A three‐goal model for patients with multimorbidity: A qualitative approach

Abstract Background To meet the challenge of multimorbidity in decision making, a switch from a disease‐oriented to a goal‐oriented approach could be beneficial for patients and clinicians. More insight about the concept and the implementation of this approach in clinical practice is needed. Objective This study aimed to develop conceptual descriptions of goal‐oriented care by examining the perspectives of general practitioners (GPs) and clinical geriatricians (CGs), and how the concept relates to collaborative communication and shared decision making with elderly patients with multimorbidity. Method Qualitative interviews with GPs and CGs were conducted and analyzed using thematic analysis. Results Clinicians distinguished disease‐ or symptom‐specific goals, functional goals and a new type of goals, which we labelled as fundamental goals. “Fundamental goals” are goals specifying patient's priorities in life, related to their values and core relationships. These fundamental goals can be considered implicitly or explicitly in decision making or can be ignored. Reasons to explicate goals are the potential mismatch between medical standards and patient preferences and the need to know individual patient values in case of multimorbidity, including the management in acute situations. Conclusion Based on the perspectives of clinicians, we expanded the concept of goal‐oriented care by identifying a three‐level goal hierarchy. This model could facilitate collaborative goal‐setting for patients with multiple long‐term conditions in clinical practice. Future research is needed to refine and validate this model and to provide specific guidance for medical training and practice.

Multimorbidity is defined as the coexistence of two or more chronic diseases or conditions, and its incidence is causing a challenge to health-care systems, patients and medical practitioners. In daily medical practice, multimorbidity challenges decision making in several ways.
Disease priorities can be interfering 8,9 and the need of adherence to multiple disease guidelines can be problematic. 10 Disease-specific guidelines are often not applicable to older patients with multiple conditions 7,9,11,12 and compliance to multiple single disease guidelines can result in polypharmacy, high treatment burden, inattention to social and personal context and failure to align care with personal goals and preferences. 13,14 Having multiple chronic conditions often leads to the involvement of several clinicians, who concentrate on managing different conditions and monitoring different disease-specific outcomes.
Patients are at risk of receiving fragmented care that might lack focus on what matters most to them. 7 Focusing care on what matters most to patients could be helpful but is also a challenge in itself.
Aligning health outcomes with individuals' values is complex, especially for older adults with multiple chronic morbidities facing conditions with clinical uncertainty (eg cancer). [15][16][17] Clinicians are challenged in helping people prioritize their values, define treatment goals and frame preferences in ways that are clinically relevant and aligned with one's values when faced with multiple diagnostic and treatment options. 18,19 Patients and clinicians may also differ in perspectives and priorities in this respect. 6,9,16,[19][20][21][22] Clinicians are at risk to make inaccurate assumptions about patient values and preferences 3 and may think that they know what is best for patients. 10 It is clear that in care for older patients with multimorbidity, incorporating values and preferences in decision making is necessary to focus on what matters most to them, but in daily practice this is complex.
To meet the challenges of multimorbidity care, including the necessary incorporation of values and preferences, Tinetti et al 2 proposed a shift from a disease-oriented to a goal-oriented approach. Taking this approach, it seems health goals can be defined using a range of dimensions (eg symptoms; physical functional status, including mobility; and social role). In goal-oriented health care, care is personalized to accommodate patients' goals, preferences and resources. 3,23 Collaborative goal-setting (CGS), defined as "a process by which health-care professionals and patients agree on a health-related goal", 15 can be useful for personalizing care and encourages patient involvement in the goal-setting process. CGS has been evaluated in several rehabilitation settings. [24][25][26][27] However, CGS in the context of older patients with multimorbidity is not common practice yet. In the 2014 Commonwealth Fund Survey of adults aged 65 or older and having a chronic condition, rates of respondents reporting the sharing goals with a professional varied from 23% (Sweden), till 59% (United Kingdom). Nine of 11 countries were having rates of less than 50% of respondents reporting the sharing goals with a professional. 28 There appears to be a relative lack of insight in goal-setting processes in the presence of complexity and little evidence to support best practices in goal-setting with complex patients. 29 Furthermore, as concluded by Knight et al, the concepts of values, goals and preferences are often used interchangeably, 30 indicating a need for establishing consistent definitions.
In the Netherlands, the current views of general practitioners (GPs) and clinical geriatricians (CGs) on the concepts of goals and CGS are as yet unknown. These views could provide valuable input into the concept of goal-orientation and into the perceived relevance of the approach in clinical practice. The objective of this study was to examine the concept of goal-orientation from a clinician's perspective, in the context of CGS and shared decision making (SDM), where patients have multiple long-term conditions.

