Codesigning a Measure of Person-Centred Coordinated Care to Capture the Experience of the Patient: The Development of the P3CEQ.

BACKGROUND: Person-centred coordinated care (P3C) is a priority for stakeholders (ie, patients, carers, professionals, policy makers). As a part of the development of an evaluation framework for P3C, we set out to identify patient-reported experience measures (PREMs) suitable for routine measurement and feedback during the development of services.
METHODS: A rapid review of the literature was undertaken to identity existing PREMs suitable for the probing person-centred and/or coordinated care. Of 74 measures identified, 7 met our inclusion criteria. We critically examined these against core domains and subdomains of P3C. Measures were then presented to stakeholders in codesign workshops to explore acceptability, utility, and their strengths/weaknesses.
RESULTS: The Long-Term Condition 6 questionnaire was preferred for its short length, utility, and tone. However, it lacked key questions in each core domain, and in response to requests from our codesign group, new questions were added to cover consideration as a whole person, coordination, care plans, carer involvement, and a single coordinator. Cognitive interviews, on-going codesign, and mapping to core P3C domains resulted in the refinement of the questionnaire to 11 items with 1 trigger question. The 11-item modified version was renamed the P3C Experiences Questionnaire.
CONCLUSIONS: Due to a dearth of brief measures available to capture people's experience of P3C for routine practice, an existing measure was modified using an iterative process of adaption and validation through codesign workshops. Next steps include psychometric validation and modification for people with dementia and learning difficulties.

Background

Measuring patient experiences of health-care delivery is now recognized as a valuable mechanism to evaluate the quality of care from the perspective of the patient (1,2). Such measures are often included in the evaluation of new models of care aimed at improving both the outcomes and the quality of care for people with long-term conditions (LTCs) and multiple long-term conditions (MLTCs) (3 –5).

People with LTCs and MLTCs often describe their care as fragmented and lacking in continuity (6 –8) and call for better coordination of their care (9). Individuals also often feel that their wishes and preferences are not taken into consideration (10 –20). These experiences can be described as non–person-centred care (PCC), which is often exacerbated by poor coordination (10,21,16). A recent systematic review identified care coordination, clinical guidelines, (non-)PCC, and shared decision-making as critical areas that require improvement in people with LTCs (22).

Person-centred care can be described as “care where individuals’ values and preferences are elicited to guide all aspects of their health care, supporting their realistic health and life goals” (23, p. 16). To recognize the importance of care coordination for the delivery of PCC for people with LTCs and MLTCs (23,24), we brought these concepts together in a single operating framework; what we call person-centred coordinated care (P3C). Our framework (25) has been influenced by the work of Ekman et al who have identified 3 core practice routines for the delivery of PCC, which emphasize personhood, equality, and personalized outcomes (26): through the (a) elicitation of a personal narrative, (b) the establishment of a partnership and treatment goals, and (c) the documentation of this in a care plan (26,27). We emphasize a fourth routine “agreeing to act with the person and their family to coordinate the care plan” (25, p. 24). For P3C to be achieved, we argue that activities within the following 5 core domains need to take place: “person-centred information and communication,” “the identification of goals/outcomes and what is important to the person,” “person-centred care planning,” “the management of health states and service transitions,” and “collaborative/shared decision making” (9,13,28,29,30).

Our domain model was created to support services develop new models of person centred coordinated care (P3C). According to these core domains, an accompanying evaluation framework probes multiple perspectives (ie, that of patients, staff, and the organization) (31). In developing this framework, our initial aim was to identify a measure capable of probing if care received is both person-centred and coordinated, from the perspective of the patient. However, due to the lack of suitable measures available, we modified our aim and developed an existing measure suitable for this purpose using stakeholder codesign.

Methods

Our methods included a rapid review, stakeholder codesign workshops and measurement modification, and cognitive testing of the modified measure.

