Interventions to improve participation in health‐care decisions in non‐Western countries: A systematic review and narrative synthesis

Abstract Background Patients' participation in medical decision making is an important aspect of patient‐centred care. However, there is often uncertainty about its applicability and feasibility in non‐Western countries. Objective To provide an overview and assessment of interventions that aimed to improve patients' participation in decision making in non‐Western countries. Method Ovid Medline, Embase, Cochrane Central Register of Controlled Trials, CINAHL, Ovid MEDLINE(R) Epub Ahead of Print, In‐Process, Other Non‐Indexed Citations, without Revisions and Daily Update and Database of Abstracts of Reviews of Effects, were searched from respective inception to February 2018. Studies were included if they (a) were randomized controlled trials, before‐and‐after studies and interrupted time series studies; (b) were conducted in non‐Western countries; (c) aimed to improve patients' participation in dyadic decision making; and (d) reported outcomes relevant to patient participation in decision making. Studies were excluded if they included children, were about triadic decision making or solely focused on information provision without reporting outcomes related to patient participation. Narrative synthesis method was used for data analysis and presentation. Results A total of 17 studies, 6 RCTs and 11 non‐RCTs, were included across ten countries. Intervention strategies included patient and/or provider communication skills training, decision aids and a question prompt material. Whilst most of the studies reported increased patient participation, those interventions which had provider or patient training in communication skills were found to be more effective. Conclusion Interventions to improve patient participation, within the context of dyadic decision making, in non‐Western countries can be feasible and effective if communication skills training is provided for health‐care providers and/or patients.


| INTRODUC TI ON
Participation in decision making is a process where engaged patients and health-care providers partake in shared decision making through the meaningful exchange of information and experiences. 1 It is a key characteristic of patient-centred health care, a paradigm that has become popular in recent decades, replacing more paternalistic healthcare models. Recent evidence shows that greater participation in health-care decisions increases patients' satisfaction, improves patient-provider relationships, facilitates medication adherence and decreases health-care costs. 2,3 There is also emerging evidence that participation in decisions may reduce health inequalities experienced by vulnerable groups such as racial and ethnic minorities, low literacy groups and seniors. 4 However, issues such as time constraints, patient characteristics, low health literacy and cultural factors are often reported as barriers to participative decision making, with some saying that it is impractical amongst certain groups. [5][6][7] Globally, this paradigm shift was reflected in the pronouncement of the Alma-Ata Declaration in 1979, a landmark moment calling for greater participation from individuals and communities in their health-care planning and implementation. 8,9 More recently, the World Health Organization (WHO) published the Framework on Integrated People-Centred Health Services (IPCHS), promoting Universal Health Coverage through equal, responsive, affordable and quality health-care services. 10 An important strategy proposed by this Framework is to engage and empower individuals and families. 11 A number of strategies, including shared clinical decision making, were proposed. However, the report fell short of providing recommended strategies, concept analyses and best practices.
In many Western countries, policies have been developed to support patients' participation in health-care decision making and the use of decision aids, question prompt lists and training for both clinicians and patients. 12,13,55 For example, in Australia, the statement 'I have a right to be included in decisions and choices about my care' is part of the Australian Charter of Healthcare Rights. 14 However, less is known about how to effectively involve patients in health-care decisions in non-Western countries. In these settings, some have argued that the concepts of 'patient centeredness' and 'active participation' are based on the Western ideology of individual autonomy and are therefore less applicable. 16,17 In cultures where individuals see themselves as agents of a family, community or a tribe, within a hierarchical community, health-care professionals are often to be respected. [18][19][20] Questioning by patients is to be avoided to bring harmony during encounters. [18][19][20] Other factors that may be prevalent in some non-Western countries are high patient loads, lack of skills in participatory communication amongst health providers, a lack of relevant research evidence and low health literacy amongst patients. [20][21][22] These lead to the question of whether patient-centred care, and more specifically, patients' participation in health decisions, is a feasible and appropriate strategy in non-Western country contexts. This systematic review aims to identify interventions designed to improve adult patients' participation in health-care decisions in non-Western countries, assess their feasibility and synthesize factors that influence their effectiveness.

| ME THODS
This systematic review is reported in accordance with preferred reporting items for systematic reviews and meta-analyses (PRISMA) (Appendix S1). 23

| Inclusion and exclusion criteria
The PICOS (participants, intervention, comparator, outcome and study design) 23 approach was used to define the following eligibility criteria for study selection.

| Participants
Studies were included if participants lived in non-Western countries (defined as countries that are not members of UN classification of Western European and Other States Group (WEOG)). 24 The same classification method was used in a previous systematic review 25 We excluded studies that included children (aged <18).

| Interventions
We included studies which aimed to improve the participation of

| Outcomes
Outcomes related to patient activation, patient or provider participatory behaviours during the decision-making encounters were analysed. Patient activation is a broad concept with a definition of 'an individual's knowledge, skill, and confidence for managing their health and health care'. 26 In this systematic review, we only included studies which reported patient activation outcomes in relation to individuals' skills and confidence in participating in health-care decision making.

