Article Text

Original research
Understanding of empathetic communication in acute hospital settings: a scoping review
  1. Jaquille Haribhai-Thompson,
  2. Karen McBride-Henry,
  3. Caz Hales,
  4. Helen Rook
  1. Faculty of Health, Victoria University of Wellington, Wellington, New Zealand
  1. Correspondence to Dr Helen Rook; Helen.rook{at}; Dr Caz Hales; caz.hales{at}


Objective Empathy and empathy education have been reviewed a number of times through systematic reviews and meta-analyses; however, the topic of ‘empathetic communication’ remains poorly understood when considering engaging in hospital-based research. Therefore, this scoping review aimed to explore the existing literature concerning empathetic communication in hospital settings and to evaluate the definitions presented.

Design Scoping review.

Data sources Systematic searches of the PubMed, CINAHL, Cochrane, PsycINFO, and PsycArticles databases were conducted.

Study selection All English studies in which empathetic communication in hospital settings were explored. The search terms used included empathy, communication, hospital settings, providers, and consumers.

Data extraction Data were assessed through the use of a pre-set analysis tool.

Results After conducting the searches, 419 articles were identified, of which 26 were included in this review. No single article specifically defined the term ‘empathetic communication’; however, 33 unique definitions of ‘empathy’ were identified, of which 23 considered communication to be a component of empathy. There was a considerable lack of consistency between the empathy definitions, with some classifying communication in empathy as an ability and others classifying it as a dynamic process.

Conclusion Future and contextually focused research is needed to develop a consistent and clear definition of empathetic communication and empathy within a hospital setting to better build positive healthcare cultures.

Practice implications Inconsistencies between definitions of empathy in empathetic communication research could reduce the efficacy of future research gains and impact the translation of research findings into clinical practice.

  • health services administration & management
  • quality in health care
  • medical education & training

Data availability statement

No data are available.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • This is the first review of current knowledge about ‘empathetic communication’ within acute hospital settings.

  • Identification of how empathetic communication was defined and measure for research purposes in acute hospital settings was performed.

  • While five databases were searched, and multiple search terms were used, the amount of literature on this topic was relatively small.

  • This review has identified the need for a consistent definition of empathetic communication for use in clinical settings.


Empathy and empathetic communication are important elements in effective patient-centred care.1–3 Patient-centred communication is foundational in building trusting relationships between healthcare providers and patients.4 There is evidence to suggest that empathetic provider–patient communication can lead to better outcomes,5 treatment adherence,6 and patient satisfaction.7 8 A fundamental element of high-quality healthcare is for clinicians to recognise and respond to individual and families’ perspectives. Developing an understanding of differing perspectives and cultures builds responsibility, adaptability and empathetic communication skills that are critical in ensuring patients receive high-quality care and for developing therapeutic provider–patient relationships.9 10 Therefore, the development of empathetic communication should be the cornerstone for all health service provider–patient interactions.

Empathy is a broad umbrella term commonly described as consisting of different dimensions—including cognitive empathy, the ability to understand another’s mental state and affective empathy, the ability to respond to another’s mental state with an appropriate emotion.11 These are independent from one another,12 and empathy as a whole is modifiable with interventions.13 A component of affective empathy often described is empathic concern—displaying compassion/sympathy in response to another’s suffering.14 However, there remains significant inconsistency among scholars about the terms ‘cognitive empathy’ and ‘affective empathy’ and how they are defined impacts the method of measurement, for example, observable interpersonal behaviour or self-reports.15 16 Empathy and empathy education have been reviewed a number of times though systematic reviews and meta-analyses16–19; however, the topic of ‘empathetic communication’ remains poorly understood when considering engaging in hospital-based research.

In 2020, a feasibility research project was conducted that aimed to establish a positive healthcare culture across several different inpatient wards/units in a hospital in Aotearoa/New Zealand (Author). The feasibility study aimed to determine whether an empathy education programme could be implemented, and whether the proposed research measurement tools employed to assess staff empathy levels and patients’ reports of staff empathy were valid. The study measures intended to capture patients’ perceptions of staff empathy during care interactions post-intervention. The feasibility study highlighted that there was an unclear understanding of what constituted ‘empathetic communication’ or how to observe or measure this meaningfully. As a result of the feasibility study, the research team knew there was a need to understand how empathetic communication was defined and measured in hospital-based clinically focused research. This feasibility study was the genesis of this scoping review presented in this paper.

