often patients with certain conditions should be seen in practice for their care appointments. Our findings offer support for including timings of appointments within such guidelines, as the findings demonstrated the importance of regular care and co-ordination appointments, particularly at key stages of a patient’s journey or condition. However, in some situations, on-demand appointments or a combination of regular and on-demand appointments were felt to be appropriate. Therefore, these findings suggest that the frequency of co-ordination and care appointments/meetings need to take into account factors such as patient and provider preferences.

Findings support previous research that indicates that a combination of methods can be used to deliver health care, including methods to share information, methods to facilitate co-ordination and methods to communicate with professionals and patients. In particular, findings highlighted the potential for remote methods of co-ordination, including digital information-sharing (e.g. through electronic records), virtual clinics and care co-ordination appointments. This shift to digital methods has been accelerated during the COVID-19 pandemic and supports previous research217,218 that indicates that digital methods may show some potential for use in health-care delivery. Our findings suggest that this may also apply to care co-ordination. However, each mode of communication has benefits and challenges, and findings indicate that the mode of co-ordination should take into account many factors, including individual preferences and resources. In addition, findings indicate that digital appointments must not replace face-to-face appointments completely in terms of care delivery and co-ordination. Face-to-face appointments were felt to be integral, particularly at key points of the patients’ journey (e.g. initial meetings and diagnosis), for certain conditions or for patients requiring more in-depth clinical care co-ordination because of additional difficulties. This extends previous research219 by highlighting the limits of digital methods of care delivery and co-ordination, and emphasises the need to offer multiple modes when co-ordinating care for patients with rare conditions.

Limitations

Although our sample of participants included a variety of rare conditions, locations and sectors, we were unable to include every rare condition. Some groups of participants, including individuals from minority ethnic groups and certain roles (e.g. GPs), were under-represented in our sample. Therefore, we are unlikely to have captured every possible option of care co-ordination for rare conditions. However, we included as many different views as possible throughout the study.

Our topic guides for both the interviews and focus groups were comprehensive and included an extensive number of questions. Therefore, it is possible that this may have compromised depth. However, we obtained a large number of in-depth data from these interviews and focus groups and, therefore, it is unlikely that depth was compromised.

Care co-ordination is a complex concept. Therefore, it is possible that we may have missed relevant constructs. However, we have minimised this risk through the extensive data collected in this study, which, together with the survey and scoping review findings, provide a comprehensive overview of the different ways in which care can be co-ordinated for rare conditions.

In addition, owing to COVID-19, we were unable to conduct five face-to-face workshops as planned. Instead, we held two remote workshops with patients/carers and HCPs. This resulted in fewer participants overall. However, in-depth discussions and feedback indicated that findings were appropriate and, therefore, this did not have an overall impact on the study.

Our taxonomy included examples of care co-ordination that currently exist in practice but also included potential new options for co-ordinating care. Therefore, it is not fully clear from the findings presented in this chapter which options of care co-ordination are currently available in practice and which areas may need improvement. However, our findings demonstrate that there are examples of real-world practice for each of the six domains and subdomains presented in this chapter, and existing services may not include elements from each of the six domains at present. Example quotations (see Table 20) provide context regarding the availability of different options in practice. Our study aimed to identify existing and new models of care co-ordination and including both existing examples and new options enabled us to achieve this aim to develop a taxonomy that captures as many options as possible, offering a broader view of findings in relation to potential hypothesised future models of care co-ordination.

Implications

The taxonomy developed in this study can be used as a menu for service planners, researchers and commissioners to consider when developing new and/or existing models of co-ordination. For example, we have used the taxonomy, together with the qualifier findings presented in this chapter, to develop some hypothetical models of care co-ordination that may be applicable in different situations (see Chapter 8). We have also developed a flow chart that may inform how the findings are used to develop such models (see Chapter 8). These models can be costed and evaluated by researchers and services.

The qualifier findings can also be used to inform decisions about which models of care co-ordination may be suitable for use in different situations. This is particularly helpful given the complexity of care pathways and funding for rare conditions.

