Covering patient’s perspective in case‐based critical review articles to improve shared decision making in complex cases

Abstract Background The patient has always been at the centre of the evidence‐based medicine model. Case‐based critical reviews, such as best‐evidence topics, however, are incomplete reflections of the evidence‐based medicine philosophy, because they fail to consider the patient's perspective. We propose a new framework, called the ‘Shared Decision Evidence Summary’ (ShaDES), where the patient's perspective on available treatment options is explicitly included. Methods Our framework is grounded in the critical appraisal of a clinical scenario, and the development of a clinical question, including patient characteristics, compared options and outcomes to be improved. Answers to the clinical question are informed by the literature, the evaluation of its quality and its potential usefulness to the clinical scenario. Finally, the evidence synthesis is presented to the patient to facilitate the formulation of an evidence‐informed decision about the treatment options. Key results Using three similar but contrasted clinical scenarios of patients with low back pain, we illustrate how considering the patient's preferences on the proposed treatment options impact the bottom line, a synthetic formulation of the answer to the focused question. ShaDES includes clinical and psychosocial components, transformed in a searchable question, with a full search strategy. Conclusions ShaDES is a practical framework that may facilitate clinical decisions adapted to psychological, social and other relevant non‐clinical characteristics of patients.

treatment, diagnostic or prognosis issue, and answering it by searching for, reviewing and critically appraising the relevant literature. The answer to the clinical question is applied to the patient 5 and allows for a shared decision making. The approach has been formalized by structuring some case-based critical reviews as critically appraised topics (CATs) [6][7][8] or best-evidence topic (BET) reviews. [9][10][11] Failure to incorporate the patient's perspective into guidelines has been criticized as a major obstacle to successfully implementing EBM with individual patients. 12 These criticisms are reflected in the development of patient-centred outcomes 13,14 and shared decision making. 13,15 The inclusion of patient choice is one of the cornerstones of theories linking EBM to quality of individual care and performance of care organization. 16 We believe that, similarly, case-based critical reviews are incomplete reflections of the EBM philosophy. Although these reviews start from and end with the patient, 5 BETs and other case-based critical reviews usually fail to include shared decision making, considering the patient's perspective on expected outcomes that can result from the available treatment options identified in the literature. Notably, the current format of BETs and other case-based critical reviews appears less useful when several treatment options are available to the practitioner and the patient. In such complex instances, according to Barry,13 patients are too often left feeling in the dark about how their problem is managed and how to navigate in the array available to them. As all patient-centred approaches, these reviews should respect patient's values, preferences and expressed needs, key dimensions of quality of care. 13 In this paper, we review the structure of current BETs and similar case-based reviews and introduce an alternative approach, called 'Shared Decision Evidence Synthesis' (ShaDES). ShaDES improves the currently used approaches by integrating the relevant scientific evidence and the perspective of a patient to choose an acceptable evidence-based treatment. We illustrate the application of ShaDES using three clinical scenarios.

| S TRUC TURE OF B E TS AND OTHER C A S E-BA S ED CRITI C AL RE VIE WS
The typical CAT or BET starts with a clinical scenario, which describes a patient's complaint, medical history, clinical examination and diagnostic tests, and prognostic or treatment challenges ( Figure 1) Answers to the clinical question are then searched from the literature, by retrieving evidence-based guidelines, systematic reviews or original studies. 5,18 This process involves identifying clinically relevant keywords and combining them in search equations, describing the retrieved studies, critically appraising the studies and extracting key results. 19 Results of the process, regarding the available evidence, its strengths and weaknesses and its potential usefulness to answer the clinical question are finally translated into what is called the 'clinical bottom line'. The clinical bottom line is a synthetic formulation of the answer to the focused question initially raised. For instance, in the above illustration, 17 the bottom line could be 'For this patient, there is no evidence that the benefits of biliary stenting would outweigh the risk related to this procedure'.

| CLINI C AL SCENARI OS NEED ING D IFFERENT BOT TOM LINE S
The typical case-based critical review is adapted to situations where the bottom line proposes a straightforward answer to the clinical question. The bottom line, derived from the evidence in the literature, allows the practitioner to decide whether to prescribe a treatment, or how to interpret results of a test. We believe the current format of BETs and similar case-based critical reviews is F I G U R E 1 An algorithm integrating steps of current casebased critical reviews and the proposed Shared Decision Evidence Synthesis not well adapted to situations where evidence is weak or lacking, or when there are several options among which the practitioner, and therefore the patient, must choose. We present three clinical scenarios of patients with similar clinical history and examination, but different psychosocial (anxiety, social isolation, apprehension and worries of pain) and socioeconomic context (country, occupation and possibly health insurance). We demonstrate how different patient profiles may result in different evidence-based therapeutic options and management of the patients.