No review to date has explored the current knowledge regarding definitions of ‘empathetic communication’ as a distinct concept. Identifying a formal definition will help inform future research that can assist in implementing empathetic communication in practice, enabling researchers to standardise and test interventions with the aim of enhancing patient care and the quality of the therapeutic relationship. The purpose of this scoping review is to ascertain current definitions of empathetic communication and explore what research is available that addresses empathetic communication. A scoping review method employs a highly structured approach to identify a range of literature available about a topic; this assists researchers to determine future possible steps in the research area.20 Three research questions were used to guide this review:

  1. What definitions of empathetic communication are used for research purposes in hospital settings?

  2. What tools have been used to measure empathetic communication?

  3. What types of research have been conducted about empathetic communication?


Search strategy

The scoping review methodology developed by Arksey and O’Malley20 was used. The scoping review has been conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews reporting statement.21 Literature searches were performed using the PubMed, CINAHL, Cochrane, PsycINFO and PsycArticles databases. Five search terms in combination were searched for in titles and/or abstracts published after 1 January 2011: ‘empathy’, ‘communication’, ‘healthcare’, ‘provider’ and ‘patient’. Various synonyms for each were used to ensure as much literature as possible was captured in the searches.

The search terms used for each database were the same and are presented in online supplemental information along with the different filters used for each database (online supplemental tables 1-3).

To be included in the review papers needed to be peer reviewed and identify empathetic communication as the focus of their research (see box 1). All empirical research were included to consider different aspects of conceptualising and measuring empathetic communication.

Box 1

Inclusion and exclusion criteria

Inclusion criteria

  • Discussed/assessed empathetic communication.

  • Focused on the provider–patient relationship.

  • Based in a secondary, tertiary, acute care or hospital setting.

  • Measured empathy.

  • Empathy included as part of the study design.

Exclusion criteria

  • Animal studies.

  • Editorials and opinion pieces.

  • Studies involving children.

  • Non-English language publications.

  • Published before 1 January 2011.

All identified article titles and abstracts were screened according to the inclusion/exclusion criteria (box 1) by three researchers (JH-T, HR and KM-H) using Rayyan,22 and any conflicts were resolved via discussion with a fourth researcher (CH). Full-text articles were then examined the same way.

Data extraction

Data were extracted by one researcher (JH-T) into a data-charting form created in Microsoft Excel jointly developed by all researchers on the research team. Data included were:

  • Title, first author, year of publication, country.

  • Study design.

  • Definition of empathetic communication.

  • Definition of empathy.

  • Definition of communication.

  • Empathy measurement tools, and whether empathetic communication was measured from the perspective of another person.

  • Interventions used in studies.

  • Purpose of study.

  • Study population and setting.

  • Summary of conclusions.

Where doubt arose as to whether the extracted data matched what was being included, a decision was made by two researchers (HR and CH).

Data analysis

Extracted definitions were then imported into NVivo12 (QSR International) for thematic analysis. Themes were developed by the research team through an iterative process of identifying consistent concepts, ideas and words within the definitions, and grouping these accordingly; this approach was in keeping with the chosen method that guided this review.20 Themes coded for included examining the roles of communication within empathy definitions, different components of empathy within definitions and whether the definitions defined a linear process.

Patient and public involvement

No patients were involved in the design of this study.


Included studies

A total of 419 articles were identified: 283 from PubMed, 82 from CINAHL, 42 from PsycArticles, 12 from Cochrane and 0 from PsycINFO. Fifty-seven duplicates were removed using the Rayyan detection tool.22 One further duplicate was removed manually.

The screening process led to 307 articles being excluded based on the inclusion/exclusion criteria, leaving 54 articles included. After examining the full text of the remaining articles, 28 articles were excluded based on the inclusion/exclusion criteria, leaving 26 articles included in the review (figure 1). See online supplemental table 4.

Figure 1

PRISMA diagram representing the scoping review literature search. PRISMA, Preferred Reporting Items for Systematic reviews and Meta-Analyses.

Study locations and designs

Eleven studies (42%) were conducted in Europe,23–33 six (23%) in Asia,34–39 five (19%) in North America,40–44 three (12%) in Brazil45–47 and one (4%) in Australia.48 Twenty-one studies (81%) were published in or after 2017,23–25 27 28 31–36 39–48 with the earliest studies being published in 2014.26 29 37

Sixteen studies (62%) were cross-sectional studies,23 25–29 32 35 37 43 45–47 49–51 four (15%) were qualitative studies,24 34 44 48 three (12%) were mixed-methods studies,30 31 40 and there was one (4%) prospective cohort study,42 one quasi-experimental study39 and one (4%) randomised controlled trial.36

