3.1 Description of included studies
As shown in Fig. 3, we screened 8616 records and 1293 full-text papers. Of those, 47 studies were included in the synthesis: 41 qualitative, two quantitative and four mixed-methods studies (Table 1). Together, these studies represent a large sample of 2997 participants (2107 patients, 698 practitioners and 192 caregivers) from 13 countries. Among the studies, we identified 69 decisional needs cases (see additional file 3 tables for more details about cases).
Table 1. Table of characteristics of included studies (n=47)
Sources
|
# of participants (total)
|
# of patients
|
# of caregivers
|
# of practitioners
|
Country
|
(52)
|
302
|
302
|
0
|
0
|
United States
|
(53)
|
57
|
40
|
0
|
17
|
United States
|
(54)
|
142
|
0
|
0
|
142
|
Norway
|
(55)
|
41
|
41
|
0
|
0
|
United Kingdom
|
(56)
|
32
|
19
|
0
|
13
|
United Kingdom
|
(57)
|
33
|
25
|
8
|
0
|
Canada
|
(58)
|
10
|
5
|
0
|
5
|
Switzerland
|
(59)
|
14
|
14
|
0
|
0
|
United Kingdom
|
(60)
|
180
|
0
|
0
|
180
|
Sweden
|
(61)
|
36
|
21
|
0
|
15
|
Germany
|
(62)
|
53
|
53
|
0
|
0
|
United States
|
(63)
|
40
|
20
|
0
|
20
|
United Kingdom
|
(64)
|
41
|
28
|
13
|
0
|
Lebanon
|
(65)
|
59
|
27
|
27
|
5
|
Canada
|
(66)
|
35
|
35
|
0
|
0
|
Canada
|
(67)
|
86
|
43
|
0
|
43
|
Germany
|
(68)
|
8
|
0
|
0
|
8
|
Australia
|
(69)
|
25
|
0
|
0
|
25
|
Netherlands
|
(70)
|
50
|
50
|
0
|
0
|
United States
|
(71)
|
30
|
30
|
0
|
0
|
United States
|
(72)
|
23
|
0
|
17
|
6
|
Canada
|
(73)
|
807
|
807
|
0
|
0
|
Belgium
|
(74)
|
65
|
16
|
37
|
12
|
Canada
|
(75)
|
18
|
18
|
0
|
0
|
United Kingdom
|
(76)
|
21
|
21
|
0
|
0
|
United States
|
(77)
|
7
|
0
|
7
|
0
|
United Kingdom
|
(78)
|
26
|
0
|
0
|
26
|
Australia
|
(79)
|
81
|
29
|
24
|
28
|
Canada
|
(80)
|
62
|
62
|
0
|
0
|
Canada
|
(81)
|
21
|
21
|
0
|
0
|
Ireland
|
(82)
|
20
|
20
|
0
|
0
|
United States
|
(83)
|
25
|
25
|
0
|
0
|
United States
|
(84)
|
40
|
20
|
0
|
20
|
United Kingdom
|
(85)
|
83
|
27
|
28
|
28
|
Canada
|
(86)
|
36
|
36
|
0
|
0
|
United States
|
(87)
|
55
|
35
|
0
|
20
|
United Kingdom
|
(88)
|
20
|
0
|
20
|
0
|
Australia
|
(89)
|
37
|
8
|
0
|
29
|
United Kingdom
|
(90)
|
50
|
50
|
0
|
0
|
United Kingdom
|
(91)
|
16
|
0
|
0
|
16
|
New Zealand
|
(92)
|
73
|
60
|
0
|
13
|
Netherlands
|
(93)
|
16
|
4
|
0
|
12
|
Sweden
|
(94)
|
26
|
26
|
0
|
0
|
Australia
|
(95)
|
28
|
28
|
0
|
0
|
United Kingdom
|
(96)
|
30
|
30
|
0
|
0
|
United Kingdom
|
(97)
|
15
|
0
|
0
|
15
|
United States
|
(98)
|
22
|
11
|
11
|
0
|
Netherlands
|
Overall, the quality of the 47 included studies was deemed good, according to the 2018 version of the MMAT (see additional file 4). Among the 41 qualitative studies, 14 addressed all five MMAT qualitative criteria, 23 did not specify the qualitative design (one criterion unmet), and four did not meet one or more criteria regarding data collection, analysis, interpretation, or coherence. One non-randomized study did not meet the measurement criterion and one descriptive quantitative study (descriptive statistics on baseline data of a trial) did not meet the criterion on sampling and sample representativeness. Of the four mixed methods studies, two met the five MMAT mixed-methods criteria, one presented an adequate rationale for using a mixed methods design to address the research question, and one did not adequately integrate the various components of the study to answer the research question or address divergences and inconsistencies.