Future research

Future research is needed to explore the implementation, clinical effectiveness and cost-effectiveness of different models of care co-ordination for rare conditions in practice. This is important, given that it is not yet clear whether or not co-ordinated care leads to better outcomes (e.g. patient outcomes, professional outcomes and organisational outcomes). Further research is also needed to operationalise these models of care co-ordination so that delivery of care co-ordination can be measured.

What was already known?

• There are numerous ways to co-ordinate care for people living with rare conditions.

• There are also several factors to take into account when deciding how best to co-ordinate care, including stakeholder preferences, the benefits and challenges of different models of co-ordination, factors influencing co-ordination, and barriers to and facilitators of co-ordination.

What this chapter adds

• Very little is known about the costs of different ways of co-ordinating care for people affected by rare diseases. This chapter outlines 10 hypothetical models of care co-ordination. These models take into account different factors that influence co-ordination and provide an insight into which models of care co-ordination may suit different circumstances.

Development of illustrative models

Measuring costs

In accordance with our study protocol, to address RQ5 (i.e. ‘how much do different models of co-ordinated care cost?’) we originally planned to undertake preliminary cost analyses of the models of co-ordinated care that we developed. These analyses were to include the cost to set up and implement each model and the cost to run each model. We envisaged that these analyses would not be a formal analysis of the incremental costs of care co-ordination (as such an analysis would not be possible without detailed evidence of the long-term impacts of co-ordination on health outcomes and health-care use), with the focus, instead, on the ‘intervention’ costs associated with setting up and running the different models of co-ordinated care. We envisaged that data for these analyses would be based on data from the survey (see Chapter 5) and data from the workshops used to refine the taxonomy (see Chapter 7). Unfortunately, it was not possible to use either of these sources. In the case of the survey, as shown in Chapter 5, most people did not experience co-ordinated care. In addition, the survey did not ask questions about the ‘intervention costs’ associated with different types of care co-ordination. Furthermore, it was not possible to attribute the 10 hypothetical models that were developed to survey respondents’ experiences. In the case of the workshops, the health service utilisation associated with each hypothetical model was unknown by workshop participants. This was primarily because the use of services associated with each model was likely to vary according to the situation-specific factors described above (see Chapter 8, Development of hypothetical models). The result was that it was not possible to generate costs associated with each hypothetical model from the survey data or from the workshop data.

In an attempt to find indicative costs of the models, we instead undertook a review of the costs of different characteristics of co-ordinated care. We identified from the 10 hypothetical models that costs would likely be incurred for the following characteristics:

• whether or not the patient attended a specialist centre

• whether or not the patient had a care co-ordinator

• whether or not the patient had a care plan (i.e. a formalised care agreement)

• whether or not the patient’s care was discussed in MDT meetings.

We then attempted to identify the costs of each of these characteristics from previously conducted research and administrative data.

We know that evidence on the costs of interventions to co-ordinate care is extremely limited. 219 Nonetheless, to identify what costs were available, we did the following:

1. We adapted the search strategy we used for the scoping review in Chapter 3 (see Appendix 1). We reran the PubMed search, replacing the search terms relating to the type of study (search number #17; the scoping review was a review of reviews) with ‘cost*’ in any field. This meant that the search focused on studies of care co-ordination of rare and chronic diseases where costs were mentioned. We reviewed the titles and abstracts of the identified papers to identify cost analyses of the characteristics of co-ordinated care included in the models. Studies focusing on any condition were included, not just rare conditions. As we are interested in the costs of co-ordination from a UK perspective, we included UK studies only.

2. We searched the NHS Economic Evaluation Database (NHS EED)220 using search terms ‘coord* OR care plan OR specialist centre OR care co-ordinator’. Note that NHS EED includes economic evaluations published up until 31 March 2015 only. We reviewed the titles and abstracts of the identified studies. As above, studies focusing on any condition were included, not just rare conditions, and we included UK studies only.