Definitions of empathetic communication, empathy and communication

‘Empathetic’ and ‘empathic’ were assumed to be equivalent terms, where empathic was used in the recorded definition this was left unchanged in online supplemental table 4. None of the 26 examined studies included an explicit definition of ‘empathetic communication’. However, 21 studies defined empathy at least once,23–29 32–39 42–44 46–48 with a total number of 36 definitions of ‘empathy’ (see online supplemental table 4). Each of these 21 studies provided a definition of empathy that referenced communication in some way as a component of empathy. Most definitions of empathy were described as ‘empathy’, though others defined ‘physician empathy,27 29 42 ‘medical empathy’,27 44 ‘clinical empathy’,44 ‘relational empathy’,23 ‘nursing empathy’34 and ‘ therapist empathy’.34 These all referred specifically to empathy from the provider to the patient in a provider–patient relationship, and otherwise appeared to be the same as the other definitions of empathy. Two studies also defined empathic episodes32 48; one describing the processes that allow someone to recognise the emotional state of another both on an intrapersonal level and behaviourally,32 and the other as a four-component linear process leading to someone responding to the experiences of another.48 One study referred to empathy as ‘perceived warmth’.47

Two definitions of empathy appeared to be identical to a definition by Kurtz et al,39 44 52 where it is defined as a two-phase process where the first involves understanding and appreciating another’s feelings and emotions and the second communicating understanding back in a supportive way. Two definitions simply described empathy as the essence of all nurse–client communication and nurse–patient interaction, respectively,34 48 which appear to be identical. Most notably, 6 of the 36 definitions cited an article by Mercer and Reynolds,23 25 26 29 42 46 53 which concludes that empathy is an ability that involves three components—understanding a patient, communicating that understanding, and acting on that understanding with the patient in a helpful way. Three definitions appeared to be identical to this definition23 26 29; however, two of these definitions in two articles by Steinhausen et al26 29 further develop the definition of empathy by emphasising that ‘the physician’s sensitivity to patient concerns’ is also essential. Like many of the definitions, Mercer and Reynolds53 characterise empathy as an ability to communicate rather than a dynamic process of ‘empathetic communication’. Treating the definition by Steinhausen et al as distinct from, but similar to, Mercer and Reynolds’ definition, results in 33 unique definitions of empathy.

Communication is clearly recognised as an important aspect of empathy. Twenty-three of 33 (70%) unique definitions referenced communication as part of the definition,23–29 32–39 42–44 46–48 using verbs like ‘communicate’, ‘convey’and ‘share’. Thirteen (57%) of these referred to empathy being an ability or capacity to communicate in some way,23 25–29 33 35 36 43 46 47 while nine (39%) referred to communication being a dynamic process or behaviour of empathy.24 27 34 37–39 42 44 One (3%) unique definition describes empathy differently—describing it as ‘the essence of all nurse–client communication’.34 48

Three of 33 (9%) unique definitions described empathy as a linear step-by-step process.39 48 54 For example, first understanding and appreciating another’s feelings and then communicating back to them in a supportive manner.39

Eleven of 33 (33%) unique definitions explicitly mention cognitive empathy, which is also described as perspective taking,32 48 and affective/emotional empathy as separate components of their definition.24 32 34 37 44 46–48 However, many of the definitions mention features of cognitive empathy, but do not explicitly define it. Cognitive empathy is described as having an ability to understand someone else’s perspective, feelings or emotions in the identified definitions,32 34 46–48 and 15 of 22 (68%) definitions that did not explicitly mention cognitive and emotional empathy, mentioned understanding someone else’s perspective, feelings or emotions in their definition.23 25–29 32–39 43 44

Emotional and affective empathy were considered to be equivalent by Savieto et al,46 describing them as ‘the ability to put oneself in another person’s shoes’, and Wu34 considered affective empathy as partaking in the same feelings as someone else is experiencing them, and both of these appear to be similar. However, a definition mentioned by Gerace et al48 describes emotional empathy as being part of ‘emotional reactions to another person’s experiences’, and a definition mentioned by Brooke et al47 describes an affective component of empathy being the ability to share the emotions of others. Another definition mentioned by Moreno-Poyato et al32 considers affective empathy to consist of two dimensions—the ‘tendency to experience feelings oriented towards others such as compassion and concern’ or ‘empathic concern’, and the tendency to emotionally react based on another’s suffering, which is consistent with the definition noted by Gerace et al.48 These multiple definitions suggest that there are possibly inconsistencies in how emotional and affective empathy are defined.

Furthermore, two authors’ definitions mention behavioural empathy as another component, defining it as ‘effectively communicating the understanding of the situation’46 and having a ‘cognitive and affective part, and is the expression of understanding the patient’s perspective with recognition of the patient’s situation and a feeling of identification with the patient’s suffering”.24 One definition of ‘therapist empathy’ is split into other components, including ‘sharing empathy’: ‘sequences where the therapist displays that he/she has something in common with the patient’ and ‘nuturant empathy’: ‘characterised by the therapist being supportive, security providing or totally attentive’ as well as cognitive and affective components.34

No studies explicitly defined communication. However, one study described that both ‘health communication and the doctor–patient relationship’ are ‘the means by which the physician can convey the intended information to the patient’.33 While it appears that the provider–patient relationship is fundamental to communication, the authors do not elaborate on what this means.