3.2 Decisional needs typology
The decisional needs cases showed five types of decisional needs: “prioritization” (n=26 cases), “use of services” (n=22 cases), “prescription” (n=12 cases), “behaviour change” (n=4 cases), and “institutionalization” (n=5 cases). (Table 2)
Table 2. Five types of decisional needs
DECISIONAL NEEDS TYPES / CONDITIONS AND OUTCOMES
|
PRIORITIZATION
(26 cases[i]; 1210 participants)
|
USE OF SERVICES
(22 cases; 769 participants)
|
PRESCRIPTION
(12 cases; 460 participants)
|
BEHAVIOUR CHANGE
(4 cases; 267 participants)
|
INSTITUTIONALIZATION
(5 cases; 982 participants)
|
WHAT
Decision
|
Which health and psychosocial issues should I prioritise? (n=14)
Which health and psychosocial issues am I suffering from? (n= 5)
Which discussion topic should I present in clinical encounters? (n= 3)
Should I prioritize issues with daily living activities or with health? (n=3)
Which medication should I prioritise? (n=1)
|
Should I consult or not? (n=8)
Should I use home care / social care services or not? (n=5)
Should I engage or not in a program? (n=3)
Should I use an intervention or not? (n=3)
When to go to hospital or emergency? (n=3)
|
Should I prescribe or not ? (n=5)
How to deprescribe? (n=3)
Adhere or not? (n=4)
|
Stop or maintain unhealthy lifestyle behaviour (smoking, non-healthy eating & sedentariness)? (n=3)
Continue to drive a car? (n=1)
|
Stay home or institutionalization? (n=5)
|
WHO
PCCN bio-psycho-social characteristics
|
Social vulnerability (n=11)
Mental health issues (n=12)
Multimorbidity (n=23)
Polypharmacy (n=9)
Frequent user (n=1)
Under user (n=2)
|
Social vulnerability (n=16)
Mental health issues (n=6)
Multimorbidity (n=17)
Polypharmacy (n=9)
Frequent user (n=7)
Under user (n=4)
|
Social vulnerability (n=9)
Mental health issues (n=8)
Multimorbidity (n=12)
Polypharmacy (n=11)
Frequent user (n=1)
Under user (n=1)
|
Social vulnerability (n=4)
Mental health issues (n=3)
Multimorbidity (n=1)
Polypharmacy (n=1)
Frequent users (n=1)
Under users (n=1)
|
Social vulnerability (n=5)
Mental health issues (n=5)
Multimorbidity (n=2)
Polypharmacy (n=1)
|
HOW
Made in partnership or independently (decision driven by)
|
Partnership (n=9)
Independently (n=17)
(8 by practitioner; 9 by patient)
|
Partnership (n=4)
Independently (n=18)
(3 by practitioner; 13 by patient; 2 by caregiver)
|
All independently = (n=12)
(8 by practitioner; 4 by patient)
|
All independently = (n=4)
(3 by patient; 1 by caregiver)
|
All independently (n=5)
(4 by caregiver; 1 by patient)
|
WHERE / WHEN the decision is made
|
During (n=19)
Between (n=7)
|
During (n=7)
Between (n=15)
|
During (n=8)
Between (n=4)
|
All between (n=4)
|
All between (n=5)
|
OUTCOMES quality, actions, impactsi
|
Positive (n=8)
Negative (n=16)
Equivocal[ii] (n=2)
|
Positive (n=4)
Negative (n=18)
|
Negative (n=11)
Equivocal (n=1)
|
All Negative (n=4)
|
All Negative (n=5)
|
3.2.1 Prioritization
The most common decisional needs concerned the difficulty prioritizing mental, physical, social, or other issues (26 of 69 cases). We identified 26 “prioritization” cases derived from 20 studies, which represented 1210 participants (740 patients, 55 caregivers and 415 practitioners). Cases concerned patients with overlapping bio-psycho-social characteristics: multimorbidity (n=23/26), polypharmacy (n=9/26), mental health issues (n=12/26) and social vulnerability (n=11/26) (Table 2).