Rapid Review

A rapid review (32) was undertaken to identify candidate measures that capture experiences of P3C (Figure 1). We searched Pubmed, Medline, and Cochrane databases of systematic reviews for recently published literature on the topic using the terms “integrated care,” “coordinated care,” “person-centred care” and “continuity of care” and “systematic/review” as keywords. Seven relevant systematic reviews were identified of which 2 were on integrated care (33,34), 2 on continuity of care (35,36), 2 on person-centredness (37,38), and 1 on care coordination (39). Reports from key organizations such as the Health Foundation (40), Picker Institute Europe (41), Agency for Healthcare Research and Quality (42), Department of Health (DH) (43), and gray literature were considered next. Outcome measure databases such as the database by the Patient-Reported Measurement Group of Oxford University (http://phi.uhce.ox.ac.uk) and Patient-Reported Outcome and Quality of Life Instruments Database (PROQOLID) (http://www.proqolid.org/) were also searched. Finally, measures that were used or currently being used by new care models were taken into account (44).

Figure 1.

A schematic representation of the rapid review.

Measures were included if they covered the patient’s perspective (a) of PCC and/or (b) their experiences of coordinated/integrated care. Of the measures that were identified, 71 were considered for inclusion (Appendix A) of which only 7 met our inclusion criteria (see Table 1).

Table 1.

Brief Outline of the Selected Measures.

Name of Measure	No. of Items	Country of Development	Aim of Measure
1. Patients Perceptions of Integrated Care Survey (PPIC) (45)	80	Harvard School of Public Health, United States	To measure patient’s perception of integrated care
2. Patient Perceived Continuity from Multiple Clinicians (CC-MC) (46)	34	Canada	To measure continuity of care from the perspective of patients who regularly see more than 1 clinician
3. Patient Assessment of Chronic Illness Care (PACIC) (47)	20	United States	To measure the extent to which patients receive care consistent with the chronic care model (CCM)
4. Client Perception of Coordination Questionnaire (CPCQ) (48)	31	National Centre for Epidemiology and Population Health, Australia	To measure coordination of health care
5. Long-Term Condition 6 (LTC6) questionnaire (43)	6	DH Quality, Innovation, Productivity and Prevention (QIPP) LTC team, United Kingdom	To measure patient’s confidence in services/care given and own abilities to self care
6. Patient-Service User Questionnaire—Integrated Care Evaluation Pilots (44)	26	United Kingdom	To evaluate patient experiences of integrated care
7. North West London Integrated Care Pilot: Patient Survey (49)	19	Integrated Care Pilot team in North West London, United Kingdom	To measure patient’s perception of integrated care

Reasons for excluding measures were (a) not aimed at LTCs or older people (eg, generic national surveys), (b) only 1 domain of integrated or PCC covered (eg, communication, empathy), (c) minimal questions of relevance to P3C, (d) covering only episodic care (eg, hospital visit), (e) condition specific (eg, chronic pain, cancer), or (f) measure still in development.

To critically examine included measures, these were mapped to the core domains of our P3C framework. The identified domains and questions were then compared with existing patient experience frameworks (Institute of Medicine framework (50), the Picker Institute framework (51), Warwick Patient Experiences framework (52), and NICE guidelines (53)) to identify missing constructs.

Codesign Workshops

The 7 included measures were presented to stakeholders in a series of workshops to explore their acceptability, utility, and strengths/weaknesses. The first workshop with 12 attendees comprised patient and carers (patients n = 3; female aged 70, male aged 67, female aged 57; carer n = 1, female aged 65), commissioners (n = 2), clinical staff (n = 2), social care representatives (n = 1), and the academic research team (n = 3) and lasted for 3 hours. First, each individual measure was discussed in detail in small groups. The top 3 measures were then selected by a final voting process and agreed by all.