| Participants
This qualitative study was conducted by inviting CGs and GPs to participate in semi-structured interviews. Participants were selected using a purposive and snowball method, aiming to recruit professional experts, and contacted by email. We invited experienced GPs and CGs performing research, teaching, developing or implementing specific innovations in care for older patients. The first two participants were acquaintances of the interviewer (first author). Some GPs were recruited at a meeting of GPs holding a specialization in geriatric care.
In the sampling, we aimed to recruit comparable numbers of CGs and GPs. To obtain diverse perspectives, we tried to ensure that different types of practice and practice location (rural or urban) for GPs and different types of hospitals (CGs) would be represented. Furthermore, we tried to ensure that all Dutch regions would be represented as much as possible. The response rates of CGs and GPs approached were 86% and 54%, respectively. The final sample consisted of 18 CGs and 15 GPs.

| Procedures
An interview guide (Table 1) was inspired by two perspectives on goal-oriented health care for elderly patients with chronic multimorbidity 2,3 and professional experiences (general practice and clinical geriatrics) in our research team. Two pilot interviews were conducted with a CG and a GP. Main topics and subtopics were not changed based on the pilot interviews nor during the conducting of the interviews. The interview guide covered three main topics: CGS, SDM and effective collaborative action. We defined effective collaborative action as clinicians and patient deciding on and performing diagnostic and treatment steps in line with collaborative goals, which were set between patient and clinicians or with other involved caretakers. Definitions were not given to the interviewees. At the start of the interview, the clinicians were asked to use the context of regular care for community-dwelling older patients (age >75 years) with a chronic disease or multimorbidity without further specifications. It was also suggested to keep one or more cases in mind in answering the questions. All topics and subtopics were covered in all interviews.
Interviews could differ in asking further questions for a better understanding of an interviewee's answer.
The first author, who is trained as a GP, conducted the interviews between November 2012 and April 2013. The interview duration was approximately 60 minutes and they were conducted face-to-face or by telephone. All interviews were audio-recorded and transcribed.
Detailed field notes were made after each interview. Theoretical memos were drafted throughout the data collection and analysis process. The two final interviews confirmed theoretical saturation as they did not reveal new issues or topics.

| Analyses
Inductive thematic analysis was used for analysis. 31 Thematic analysis is an approach for qualitative research focusing on identifying, analyzing and reporting patterns (themes) within qualitative data and the interpretation of aspects of the research topic. In an inductive approach, themes are data-driven. An iterative process of interviewing and analysis was followed. During the interviewing phase, preliminary analyses were conducted based on reflections and discussion of the interviews (first and last author) and by constantly comparing the interviews with the field notes. These preliminary analyses were conceived in theoretical memos, and the interview guide was continually adapted to reflect emerging insights.
In the coding process, data were conceptually interpreted and labelled accordingly. The two data coders (first and second author) applied open coding to the first five transcripts. Initial codes were compared, discussed, grouped and categorized to develop an initial coding tree. The first five interviews were coded independently by both data coders. The remaining interviews were coded by one researcher (second author) and checked by the other (first author).
In weekly meetings, the researchers (first and second author) compared, discussed and agreed on the coding of the transcripts, including the creation of additional codes and further refinement of categories and subcategories. Similarities, differences, regularities and patterns were interpreted and discussed to identify themes and to generate hypotheses. Illustrative quotations were selected to underpin and illustrate our findings. In addition, Box 1 presents two case examples, one of a GP and one of a CG to illustrate daily practice of this topic.
The quotations were translated from Dutch into English by a professional translator. The translator and first author discussed the translations to ensure that proper meaning of words and nuances were kept in the translation process. For data coding and analysis, Atlas-ti 7.1.15, (GmbH, Berlin, Germany), was used.