Prioritization occurred between or during clinical encounters and concerned three levels: the sharing of information; identifying care goals; and considering activities of daily living. In the context of multimorbidity, the patient often had a lot of information to share during clinical meetings. They had to decide which information to present, or sometimes intentionally omit, during a time-limited clinical encounter. Then, the practitioner would choose from among the patient’s multiple concerns which topics to focus on. After an information exchange, the patient, and the practitioner, independently or in partnership, determined the priorities and goals to address. PCCNs who were socially vulnerable might feel they needed to prioritize activities of daily living over health care issues.
In nine cases, practitioner and patient prioritized in partnership during the clinical encounter (see additional file 3.1). The prioritization was explicit, with discussion and decision-making on a few selected issues. Patients and caregivers tended to prioritize issues related to functional health, quality of life and autonomy, e.g., appropriate support for maintaining activities of daily living. In contrast, practitioners tended to prioritize issues related to the disease itself, e.g., the most severe health issues. In all nine cases the parties used a prioritization tool or strategy. First, some patients used tools (e.g., Collaborative goal-setting aid) to prepare for the clinical encounter or to use with the practitioner during the encounter. Second, they used another strategy if there was no tool available, e.g., choosing a few priorities beforehand to address in the clinical encounter, and scheduling several encounters to address all their priorities. Another strategy consisted of finding a balance between the practitioner’s disease-related priorities and the patient’s illness-related priorities during the clinical encounter. Eight of these nine cases reported positive outcomes and one presented an equivocal outcome (see section 3.3.1 Well-managed for more details).
In 10 cases, practitioners or patients prioritized independently during clinical encounters (see additional file 3.2). Of those, eight cases reported a lack of patient-practitioner communication. Typically, the practitioner led the discussion during the clinical encounter and selected the topics to be addressed (paternalistic prioritization). The patients realized that the practitioner was omitting certain issues but decided to remain silent. Some practitioners complained about the overloaded agenda of PCCNs. For their part, patients sometimes selected the topics they wanted to discuss and deliberately avoided others that were clinically important, e.g., masking depression symptoms to avoid being given antidepressants (prioritization by omission). Others patients thought their point of view was trivial and feared that the practitioners would laugh at them. One case reported an under-diagnosis resulting from lack of patient-practitioner communication. Three cases reported a lack of inter-professional coordination around prioritization decisions among numerous prescribers, leading to polypharmacy. We detected dissonance between patients’ and practitioners’ perceptions of priorities. Patients and caregivers in these cases tended to prioritize information needs, self-care, loss of autonomy, and functional health, while the practitioners tended to prioritize the most severe health issues, problems with realistic solutions, specific disease issues, and a few essential tests and procedures. No case presented a patient-practitioner or interprofessional prioritization tool. One strategy used by a patient was reported. Some patients tended to prioritize conditions that could not be controlled through self-management. All cases reported negative outcomes for the three evaluated outcome dimensions: action taken, impacts on health and well-being, and decision quality (see section 3.3.2 Asymmetric encounters for more details).