3. We reviewed commonly used sources of unit costs for undertaking economic evaluations in the UK (NHS reference costs for 2018/19221 and Unit Costs of Health and Social Care 2019222) to identify unit costs for each characteristic. We included unit costs relating to any condition, not only rare conditions.

We tried to identify UK-based ‘intervention’ costs for each of the four characteristics (i.e. attending a specialist centre, access to a care co-ordinator, having a care plan and having care discussed at MDT meeting) and converted these into 2019/20 GBP where necessary.

Illustrative models

We developed 10 hypothetical models of care co-ordination. These models are summarised in Table 22 (see Report Supplementary Material 4 for further details). Each type of model is a function of where the patient and parent/carer lives in relation to a specialist centre, if the patient and parent/carer can, or wants to, travel to a specialist centre, if the patient and parent/carer has the ability (and want) to co-ordinate their own care, if the patient and parent/carer has access to a specialist centre and whether or not it is clear who the patient needs to see for the management of the condition. The characteristics of the models are centred around attending a specialist centre or outreach clinic, having a formalised care agreement (i.e. a care plan), having a care co-ordinator to organise appointments (or providing a point of contact), whether or not there are meetings between HCPs to discuss care and the type of a HCP who oversees care. The 10 models described in Table 22 are essentially combinations of these characteristics. As noted above, the specificity of these characteristics will be determined by situation-specific factors.

Costs of model characteristics

Our reviews of PubMed and NHS EED identified 7254 and 190 hits, respectively. However, after reviewing the titles and abstracts of these reports, we found that evidence on the UK costs of characteristics of co-ordinated care was extremely limited. We found one study11 that calculated the costs of treatment at a specialist centre for a rare condition (Alström syndrome). This study estimated the mean cost per patient per annum, including clinic attendances and contacts with HCPs, plus consumables and capital. The estimated cost was £748 per patient per annum (i.e. £690 per patient per annum in 2015/16 prices). We found one study223 that evaluated the costs of a care co-ordinator (a key worker for disabled children), which included telephone calls and face-to-face contacts with the patient and their family and non-contact time (e.g. writing case notes, travelling, liaising with staff from their own and other organisations, and attending meetings and reviews). The estimated cost was £834 per patient per annum (i.e. £151 for 3 months in 2002/3 prices) for time spent in contact with families, which rose to £1251 to £1668 per patient per annum when including non-contact time. 223 We found estimates for the costs of maintaining and reviewing a care plan – for looked after children – with a total cost of £568 per patient per annum (i.e. £556 per annum in 2018/19 prices). 222 We were unable to find estimates of the UK costs per patient of MDT meetings, but found national average costs of £94–140 per patient per meeting for cancer MDT meetings. 221 With the exception of the cost of attending the specialist centre, it is not clear if these costs are applicable to co-ordinated care for people affected by rare conditions. In addition, we were unable to use these estimates to represent the impact of the situation-specific factors identified in the illustrative models.

Key findings

In this chapter, we developed and refined 10 illustrative models of care co-ordination for rare conditions using our taxonomy. The type of model was a function of where the patient and parent/carer lives in relation to a specialist centre, if the patient and parent/carer can, and wants to, travel to a specialist centre, if the patient and parent/carer have the ability (and want) to co-ordinate their own care, if the patient and parent/carer have access to a specialist centre and whether or not it is clear who the patient needs to see for the management of the condition.

We attempted to calculate the costs of each model using data from the study, but this was not possible. Instead, we reviewed what evidence exists on the UK costs of different components of co-ordinated care, focusing on the costs of attending a specialist centre, the costs of having a care co-ordinator, the costs of having a care plan and the costs of discussing a patient’s management at MDT meetings. Although some UK cost data were found, these data were limited and it was not possible to apply these data to the 10 illustrative models.