The top 3 measures were then presented and discussed in further 3 workshops. The first of these had a mixed representation of stakeholders (n = 11) involving patient representatives (n = 3; 1 male aged 58, 2 females aged 57 and 68 years), commissioners (n = 1), clinical staff (n = 2), social care representatives (n = 2), and academic research team (n = 3). The final 2 were organized solely with patient representatives (n = 5 and n = 7; age ranged: 58-80, 7 females and 5 males). All workshops lasted for an hour and a half. Stakeholders had an opportunity to comment on each of the 3 questionnaires. Discussions focused on the content and language used in the measures but also relevance to lived experience, length, and layout. All patient representatives had LTCs or MLTCs and were identified through general practice patient participant groups. Their participation was a patient representatives and codesign experts and not as research participants.

Measurement Selection and Modification

The preferred measure was subject to the modifications that focused on adding extra questions as per the feedback obtained from codesign workshops. The Modified Long-Term Condition 6 (LTC6; renamed Person-Centred Coordinated Care Experiences Questionnaire [P3CEQ]) underwent several iterations based on feedback from the team, other researchers working in P3C, and our codesign workshops where questions were redesigned by patients.

Cognitive Testing of the Modified Version

The next step was refine the modified measure through cognitive interviews with patient representatives. This is a method of collecting additional information on a questionnaire from end users which is then used to determine whether each question is generating the response/information that is intended by the developers (54). It explicitly focusses on the cognitive process that a person uses to answer questions which may include comprehension of question, retrieval from memory of relevant information, judgment, and estimation process and response process (ie, mapping answers to response options) (55). It also seeks to ensure that the questions are meaningful to those the measure is aimed at.

Patient representatives were approached through a local public engagement coordinator and 5 participants (4 males, aged 55-69, and 1 female, aged 67) volunteered to take part in these interviews. The participants either had an LTC such as diabetes and renal problems or cared for someone with health (physical and mental) difficulties. Verbal consent was obtained and the interviews were audio recorded. T.S., a trained interviewer conducted these interviews, and each interview lasted around 1 hour. Monetary compensation was provided for their time. Both the “think aloud” and “verbal probing” techniques were used throughout to comprehend how participants were responding to each questions (54,55).

Results

Our review work identified 13 important subdomains of patient experience which fall within our 5 core domains (28): (a) goal setting, (b) empowerment/activation, (c) self-management, (d) carer involvement, (e) care plan, (f) case manager/key person, (g) single point of contact, (h) care coordination (i within teams; ii across teams), (i) continuity of care (eg, regular appointment, follow-up), (j) involvement in decision-making, (k) relational continuity, (l) information gathering/sharing, and (m) knowledge of patient/familiarity.

Each item from the 7 included measures (see Table 1) was mapped to these domains and subdomains by 2 reviewers (T.S. and H.L.) independently to ascertain that if the measure probed that subdomain, any disagreements were resolved on discussion.

This process examined the extent a measure probed P3C, according to the definition and core domains we identified. We also sought to identify the extent to which a measure had been psychometrically validated. The mapping process for the selected 7 measures is presented in Table 2. This table depicts if the measure probes that domain and subdomain and the number of questions in each. The top row details the domains with the subdomains presented underneath.

Table 2.

Item Mapping of the Selected Measures to Core Coordinated Care Domains.