In seven cases, patients prioritized independently between clinical encounters (see additional file 3.3). Six of these cases reported negative outcomes and one presented an equivocal outcome (see sections 3.3.3 Self-management by default and 3.3.4 Chaotic) and none presented a prioritization tool. One strategy used by the patient was reported. Some patients weighed the current and daily impact of some of their long-term conditions against those which they perceived could produce more serious and negative outcomes in the future.
3.2.2 Use of services
Another common type of decisional needs concerned use of services (22 of 69 cases). The 22 cases, derived from 19 studies, represented 765 participants (463 patients, 52 caregivers and 250 practitioners). Patients with this type of decisional need were in situations of social vulnerability (n=16) and faced problems with patient-practitioner communication (n=15) and interprofessional coordination (n=8) (Table 2). In seven cases, practitioners perceived these patients as frequent users of services, although their patients did not share this perception. Patients reported that they used services when their self-management burden threatened to overwhelm them. Four cases concerned patients in situations of social vulnerability who underused health services due to limited access or being refused services, e.g., marginal people.
Four cases reported use of services decisions made in partnership with the practitioner during the clinical encounter (see additional file 3.4). None of these four cases reported patient-practitioner communication or inter-professional coordination issues, and all reported positive outcomes (see section 3.3.1 Well-managed). In all four cases, stakeholders used a tool, a strategy, or a program to facilitate decision-making. One case included a tool for helping caregivers to select appropriate home care services: patients and caregivers had to decide (accept/refuse) among multiple service options. Two cases presented patient-centred strategies. One case seems to have followed an informed decision-making model: this patient decided to refuse treatment that was legitimized by the care team. Due to the patient’s nursing background, they had confidence in the patient's knowledge and ability to decide. Finally, one case presented a management program that established individualized care plans for patients identified by their physicians as frequent users.
Three “use of services” cases presented decisions made independently during clinical encounters (see additional file 3.5). Decisions were made independently by the physicians in two cases, and by patients or caregivers in one. All cases displayed a lack of patient/caregiver-practitioner communication, often associated with negative perceptions of the other stakeholders, perhaps due to an unpleasant previous interaction. Some patients and caregivers, for example, withheld health information, refused to comply with regulations, or declined assistance with activities of daily living for instance. Some practitioners felt dissatisfied about not being able to help. On the other hand, some practitioners felt uneasy with patients whom they saw as ‘difficult’ because they were non-compliant, frequent users of services, in situations of social vulnerability, time-consuming, or seriously ill and in need of high-intensity care. Some practitioners saw self-management as positive as it reduces use of services, while many patients considered seeing their doctor and nurse as a last resort, and their motivation to self-manage did not reflect a desire to reduce their use of healthcare services. Other difficult decisional needs concerned high-risk interventions, e.g., aneurysm surgery. Four of these resulted in negative outcomes and one in an equivocal outcome (see section 3.3.2 Asymmetric encounters). None of these cases reported a decision-making tool or strategy.