Limitations

The main limitation of the research presented in this chapter was the lack of cost data, which prevented us from estimating the costs of the illustrative models. We were unable to use data from the national survey or the taxonomy and so, instead, searched for previously estimated costs from UK-based studies for the key characteristics of the models. In terms of the survey, it was not possible for survey respondents to report the costs of different aspects of care co-ordination, partly because many of these aspects are not experienced by participants and also because the NHS costs incurred or resources used to provide these aspects of care were unknown. For the work on the taxonomy, we had originally envisaged that for each of the models that were developed it would have been possible to provide specific information around the resources needed to provide that model (e.g. the number, type and hours of staff required, and the non-staff resources required). With these data, it would have been possible to undertake a detailed bottom-up costing. However, this information was not known by study participants and so this level of specificity was not possible. There were few UK studies that calculated the costs of care co-ordination and it was unclear if the available data are applicable to the co-ordination of care for rare diseases. In addition, we were unable to use these estimates to represent the impact of the situation-specific factors identified in the illustrative models.

Further research

Further research would be beneficial to produce accurate estimates of the costs of the different elements of co-ordinated care for people affected by rare conditions in the UK. One possible approach would be to find empirical examples of each element of co-ordinated care currently in existence and undertake a detailed bottom-up costing. In addition, further research into the feasibility of adapting existing rare disease services or implementing these hypothetical models into rare disease services in future is needed.

Implications relating to whether or not care for rare diseases is co-ordinated

The UK government recently highlighted the problem of co-ordinated care for people affected by rare diseases in The UK Rare Diseases Framework,9 although the evidence base was largely anecdotal. 9 The UK Rare Diseases Framework9 states that co-ordination of care is one of the top challenges facing people affected by rare diseases and better co-ordination was listed as one of the four top priorities. In a ‘national conversation’ survey of 6293 members of the rare diseases community, conducted in 2019, co-ordination of care was identified as one of the top challenges facing the rare diseases community. Our study findings are consistent with The UK Rare Diseases Framework,9 but add further detail of the extent of these problems. We identified gaps in provision of services that might better facilitate co-ordination, identified aspects of care that are not co-ordinated, provided evidence of the impact of these problems and provided further evidence that people affected by rare diseases have a strong preference for better co-ordination.

Our definition of care co-ordination for rare diseases highlights the complexity of achieving co-ordinated care, indicating that there are several components of care that ought to be addressed to improve co-ordination. This definition serves as a useful guide for researchers, policy-makers and other stakeholders seeking to improve care co-ordination.

Evidence of the lack of co-ordinated care for people affected by rare diseases is supported by our national survey, which found that for the majority of people affected by rare diseases care is not well co-ordinated, with limited access to care co-ordinators, specialist centres and care plans.

The importance of the finding from our national survey that people affected by rare diseases have limited access to co-ordinated care was made clear by our exploratory qualitative interview study, which highlighted how patients and carers are negatively affected by poorly co-ordinated care in terms of their physical and mental health and their financial well-being.

The importance of co-ordinated care was further strengthened by the findings of our taxonomy and our analysis of preferences, which showed that patients, parents/carers and HCPs all have a clear preference for better co-ordinated care. This preference related to all the aspects of co-ordinated care that were considered.

Implications relating to the ways in which care for people with rare diseases might be co-ordinated

As well as providing further evidence of lack of care co-ordination, our findings also have implications for the ways in which care for people with rare diseases might be co-ordinated and propose new models of care co-ordination. Following the publication of The UK Rare Diseases Framework,9 the four devolved nations of the UK will develop action plans that set out how the four priorities identified in the framework will be addressed (note that the third priority is ‘better co-ordinated care’). Our findings around the hypothetical models of co-ordination have potentially useful implications for this work, for example by being included in national and local action plans, and for others seeking to improve co-ordination of care.

Our definition of care co-ordination and description of the components of care co-ordination can be taken into account when considering how to improve co-ordinated care.

The taxonomy developed in this study can be used as a menu for service planners, researchers and commissioners to consider when developing new and/or existing models of co-ordination. In addition, the qualifier findings from the taxonomy can also be used to inform decisions about which models of care co-ordination may be suitable for use in different situations, accounting for the preferences of stakeholders. This is particularly helpful given the complexity of care pathways and service funding for rare conditions.