No.	“I” Statements	My Goals/Outcomes				Care Planning					Transitions	Decision-Making	Information and Communication
	Person-Centred Coordinated Care Domains	Goal setting	Empowerment/activation	Self- Management	Carer Involvement	Care Plan	Case Manager/Key Person	Single Point of Contact	Care Coordination		Continuity of Care	Involvement in Decision- Making	Relational Continuity	Information Gathering/Sharing	Knowledge of Individual/Familiarity
									Within Teams	Across Teams
1	Patient Perceptions of Integrated Care (PPIC)	✓ (2)	✓ (11)	✓ (13)	✓ (1)	✓ (1)	✓ (5)	✓ (4)	✓ (3)	✓ (8)	✓ (12)	✓ (2)	✓ (12)	✓ (8)	✓ (12)
2	Patient Perceived Continuity from Multiple Clinicians (CC-MC)	✓ (3)	✓ (14)	✓ (1)	–	✓ (6)	✓ (3)	✓ (4)	✓ (7)	✓ (12)	✓ (12)	✓ (12)	✓ (1)	✓ (11)	✓ (6)
3	Patient Assessment of Chronic Illness Care (PACIC)	✓ (3)	✓ (7)	✓ (3)	–	✓ (3)	–	–	✓ (1)	✓ (1)	✓ (5)	✓ (2)	✓ (2)	✓ (7)	✓ (2)
4	Client Perception of Coordination Questionnaire (CPCQ)	–	✓ (7)	✓ (2)	✓ (3)	–	✓ (1)	–	✓ (2)	✓ (6)	✓ (5)	✓ (5)	✓ (2)	✓ (1)	✓ (1)
5	Long-Term Condition 6 (LTC6) questionnaire	✓ (1)	✓ (2)	✓ (3)	–	–	–	–	–	✓ (1)	–	✓ (1)	–	✓ (2)	–
6	Patient-Service User Questionnaire—Integrated Care Evaluation Pilots	–	✓ (5)	✓ (5)	–	✓ (3)	–	✓ (1)	✓ (4)	✓ (3)	✓ (4)	✓ (3)	✓ (1)	✓ (8)	✓ (5)
7	North West London Integrated Care Pilot: Patient Survey	✓ (1)	✓ (5)	–	–	✓ (5)	–	–	✓ (3)	✓ (3)	–	✓ (5)	✓ (2)	✓ (1)	✓ (1)

Codesign Workshops and Measurement Modification

Our codesign workshops selected the LTC6 questionnaire as the most preferred for routine measurement due to its content, short length, utility, and tone. The other measures were rejected due to length and/or because they lacked a person-centred focus.

In response to our codesign workshops, we modified the LTC6 (43). As presented in Table 2, the focus of the questionnaire was on self-management (including goal setting and empowerment/activation). There were 2 free-text questions on how care could be improved and what support is needed to make them feel more confident. Initial development data showed that the LTC6 has been used with 1592 individuals across 29 sites participating in the Quality, Innovation, Productivity and Prevention LTC Work stream in September 2012, with no apparent floor and ceiling effects (unpublished DH report).

The Modified LTC6 (renamed P3CEQ) underwent several iterations based on feedback from the team, other researchers working in P3C, and our wider codesign group. The modifications are described below:

The term “well-being” was added to “health” to questions 1, 3, 4, and 6 to ensure the questionnaire was applicable to both health and social care settings.

As per (i), the terms “treatment” in question 2 and “support” in question 5 were replaced with the term “care.”

The original question 3 on the amount of information received by the person was modified to reflect not only the quantity of the information but also its perceived usefulness.

A comments section was added to each question to allow individuals to expand their views or provide an example in their own words.

Eight new questions were included that probed: “being considered as a ‘whole’ person rather than a disease/condition,” “repeating information that should have been in care records,” “carer involvement”; “a single named person to coordinate the care,” and 4 questions on care planning (covering availability, accessibility, utility, and applicability).

Cognitive Interviewing

We used cognitive interviewing to understand if the modified LTC6 measure was understandable to people with LTCs and MLTCs. We also wanted to ensure that specific terms such as “well-being,” “whole person,” “plan of care,” “joined up care,” “coordinated care,” and “information” were meaningful to people. Positive feedback from the interviews related to the relevance of the questionnaire to P3C, overall language, response codes, length, and the time taken for completion. Minor suggestions were made in relation to layout, formatting, ordering of questions, and language. Major changes included in the reduction in the number of main questions to 11. This was achieved by including a trigger question on care plans as part of the main questions, and 3 further care plan questions that would be applicable only to people who respond positively to the initial care plan question.

The positive feedback from the cognitive interviews and from the ongoing stakeholder engagement in terms of relevance to P3C and understanding of the questions ensured face validity (ie, the measure was investigating a person’s perspective on whether care was experienced as person centred and coordinated). Throughout the 12 iterations of the P3CEQ, items within the revised version were continuously being mapped to the core domains of our model to maintain the relevance of the measure for capturing peoples experiences of P3C and thus ensuring content validity. How the original and modified LTC6 questionnaire compared to the core P3C domains is presented in Table 3.