| Quality assurance
The Consolidated Criteria for Reporting Qualitative Research (COREQ) and a 15-point checklist for thematic analysis by Braun and Clarke were used for design, performance and reporting. 31,32 Appendix S1 reports on these COREQ criteria in relation to our research. All interview topics were analyzed in one process to secure consistency and theoretical interrelatedness.

| RESULTS
Participating GPs' (n = 15) mean age was 51 years, being 40% male and on average having 16 years of professional experience. Participating CGs (n = 18) had a mean age of 48 years, being 50% male, and having on average 10 years of professional experience. Further participants' characteristics are presented in Table 2. Three themes were identified (Box 2).

| Clinicians draw distinctions between different types of goals
From the data, three types of goals were identified, that is diseasespecific or symptom-specific goals, functional goals and a third category labelled as fundamental goals. "being of help to others and/or society", and "no wish for changes".

| The consideration of fundamental goals
The practitioners differed in their consideration of fundamental goals, creating three orientation categories, that is (i) no consideration of fundamental goals, (ii) implicit consideration of fundamental goals and (iii) explicit consideration of fundamental goals.

| No consideration of fundamental goals
Practitioners in this category mentioned a primary focus on functional goals and/or disease-specific or symptom-specific goals. Functional goals and disease-specific or symptom-specific goals can be connected to each other as described by CG_12:

Box 1 Two Case Examples from Daily Practice
Case Example One GP_10 spoke about a patient of over 90 years old whose hip surgery had failed.
[The prosthesis] got infected, so her hip had to be removed (…) She was admitted to a nursing home (…), but she really wanted to return home. I understood why, because she had an unusual background. She had been interned in a concentration camp years before that (…). All she really wanted was to go home, because that was the only place she felt safe (…). Everything around her reminded her of her traumatic experiences (…). She actually returned to her apartment in that severely disabled state. But she coordinated all her care and assistance there (…) and lived for years, in fact. Naturally, this is an extreme case, but if you look at the patient's circumstances and history, it is completely understandable (…) and her final years were wonderful. Yes, they were.
Case Example Two CG_11 spoke about a patient who was referred by the GP because of abdominal pains, whereas this patient had been screened by the internist 3 years earlier revealing no major diseases.
The GP still was not sure: Couldn't there still be a malignancy, isn't there anything else still? He did not have a conversation with the patient asking: "If we refer you to that hospital now, what would be your goal? And what is your goal in life in general?" (…) I came to an agreement with (…) the patient: "OK, we are going to do some examinations" (…), but we also immediately talked about: "What would you actually want?" And then she said: "I really just want the abdominal pains to go away" (…) She was very clear about her concerns: "It is not my main concern whether there is a malignancy or not." (…)Then you talk it over in a conversation with the patient. If you've set that goal for yourself: "Now, how far do we want to go to see if we can help you get there?" And together you decide that, at this moment, a colonoscopy and a gastroscopy are really too much for the patient. And yes, a patient then accepts that certain issues cannot be completely figured out.
But we do as much as we can to help her achieve her goals.
Note: These are two case examples from daily practice that show the importance of aligning care with patients' personal history, values and priorities and its difficulties.

Those [patients] usually come to me with problems (…).
Their complaints vary from "more trouble walking" to "tiring out faster", "forgetfulness", "falling" and a whole range of other problems. You try to unravel all their problems and often come back to their medical diagnosis.
At that point, you try to figure out how you can help.
But the foundation is still the patient's functioning (…).

(CG_12)
The practitioners in this category did not mention setting or taking into account fundamental goals.

| Implicit consideration of fundamental goals
Practitioners in this category were aware of fundamental goals.
However, these goals were presumed but not made explicit in a discussion with the patient. GP_04 illustrates that they are aware of im-

| Explicit consideration of fundamental goals
The third category constitutes practitioners who have an orientation towards disease-specific and/or functional goals, while explicitly This theme "The consideration of fundamental goals" makes clear that although aspects of implicit fundamental goals may be taken into account, discussion and consideration of aspects of explicit fundamental goals, are not regular practice yet. Table 3 provides several quotations of questions asked to elicit fundamental goals, as mentioned by the practitioners. These practitioners' examples were transformed by the authors into possible questions, which may be helpful to use in clinical practice to start a discussion on fundamental goals.