In 15 other “use of services” cases, decisions were made independently, mostly by patients, between clinical encounters (see additional file 3.6). The patients decided independently in 14 cases and the practitioners in one case (exclusion of a patient from a program). Patients were mostly in situations of social vulnerability (10 of 14 cases). In six cases, the practitioners perceived patients as frequent users of services or under users of services (compared to their needs). Twelve cases explained the difficulty of the decision by lack of patient-practitioner communication (n=10) and/or a lack of inter-professional coordination (n=7). These 15 “use of services” cases concerned consultations with practitioners (n=6), social and home care services (n=4), emergency room or hospitalization (n=2), engagement in a program (n=2) and interventions (n=1). Regarding consultations, some patients preferred self-management and refused services, while others were lost in the complexity of services, felt abandoned. With respect to social and home care services, patients’ and caregivers’ decisional needs fluctuated over time depending on the evolution of patients’ multimorbidity and financial resources, or on their social support. Typically, elderly patients felt overwhelmed, overestimated their autonomy, and refused or delayed social and home care services. Regarding hospital services, patients reported the emergency room and hospitalization as last resort options when the burden of self-management exceeded their capacities. The 15 cases reported negative outcomes (see sections 3.3.3 Self-management by default and 3.3.4 Chaotic). None of these 15 cases reported use of a decision-making tool. Two cases reported strategies (see additional file 3.6) that could be developed further such as using online communication technologies (e.g., secure messaging and video-conferencing) to aid self-management and receiving specialized support (e.g., a social worker) to navigate the health and social services system.
3.2.3 Prescription
Twelve cases out of 69 reported decisional needs related to medication prescription (hereafter, prescription), which represented 460 participants including 176 patients, 35 caregivers and 249 practitioners. Five cases concerned the practitioners’ decision-making about prescribing medication or not, three cases concerned de-prescribing (polypharmacy reduction), and four cases concerned the patients’ adherence to medication treatment. Patients faced multimorbidity, had polypharmacy and were in situations of social vulnerability (or all three) in 12, 11 and 10 cases, respectively (see Table 2). In all “prescription” cases, the patients or practitioners decided independently. Eight cases presented a lack of inter-professional coordination and six cases a lack of patient-practitioner communication.
In eight cases, “prescription” decisional needs concerned practitioners’ decisions made independently during clinical encounters (see additional file 3.7). The eight cases concerned prescribing vs no-prescription (n=5) or deprescribing (n=3). Seven cases reported negative outcomes, and one case an equivocal outcome (see section 3.3.2 Asymmetric encounters). None of these cases reported a decision-making tool, while two mentioned strategies to improve decision-making processes (see additional file 3.7). In one strategy, practitioners presented the positive and negative effects of the medication to the patient. In another strategy, patients enhanced the quality and amount of information provided by practitioners by bringing someone to the clinical encounter, preparing the encounter with a list of questions written in advance, reporting things mentioned previously by other practitioners, and searching information by themselves in diverse sources, e.g., the Internet.
In four cases, “prescription” decisional needs concerned patients’ decisions made independently, between clinical encounters (see additional file 3.8). Patients chose on their own not to adhere to prescriptions (medication not bought, or not taken). In all cases, patients were in situations of social vulnerability and faced mental health issues. None of these cases reported a decision-making tool or strategy and all of them resulted in negative outcomes (see section 3.3.4 Chaotic).
3.2.4 Behavior change
Four cases reported decisional needs pertaining to behavior change (see additional file 3.9), and corresponded to 267 participants including 80 patients, 40 caregivers and 147 practitioners. Two cases reported alcohol use, sedentary lifestyles and smoking by patients with chronic obstructive pulmonary diseases. In a third case, patients with severe mental disorders did not alter their unhealthy lifestyles. In the fourth case, elderly patients with dementia continued to drive. All cases involved patients in situations of social vulnerability, combined with mental health issues in three cases. All cases reported a lack of patient-practitioner communication and two cases a lack of inter-professional coordination. Patients or caregivers decided independently between clinical encounters. Patients and caregivers did not adhere to prescribed behavioral strategies. All cases reported negative outcomes (see section 3.3.4 Chaotic) and none presented a decision support tool or strategy.