| Common background
The three patients complain of low back pain (LBP) without specific lesions. They use analgesic and were prescribed exercises by a physiotherapist. These interventions provided moderate benefits. The patients consult their general practitioner (GP) for the next available options. All patients are currently working or willing to work, but they express worries about their ability to continue.

| Clinical vignette #1
Sophie is a 40-year-old bus driver in her small French home town and is also responsible for planning work assignments for a team of bus drivers. Her new manager is difficult and not inclined to consult others for their opinions about work organization. Sophie is married to Georges, a bank employee; they have two teenage children. When Sophie comes home from work, she has to supervise the children's homework and prepare the meal. Georges contributes little to home chores, which sometimes leads to conflicts.
Sophie has chronic recurrent LBP. Her most recent episode started six weeks ago while gardening. She felt a sharp pain in her lower back. She takes analgesics and has remained at work. However, her home activities are limited and her husband is worried that she cannot complete her 'chores'. Sophie feels guilty and fears that her manager believes that she might not be able to do her job appropriately. She believes that there is something wrong with her back and that her pain is caused by pinched nerve due to her occupation. Sophie was prescribed exercises by her physiotherapist and, although she never missed her appointments, she is fearful that doing exercises will harm her back and make her condition worse.
She is distressed and depressed.

| Clinical vignette #2
Mary is a 45-year-old registered nurse employed in a large Canadian She has suffered from recurrent LBP for eight months. Mary uses muscle relaxants and she enthusiastically participated in active physical therapy. These interventions provided moderate benefits. Her pain is particularly intense at the end of the day, and she occasionally experiences sharp bouts of pain at work, which makes her anxious.
She knows that chronic LBP is a benign disorder and that fear of pain can be disabling. She has never been on sick leave because of her back, but she now fears that she may not be able to continue, which amplifies her anxiety. She is very concerned that she will not be promoted at work. This makes her even more anxious, because she has been working hard for years to obtain the promotion and has gained the support of her colleagues. She is considering stopping building the boat despite the help of her family.

| Clinical vignette #3
Gloria has been suffering from LBP for the past four years. Her pain started insidiously and has been persistent for the past two years. She is 46-year-old, married, with an 18-year-old daughter. She was employed as a secretary in a small factory in Switzerland. She also worked as a part-time cook, and launched a successful blog which publishes her best recipes. However, she has to rest most of the time because of the pain. Her condition has not improved despite various courses of NSAIDs and duloxetine and physical therapy pool-and land-based exercises.
She describes her family as supportive but she feels extremely guilty, because she cannot perform the activities she used to do when she was 'healthy'. Overwhelmed by pain, she is pessimistic and gloomy, invaded by dark thoughts and feelings. She tries to stay active in her everyday life, but she does not succeed. When asked about what gives meaning to her life, she can hardly respond. While Sophie and Mary consider that treatment may be helpful, Gloria's negative appraisal also applies to treatments.

| S HARED DECIS ION E VIDEN CE SYNTHE S IS
ShaDES includes four steps ( Figure 1): First, the clinician builds the clinical and psychological scenario. Second, the clinical information collected in the history and examination informs the literature search and critical appraisal of the retrieved evidence. Third, the clinician synthesizes the evidence, as is done when developing decision aids. [25][26][27][28][29] Finally, the clinician and patient enter in a shared decision process where the patient expresses his/her preferences regarding the available treatment options.

| Starting with a clinical and psychological scenario
Implementing ShaDES starts with understanding the clinical scenario by describing the patient, his/her medical and personal history and clinical examination, and identifying diagnostic, prognostic or treatment issues ( Figure 1). As seen in our three vignettes, the

| Elements common with BETs and other casebased critical reviews
This part is not specific to ShaDES, and readers can refer to Elwyn et al 15

| Integrating patient preferences
In the next step of ShaDES, the evidence synthesis is presented to the patient to facilitate the formulation of an evidence-informed decision about the treatment options. To successfully achieve this step, one can use, for instance, O'Connor et al.'s approach to construct decision aids. [25][26][27][28][29] Basically, each available option is presented to the patient with an explanation of: (1)  In such decision aids, the patient is invited to express his/her feeling by using sliders or cursors through visualization of the relevant dimensions (outcome possibly improved, risks, side effects…) that are important to the patient.
In ShaDES, we propose to present the evidence in a comparative