We have developed a flow chart that may inform how the findings from the taxonomy may be used to develop such models, and their potential costs.

The trade-offs from the DCE could be used to value the potential benefits of different models of care co-ordination. For example, the willingness to pay for each aspect of care co-ordination is a measure of the value of the benefit of each aspect on average per patient. The willingness to pay could be summed across all patients receiving that aspect of care and balanced against the total costs of providing that aspect of care in a future cost–benefit analysis.

Strengths

There are several strengths of our study. The study used a combination of qualitative and quantitative research methods to better understand (1) what co-ordinated care for people affected by rare diseases means and (2) how to address the problem of poorly co-ordinated care. Input into addressing the RQs was obtained from a range of stakeholders, including patients, parents, carers, HCPs from a range of occupations and clinical backgrounds, commissioners and patient organisations. We undertook a large and comprehensive national survey, covering a wide range of people affected by rare diseases. The study had a strong patient and public involvement (PPI) component throughout all aspects of the study. Members of Genetic Alliance UK, PPI co-applicants on the study team and the study PPIAG, inter alia, advised on the content and wording of the survey, gave feedback on the findings of the scoping review and taxonomy, and advised on dissemination opportunities.

Limitations

There were several limitations to the study. These limitations are described in detail in each chapter, but the main limitations were as follows.

First, it was not possible to capture the experiences of people affected by every rare condition. However, the principles of care co-ordination, and what these ought to entail, are likely to be common across many rare conditions, which means that it was still feasible to make recommendations about what co-ordinated care for rare conditions should involve. We ensured that we included a wide range of models of care co-ordination by involving many different stakeholders during the study and eliciting their views throughout. These stakeholders included patient organisations/charities that work on behalf of people affected by rare conditions, patients and carers affected by rare conditions, HCPs caring for people with rare conditions, commissioners and providers at local and national levels who commission/provide health services for people affected by rare conditions, academics interested in the organisation of care for people with rare conditions and policy-makers with an interest in rare conditions.

Second, our sampling for both the qualitative and quantitative aspects of the study may have been biased if study participants were systematically different from the population affected by rare conditions. For our national survey and DCE in Chapters 5 and 6, it was not possible to determine the representativeness of the sample and, therefore, the generalisability of the results. We had high proportions of patients and parents/carers who were female and from more highly educated groups, which is likely to indicate that these groups were over-represented, but the true extent of this is difficult to quantify. There is a paucity of evidence regarding the total number of people living with rare conditions in the UK, and their characteristics, such as gender, age distribution, ethnicity, level of education and socioeconomic status, are unknown. Moreover, the methods used to recruit meant that people without links to a patient organisation that was known to the study partners, and who do not use NHS hospital services, were likely to be under-represented in the survey respondents. For the taxonomy and models in Chapters 7 and 8, although we sampled from a variety of rare conditions, locations, sectors and populations, some groups are likely to have been under-represented in our research and, therefore, we are unlikely to have captured every possible option of care co-ordination for rare conditions. Although we used the wealth of experience at Genetic Alliance UK in undertaking research among people affected by rare diseases throughout the study, and worked with the PPIAG to explore the right methods for including hard-to-reach patients and families affected by rare conditions, our sampling strategy may have introduced bias into our results and conclusions.

Third, our review of reviews in Chapter 3 was a scoping review and, therefore, was unlikely to have captured every relevant study. To identify as many studies as possible, we conducted a comprehensive search that included contacting experts and searching the reference lists of included reviews.

Fourth, our qualitative study in Chapter 4 was based on only 15 participants.

Fifth, in terms of the DCE, this method elicits hypothetical choices and, therefore, might lack external validity if individuals do not make the same choices in real-life situations. We also acknowledge that preferences from the DCE might vary by subgroups within our study groups, but sample size considerations make subgroup analyses problematic.

Sixth, our cost analysis in Chapter 8 was limited in scope because of the paucity of available data, and there is considerable uncertainty in the costs associated with different co-ordination models.