| Study designs
Randomized controlled trials (RCTs), controlled or uncontrolled before-and-after studies with pre-and post-test data available and interrupted time series studies were included.

| Search strategy and study selection
We systematically searched databases using keywords and Medical Subject Headings (MeSH) related to pre-specified PICOS criteria.
Some segments of our search strategy were adapted from other published systematic reviews with similar concepts. 4,27 The search strategy was originally developed in Medline via OvidSP (Appendix S2) and later modified to other databases. We initially limited our search to humans, adults and the English language, and later expanded the search to the non-English language records. Returned records from database searches were combined, duplicates removed using Endnote X8 software, and remaining references imported to the Covidence tool 28 for screening, data extraction and quality assessment purposes. Two reviewers conducted title and abstract screening and full-text screening of eligible studies on Covidence. Disagreement on the selection of certain studies was resolved by consensus.

| Data extraction and quality assessment
We extracted data using the Covidence online tool 28 and an adaptation of the Cochrane Consumers and Communication Group data extraction template. 29 We recorded country of origin, study design, participant numbers, intervention characteristics, theoretical framework, setting/conditions, outcome measures and detailed outcome results.
The quality of RCTs was assessed using the Cochrane risk of bias tool (sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other sources of bias). 30 The quality of non-randomized studies was assessed using the modified Downs and Black's checklist, 31 rating each study numerically against 27-item questions, and the total score ranged from 0 to 28.

| Data analysis
Due to the wide variation of study designs, intervention strategies and outcome measures used in the included studies, a narrative synthesis method was used. Narrative synthesis is a process of exploring study characteristics and their relationships within (and between) included studies in order to identify factors influencing the effectiveness and implementation of interventions. 32 The process of narrative synthesis was partially guided by recommendations by Popay et al. 32 We used textual description, grouping and tabulation methods for preliminary synthesis and exploration of patterns across studies.  Table 1 and Table 2).

| Characteristics of included studies
The methodological quality of the included RCTs varied across studies ( Figure 2). Two (2/6) did not provide sufficient information on random sequence generation methods, and three (3/6) studies did not describe or have allocation concealment. Blinding of participants and personnel was lacking in two (2/6) studies, and in one study, outcome assessment was not reported in detail to permit a judgement. The Downs and Black quality scores for non-randomized studies ranged from 12 to 21 (see Table 2).

| Data synthesis
There were four types of intervention strategies in the included studies; provider communication skills training (n = 4), patient communication skills training (n = 3), question prompt material (n = 1) and patient decision aids (n = 9). Details of elements of each intervention strategies, theoretical background and development processes are summarized in Table 1. Based on our pre-defined outcome inclusion criteria and emergent patterns in the extracted data, we categorized study outcomes into three groups: (a) change in decisional conflict or preparedness; (b) patient participatory behaviours; and (c) provider participatory behaviours (see Table 2).

| Provider communication skills training
Four studies used the conceptual framework of Interpersonal Communication and Counselling (IPC/C) [33][34][35] or client-centred counselling 36 for provider communication skills training, and all four were controlled before-and-after studies. Two studies 33,34 measured the effects of stand-alone provider interpersonal communication skills training, whilst the other two 35 such as asking questions 36 and providing medical information, 33,34 were also reported.

| Patient communication skills training
Three studies provided communication training for patients. One RCT from Namibia used a curriculum with three components: learning to speak to providers, using tools to help communication and overcoming barriers to communication. 37 The second cluster-randomized study from Indonesia provided individual coaching to patients on asking questions, requesting clarification and expressing concerns prior to their consultations. 38 Finally, a controlled beforeand-after study from South Korea developed a shared decision-making training programme for people with schizophrenia. 39 The Namibian training programme for patients resulted in the doctors of trained participants performing significantly better in facilitating (adjusted difference in score 1.19, P = 0.004) and gathering information (adjusted difference in scores 2.96, P = 0.000) than control group doctors. These trained patients also asked significantly more questions during consultations (adjusted difference in score 0.48, P = 0.012). 37 The Indonesian study of individual coaching for patients also resulted in the coached patients asking significantly more questions than those in the control group (6.3 vs 4.9, P < 0.01). 38 Similarly, the shared decision-making training in South Korea found a significant positive change in self-esteem in the intervention group compared to control (mean change ± SD: 4.06 ± 4.42 vs −1.06 ± 3.66, P < 0.001) which could be seen as empowerment in decision making. 39