3.2.5 Institutionalization
Five cases reported decisional needs concerning transfer to long term care (hereafter institutionalization), which represented 982 participants including 850 patients, 92 caregivers and 40 practitioners (see additional file 3.10). These cases showed that decisional needs about this issue are emotional and sometimes heart breaking. Typically, they involved caregivers making decisions for elderly patients in situations of social vulnerability with cognitive/mental disorders such as dementia who are unable to participate in decision-making. Also specific to this type of decisional need was that all cases presented family conflicts, aggravated in two cases by migration-related and generational cultural conflict (see Table 2). Decision-making confronted cultural clashes such as (a) communitarian values (favoring home care managed by the family) versus individualistic values (favoring transfer to long term care), and/or (b) traditional values regarding gender-based roles in decision-making roles versus feminist values. Multiple reasons triggered “institutionalization” decisional needs (unmanageable or unsafe patient behavior, uncontrolled chronic conditions, patients’ dependency on full-time care services, and caregiver burden). All cases reported negative outcomes (see section 3.3.4 Chaotic), e.g., caregiver burden, sometimes worsened by family conflict. No case reported using a decision support tool. Two cases presented decision-making strategies (see additional file 3.10). In the first case, relatives and practitioners were convinced that institutionalization was needed, waited for an incident to make it happen, or disguised the institutionalization as short-term hospitalization. In the second case, patients, caregivers, and practitioners had to negotiate between conflicting cultural values (North American versus Asian) and created alliances at turning points in the patients’ lives (alliances between a wide variety of family members and health care and social care practitioners).
3.3 Decision-making configurations: patterns of conditions in association with decision outcomes
The visualization and interpretation of the 69 cases, documented by 6W-O memos using spider web diagrams (Fig. 1), revealed nine patterns of cases with similar outcomes and conditions (Fig. 2). As shown Fig. 2, we grouped patterns into four decision-making configurations: “well-managed” (n=13), “asymmetric encounters” (n = 21), “self-management by default” (n = 8), and “chaotic” (n = 27).
3.3.1 Well-managed
The “well-managed” configuration represented the best-case scenario. In this decision-making configuration, 13 cases sharing conditions and outcomes (6W-O) grouped into three patterns (the green spider web diagrams in Fig. 2). Cases concerned decisional needs regarding “prioritization” (n=9) and “use of services” (n=4). Of the 13 cases, one had an equivocal outcome, and 12 reported positive outcomes. Positive outcomes were linked to actions taken (e.g., patient engagement in care plan; services use or access), to emotional impact (e.g., patient and caregiver experiencing positive emotion) and to the quality of the decision (informed and value-based). In this configuration, cases reported satisfactory patient-practitioner communication and interprofessional coordination between primary care services and other mental health and social care services, and decisions made in partnership by PCCN/caregivers and practitioners during clinical encounters. Fig. 2 shows, for this configuration, two patterns of conditions associated with positive outcomes. Of 12 cases, five corresponded to the “full circle” pattern and concerned patients with no major psychosocial issues; and seven cases corresponded to the “almost full circle” pattern concerning patients in situations of social vulnerability (e.g., socio-economic deprivation) or faced mental health issues (e.g., depression). Although some physicians reported that their patients’ psychosocial issues were challenging, their conditions were nevertheless well-managed. Decision outcomes were equivocal in one case because sharing a decision about health priorities seemed unusual for both the doctors and elderly patients. All cases reported a tool (n=5), a strategy (n=7) or a program (n=1).
3.3.2 Asymmetric encounters
In the “asymmetric encounters” configuration, 21 cases sharing similar conditions grouped into three patterns (the three blue diagrams in Fig. 2). This decision-making configuration exposed informational and power asymmetries during clinical encounters. Practitioners decided independently in 17 cases (e.g., prescribed or deprescribed a drug), and patients made independent decisions in the four other cases (e.g., decisions about what information to prioritize for presenting to the practitioner during clinical encounters). In “asymmetric encounters”, decisional needs concerned “prioritization” (n= 10), “use of services” (n = 3) and “prescription” (n = 8) decisions. Of 21 cases, 16 presented a lack of patient-practitioner communication and seven interprofessional coordination issues.