| The relevance of explicit goals for decision making
The analysis revealed several reasons to explicate fundamental goals.
The patient's preferences are not always in line with medical standards, nor with the preferences of the practitioners involved, as is illustrated by CG_03: I really do believe that care will become better for the patient, that they will finally get the care they want instead of the care that guidelines, or we together, say they must be

given, whereas that is not what they want. (CG_03)
Secondly, patients' preferences may vary. For example, CG_17 describes differences in medication preferences in a case of dementia:

Example questions Quotations
How do you see your future? How would you prefer to plan it?
GP_15: I mean, you have to consider how these individuals see their future (…) and how they prefer to shape that future….

| Main findings
The case examples (Box 1) demonstrate the importance and difficulties of aligning care with patients' personal histories, values and priorities.
Our analysis revealed three types of goals: disease-specific or symptomspecific goals, functional goals and a third type of goals, which we labelled fundamental goals. From our analysis followed that fundamental goals are implicitly and explicitly applied in daily practice. We hypothesize that the explicit setting and application of fundamental goals could lead to patient-specific clinical decisions concerning diagnostic trajectories or treatments by translating values, personal history and core relationships into useable reference points for decision making.

| Interpretation
Earlier studies confirmed our findings. Maintaining (functional) independence, fixing specific symptoms or functional challenges, day-to-day functioning, behaviour and emotional health and safety are considered important goals and priorities. 33,34 An analysis of health-related values of multimorbid cancer survivors revealed the five values: self-sufficiency, life enjoyment, connectedness and legacy, balancing quality and length of life, and engagement of care. 18 Incorporating patient values into health-care decisions is critical, especially for elderly patients since goals may change when life expectancy shortens. 35 However, there appears to be a lack of consistency in the use of the concepts of values, goals and preferences. 18 preferences in a structured and consistent manner is lacking. 36 As a synthesis of the three themes identified from the data, It must be noted that fundamental goals and CGS show similarities with advance care planning (ACP). ACP is a formal decision-making process that aims to support patients in making decisions about future care in anticipation of the incapacity to make decisions due to a worsening condition. 41 Patients consider the focus in health care on patient goals and values to be particularly helpful. 42 ACP is usually part of an end-of-life care strategy and is used in the context of progressive illness and anticipated deterioration. 43 In our view, discussing and explicating a patient's fundamental goals and specifying values and underlying beliefs and preferences, could also be valuable in earlier stages of life, especially in patients with multimorbidity.

| Strengths and limitations
The methodological strengths of this study include the following: First, we worked with an interviewer who is trained as a GP, which may have encouraged the participants to speak frankly and directly from their own professional perspectives. The second coder has substantial experience in interview analysis but has no medical background, which helped us avoid a 'medical' bias in our data in-

| Implications for practice and research
Further research on the patients' perspectives on goals is required.
Further combined theoretical and practice-based research on this topic of goal-orientation in the context of goal-setting and decision making could prepare a shift in clinical practice towards goal-oriented care for patients with multimorbidity.

| CONCLUSION
This qualitative study provides new insights into types of goals and the consideration of goals in care for elderly patients with multimorbidity.
Based on the perspectives of clinicians, we expanded the concept of goal-oriented care by identifying a three-level goal hierarchy acting as a guide to clinical care of patients with multiple long-term conditions.
Awareness of and application of explicit fundamental goals in addition to functional and symptom-specific and/or disease-specific goals could contribute in making daily care more patient goal-oriented.
Future research is needed to refine and validate the developed threegoal model and to provide recommendations for medical training and practice. 13. Non-participation The response rates of clinical geriatricians and general practitioners were 86% and 54% respectively. Of the 21 CGs approached, one CG refused and two CGs did not respond to the first and reminder emails. A total of 28 GPs were approached. There were 6 non-responders. 3 GPs responded positively, but did not respond to proposed dates. There were two drop-outs (the interview was cancelled and there was no rescheduling (i.e. no response to proposed dates)). Two GPs of the same practice chose one participant. Their lack of time was the main reason not to participate.

14.
Setting of data collection Five interviews were face-to-face, the others were held by telephone, as the medical practitioners' busy schedules and varying locations required flexibility.
The face-to-face interviews were held at the interviewee's office.