used the International Patient Decision Aids Standards (IPDAS)
Collaboration checklist to guide the development of their decision aids. 42,43,45,46,48 Of the nine included studies of PDAs, three were RCTs. 42,43,48 We note that none of the RCTs 42,43,48 included training as part of their intervention and that the studies which did were all of a weaker study design as uncontrolled before-and-after studies (n = 6). 40,41,[44][45][46][47] Nevertheless, amongst these nine studies evaluating patient decision aids, there was a consistent improvement in the patient and provider participatory behaviours in those studies which included provider and/or patient training as part of the intervention. 40,41,[44][45][46][47] The Hong Kong RCT found that the decision aid tool significantly The RCT from Japan found that decisional conflict was significantly reduced for both the decision aid groups one month after receiving them and after having the selected surgery (26.5 vs 26.9 vs 32.1), but not immediately after the intervention and before the surgery (28.7 vs 29.8 vs 31.7). 43 Conversely, the RCT study from South Korea 48 which assessed a PDA for carpal tunnel syndrome did not find any difference in the decisional conflict between the PDA (video format) group and the regular information group (control) (22 vs 23, P = 0.76).
Amongst the remaining six uncontrolled before-and-after studies, three studies using the WHO family planning decision-making tool (DMT) reported positive changes in provider and client participatory behaviours (see Table 2). 40,41,44 The other two studies from China (PDA in POAG) and Japan (PDA in birth choices) reported a reduction in decisional conflict after patients were given PDAs with a short briefing 45 and decisional support. 46 The study from mainland China, using a 3D multimedia system able to display relevant dental  where sufficient level of provider participatory behaviours were observed. These included significant increase in giving information about methods (from 58% to 80%), method efficacy (from 54% to 88%), how to use the chosen method (from 81% to 98%) and complication of the chosen method (from 67% to 94%). There were also significant increases in behaviours such as engaging clients to speak (from 75% to 93%) and answering all the client questions (from 89% to 99%) ↑ 17 Shum 2017   Abbreviations: CBA: controlled before-and-after studies; CI: confidence intervals; N: total number; OR: odds ratio; RCT: randomized controlled trial; RIAS: Roter interaction analysis system; RR: relative risk; SD: standard deviation; UCBA: uncontrolled before-and-after or time series studies; empty cells indicate that the outcome was not assessed; ↑ positive effect; ↔ no effect/difference.

TA B L E 2 (Continued)
anatomy as an aid for patient-provider communication, reported that patients felt more involved in decision making, understood decisions and treatment planning. 47

| Question prompt materials
One study (an RCT) in Japan compared the use of a Question Prompt Sheet (QPS) and standard Hospital Introduction Sheet (HIS) for advanced cancer patients. 49 Participants who received the QPS were more likely to find the materials useful in helping them asking questions compared to the HIS group (4.4 ± 3.6 vs 2.7 ± 2.8, P = 0.003); however, there were no differences in asking questions (63% vs 71%) nor total questions asked (both median 1) between these two groups. 49 and Mexico and achieved long-term maintenance of provider communication skills. 35,36 They suggested that incorporating specific supervision on provider IPC/C skills to the already existing functional provider performance supervision system could be a costeffective option in developing countries. 35,36 Communication skills training was evaluated amongst Indonesian patients of providers who had previously received client-centred counselling skills training. 36  has not progressed and should be urgently revisited.

| D ISCUSS I ON
Our review was not able to explore specific cultural aspects of patient participation, but it did highlight the potential for this to be successfully achieved, particularly if provider training is incorporated. Learnings from the studies that used IPC should be particularly noted, due to the unique way that they emphasized reducing social distance, an aspect of culture not strongly featured in Western countries. Social distance can be a 'virtual barrier' between providers and patients created by the subjective feelings of alienation in class and status due to age, sex, race and social, educational, economic and cultural backgrounds. 50 Therefore, such cultural and local aspects of patient-provider communication in non-Western should not be overlooked when designing interventions to promote patient active participation.
One of the limitations of this systematic review is that we focused our systematic review on participation in dyadic decision making amongst patients living in non-Western countries. We acknowledge that family and significant others can play a significant role in the process of decision making in some patients from non-Western cultural backgrounds. 57 However, one recent study 58 has suggested this may be less homogenous within cultures than previously thought. Another limitation is that we only included studies that were published in English and were identified from major databases in medical research. To draw more complete conclusions on the evidence from non-Western countries, a comprehensive search of the literature in non-Western country-specific databases that collect local language studies might be needed in the future.
In conclusion, people in non-Western countries can successfully be involved in their health-care decisions, and this should not be overlooked as this is a core component of a people-centred healthcare system as advocated by the Alma-Ata Declaration and the WHO framework for IPCHS. Our study highlights the ability of communication skills training for patients and providers to increase patient participation and involvement in health-care decisions. Such intervention strategies should be further developed and implemented as a priority in non-Western countries regardless of their income status.

CO N FLI C T O F I NTE R E S T
No conflict of interest.

DATA AVA I L A B I L I T Y
Data for this systematic review were derived from published articles which are available in the public domain and may be subject to copyright. Relevant data supporting the conclusions of this systematic review are included within the article and supporting files.