Table 3.

Comparison of the Original and Modified LTC6 against Core P3C Domains.

No.	“I” Statements	My Goals/Outcomes				Care Planning					Transitions	Decision-Making	Information and Communication
	Person-Centred Coordinated Care Domains	Goal Setting	Empowerment/Activation	Self- Management	Carer Involvement	Care Plan	Case Manager/Key Person	Single Point of Contact	Care Coordination		Continuity of Care	Involvement in Decision Making	Relational Continuity	Information Gathering/Sharing	Knowledge of Individual/Familiarity
									Within Teams	Across Teams
1	Long-Term Condition 6 (LTC6) questionnaire	✓ (1)	✓ (2)	✓ (3)	–	–	–	–	–	✓ (1)	–	✓ (1)	–	✓ (2)	–
2	Person-Centred Coordinated Care Experiences Questionnaire (P3CEQ)	✓ (2)	✓ (3)	✓ (4)	✓ (1)	✓ (4)	✓ (1)	–	✓ (3)	✓ (3)	✓ (1)	✓ (2)	✓ (3)	✓ (2)	✓ (1)

The P3CEQ

The P3CEQ has 11 questions and is free to use on completing a user agreement form. The P3CEQ has been translated into 5 European languages and is now being used across a range of national and international evaluations of new models of integrated care and self-management (http://www.selfie2020.eu/ and http://www.sustain-eu.org/).

Although the current FKG level of 7.3 is appropriate for most individuals, we recognize that it may not be suitable for people with learning difficulties. Feedback that the measure may not be suitable for completion by people with dementia and cognitive difficulties has also been taken into consideration. To adjust for these difficulties, we are exploring the use of Talking Mats, a pictorial framework that has been validated for this subset of this population (45).

Discussion

The importance of measuring the individual patient’s perspective of P3C is widely acknowledged in practice, research, and policy. In response to repeated calls for the need of patient-reported experience measures, there has been a rapid surge in the development of new measures. Hence, this study was initiated as a review with the aim of identifying a measure that could be used in routine practice to evaluate P3C experience from the perspective of the patient. Despite a large number of measures identified, many failed to cover the core domains of P3C or were too long for routine practice improvement projects. This finding was in line with the findings of the recent relevant systematic reviews (33 –35,37,39). Due to the lack of a suitable measure, we adapted the LTC6 to cover the core domains of P3C in a timely and cost-effective manner using codesign to produce the P3CEQ.

A particular strength of our work is in the ongoing involvement and validation by key stakeholders throughout the process of measure selection and modification. This has ensured that the focus of the work is person centred and is relevant to the end user. Recent reports/articles have confirmed the domains of interest probed by the P3CEQ as key constructs of P3C (23,24). The mapping of the items from the modified questionnaire to these core P3C domains has ensured that the measure is strongly aligned with new models of care. Finally, the use of cognitive interviewing (50) has helped refine the questionnaire, making it more user-friendly. The P3CEQ has completed wide-scale psychometric testing, the results of which will be published imminently.

Conclusion

The P3CEQ is a brief, generic measure that covers core domains of person-centred coordinated care from the perspective of the patient. This measure therefore offers an opportunity for services to develop P3C by acting on the evaluative statements of those receiving care and support. This will only be achieved, however, if the data from it are fed back to practice to drive change, a short easy to administer measure such as the P3CEQ advances progress toward this aim.

Click here for additional data file.

Supplemental Material, Appendix_1_Px for Codesigning a Measure of Person-Centred Coordinated Care to Capture the Experience of the Patient: The Development of the P3CEQ by Thavapriya Sugavanam, Ben Fosh, James Close, Richard Byng, Jane Horrell, and Helen Lloyd in Journal of Patient Experience