In this decision-making configuration, all 21 cases presented non-informed or non-value based decisions, 18 reported action-related issues (e.g., non-adherence, delaying the decision), and 18 reported negative well-being impacts (e.g., patients felt frustrated or pressured). Of 21 cases, 20 reported negative outcomes and one presented an equivocal outcome, which related to antibiotics and steroids prescriptions for patients with Chronic Obstructive Pulmonary Disease (COPD). In the “asymmetric encounters” configuration, no case reported using a tool. There is nevertheless hope for improving communication because the stakeholders made decisions during clinical encounters, i.e., patients and practitioners maintained contact for decision-making, and might consider an appropriate PCCN-oriented SDM tool.
3.3.3 Self-management by default
In the “self-management by default” configuration, eight cases grouped into two patterns with similar conditions (orange diagrams in Fig. 2). In this decision-making configuration, decisions were made by PCCN/caregivers independently and decided between clinical encounters. “Self-management by default” cases concerned “prioritization” (n=3) and “use of services” (n=5) decisional needs. Of these eight cases, seven reported negative outcomes, and one an equivocal outcome. The negative outcomes corresponded to patients who felt overwhelmed and abandoned in the self-management of their multimorbidity and experienced polypharmacy. Some patients explained that their day-to-day routine required a great deal of energy and time due to their multiple health and social issues. The eight cases reported non-informed based decision related communicational issues. One case presented underuse of services and delay in a decision leading to hospitalization (action issues). Two cases reported caregivers’ cognitive burden when the patient decided to refuse services. Two other cases presented patients/caregivers in situations of social vulnerability that affected self-management. In the case with an equivocal outcome, the patients were guided by their values and actively engaged in their care but decided not to adhere to recommended prioritization or services and had not been adequately informed of the consequence of their action. No case reported a decision-making tool. In comparison to the next decision-making configuration (chaotic), an encouraging element of the “self-management by default” is that PCCNs remained reachable and had the cognitive capacities to participate in decision-making processes and might consider a PCCN-oriented decision-making tool.
3.3.4 Chaotic
In the “chaotic” configuration, 27 cases with similar conditions grouped together into one pattern (the red square in Fig. 2). In this decision-making configuration, all cases had negative outcomes associated with mental health and social issues, unsatisfactory patient-practitioner communication and interprofessional coordination (e.g., between primary care services and other services, specifically mental health, and social care services). All the decisions were made independently, between clinical encounters by PCCN/caregivers and practitioners. The “chaotic” decision-making configuration concerned all types of decisional needs: “use of services” (n= 10), “institutionalization” (n=5), “behavioural change” (n=4), “prescription” (n=4), and “prioritization” (n=4). Almost all cases concerned patients in situations of social vulnerability (n=26), often with high disability, e.g., frailty with dementia (n=18). Some patients were homeless, or experienced financial precariousness (n=8). In addition, many cases presented patients facing mental health issues (n=19). Twenty-five cases reported action-related issues (e.g., patient non-engagement, delay of decision, access to services, and non-adherence). Emotional impacts of decisions were negative in 21/27 cases: frustration, loneliness, regret, and uncertainty about the decision made, and feeling pressured. Fifteen cases reported disagreements among stakeholders (caregivers, patients, and practitioners). Some patients (e.g., marginalized people), expressed a loss of trust in the health system. In this configuration, those making decisions involving behavioural change or institutionalization reported major family or cultural conflicts (n=9). This configuration also concerned cases reporting a refusal of social services (n=4), consultation and navigation issues in a fragmented health system (n=3), non-adherence to prescriptions by patients in situations of social vulnerability (n=3), marginalized patients prioritizing urgent problems with their daily living activities over health care (n=3), and frequent hospitalization of socially isolated frail elders (n=2). No case reported a PCCN-oriented tool. In this configuration SDM was thus unlikely to occur.
[i] See additional file 3 tables for more details about cases.
[ii] Open to more than one interpretation. Could be positive or negative outcome.