| Application
If we now consider how the elements of our proposed model can be applied to the clinical situations described above, the following issues and questions need be emphasized and addressed: Sophie: Sophie was not reassured by her GP, because of her fear of pain or recurrent injury. The clinical and psychological question for Sophie is whether her challenges with her recurrent LBP and the unsatisfactory results of evidence-based medication and exercise (P) could be better addressed with cognitive-behavioural therapy and

yoga (I & C) directed at pain and function (O). For such conditions,
there is strong evidence that CBT is more effective for pain, functional status and behavioural outcomes than placebo/no treatment/ waiting list controls. CBT requires the identification of meaningful and realistic goals to allow Sophie to address her catastrophizing with pain and activity and thus improve her readiness to engage in active pain self-management. Sophie may also consider yoga as she expresses positive expectations towards such a treatment, but limited evidence exists to support its effectiveness compared with standard treatments. 20,22 Mary: The GP could reassure Mary by addressing her concerns regarding her work. In this instance, the clinical and psychological question is whether her challenges with her recurrent LBP and the TA B L E 1 Available options to treat patients with chronic low back pain who did not benefit from an evidence-based intervention that included medication and exercise Yoga is an effective intervention for chronic LBP, but patients' fears about the treatment need to be addressed.
CBT is an effective treatment options for patients who do not want physical treatment; it can also be used in combination with manual therapy and yoga.

| D ISCUSS I ON
To our knowledge, ShaDES is the first framework that explicitly includes patients' perspectives in case-based critical reviews. Our illustration represents real life, where practitioner and patients are usually confronted to more complex situations than are usually depicted in guidelines or clinical scenarios. 12 Thus, ShaDES is closer to the full model of EBM than other short reviews, such as BETs.
ShaDES may also more ethical, because one cannot make informed choices without information, even if scientific information is lacking, scarce, poor or equivocal in the literature. 4,[32][33][34][35] We focused on including the patient's perspective. Nevertheless, we still have to consider the practitioner's experience, or availability of resources, 1,3 or even, as suggested by Siminoff et al., the patient's family preferences. 36 The inclusion of physician's experience has become possible with electronic medical records 37,38 and automatic search of a practitioner's databases or a hospital data warehouses. 39 Issues of availability and access to resources, such as delays to access professionals in the specific context where patients are seen  44 and specific recommendations regarding patient-centred outcomes. 14 Further research needs to focus on the integration of key characteristics of the clinician's psychology, such as empathy, 45 affective attitude 46 or ability to provide cognitive reassurance, 47 known to lower anxiety and distress and to be associated with better clinical outcomes. 45 We also need methods to improve the development of effective decision aids or other tools for shared decision making 15,43 and to integrate them in daily clinical practice, including in busy practices where available time might be an issue. In such instances, we believe our approach is pragmatic, as it is based on questions to assess patient preferences that are close to the usual patient-practitioner encounter; these questions are also a good reflection of the 'Choice Talk, Option Talk, Decision Talk' model of Elwyn et al. 15 We need research on the effect of implementing ShaDES in clinical practice, including comparisons with shared decision approaches, such as qualitative studies investigating patient adherence and satisfaction, and quantitative research on effectiveness. 16 We also need to better train caregivers to focus critical reading on clinically relevant issues such as proximity of population covered, relevance of outcome measurement, and magnitude and precision of effect sizes.
Above all, if we are to discuss the need to incorporate the patients' perspectives in all decision making, training must address the development of caregivers to explain evidence to patients.
Basic training of health-care professionals also must convey the importance of having the same rigour when documenting indication of a procedure and baseline and outcome measure in records and information systems.

| CON CLUS ION
We have proposed a new way to present case-based critical reviews, which might better reflect the original philosophy of EBM.
Our proposition puts a strong emphasis not only on clinical but also on psychosocial components, aimed to point to a question that can be developed into a search strategy. This proposition thus offers a practical framework that may contribute to clinical decisions using relevant psychological, social and other characteristics of the patient and of his/her environment. The applicability and impact of such an approach, which specifically considers values, preferences and needs of patients, remain to be tested.

CO N FLI C T O F I NTE R E S T
The authors declare they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analysed in this study.