Empathetic communication measurement tools

Despite the lack of a formal and consistent definition of ‘empathetic communication’, there were a number of studies that measured ‘empathetic communication’ in some way. Of the 26 studies, 21 (81%) measured empathetic communication from another person’s perspective.23–30 33–35 39–48 Twelve studies (57%) used the Consultation and Relational Empathy (CARE) Measure,23 25–27 29 30 35 42 43 45–47 a 10-question patient-completed questionnaire used to evaluate provider empathy in a consultation.55 Five studies (24%) recorded provider empathy qualitatively.24 34 40 44 48 Three studies (14%) used the Jefferson Scale of Patient’s Perceptions of Physician Empathy (JSPPPE),28 33 45 a five-question patient-completed questionnaire like CARE used to evaluate provider empathy.56 One study used both JSPPPE and CARE,45 and one study used both CARE and Warmometer,47 a tool used to measure the perceived warmth (considered to be equivalent to empathy by the study) of the patient–provider relationship from the patient’s perspective. One study used the Active Empathetic Listening Scale,41 an 11-item tool evaluating three subscales of someone else’s listening—sensing, processing and responding. One study’s research team designed a Global Rating Scale for assessing empathetic communication,39 which was based on four strategies of an empathy model by Pehrson et al.4: that included recognising or eliciting a patient’s empathetic opportunity, working towards a shared understanding of the patient’s emotion/experience, empathetically responding to the emotion/experience and facilitating coping and connecting to social support.

The CARE measure was created based on numerous concepts of empathy, including the definition of empathy by Mercer and Reynolds,53 55 which was the most frequently cited definition found in this scoping review. The JSPPPE was developed based on various literature sources; however, these sources were not specified by the original authors.56 The Warmometer was developed based on theoretical assumptions about warmth between humans, though the original paper notes that it found that ‘physician warmth’ is a broader and more genuine concept than physician empathy as it combines multiple personality attributes.57 This seemingly contradicts the study included in this scoping review that translates and validates the JSPPPE for use in Brazil,47 which appears to define empathy as ‘perceived warmth’, the ability of someone, either the healthcare provider or patient, to share and understand the emotions of others.

Types of research about empathetic communication

Of the 21 studies measuring empathetic communication from another person’s perspective, six (29%) compared patient-rated provider empathy measurements with other measures such as treatment outcomes and patient satisfaction.25–27 29 35 43 Five studies (24%) identified empathy or empathetic responses thematically,24 34 40 44 48 by coding physicians’ responses to patients expressing negative emotion as empathetic, neutral and non-empathetic,40 by coding nurse and patient interview responses into aspects of empathy,24 34 48 and coding instances of empathy into three themes: understanding patients’ experiences, communicating that understanding and acting on the understanding.44 Two studies (10%) compared patient-rated provider empathy measurements between two groups—family medicine and hospital consultations33 and in-person and telemedicine consultations.42 Two studies (10%) aimed to validate patient-rated provider empathy measurement tools (CARE23 and Warmometer47), both assessing empathy in the process. Two studies (10%) compared patient-rated empathetic communication measurements after an intervention: asking a patient a question about dignity30 and after simulation-based empathetic communication training.39 One study measured active empathetic listening of nurses from the patient’s perspective.41

Discussion and conclusion


Although many definitions were found that describe communication as part of empathy, there was a lack of consistency across them. Some tended to describe empathy as an ability or capacity to communicate, while others described it as a dynamic process. Similarly, several definitions defined empathy using terms like ‘clinical empathy’ and ‘physician empathy’, seemingly restricting their definition to a particular group of clinicians despite appearing similar to broader definitions of ‘empathy’. The value of professional distinctions is unclear in the context of the broad definitional similarities. Furthermore, we identified an instance of one empathy definition contradicting the tool used in the study,47 equating perceived warmth to empathy when the article discussing the tool’s creation clearly differentiates the two.57 Several studies noted the lack of clarity and numerous inconsistencies between definitions of empathy in the literature,27 28 39 44 48 and one study mentioned the difficulty of defining a vague concept like empathy.24 Without a clear definition of empathy, empathetic communication is not easy to build a consentient body of knowledge or develop high-quality and transferrable research within a hospital setting.

No definition of ‘empathetic communication’ was described in the studies despite multiple uses of the term in the included studies.23 25 26 33 37 40 44 48 While not positioned as a definition of empathetic communication, one study identifies features that are core components of it, these being the need to feel listened to, validated and understood,25 yet the three articles cited in this study to support the assertion do not define empathetic communication.58–60 Further research could aim to define empathetic communication and to explore the differences (or not) between it and empathy, especially given that many sources seem to consider communication as part of empathy. From this definition, meaningful and useful behavioural measures could be created that can be used within clinically focused research that can be replicated across settings.

None of the identified definitions of empathy included a component relating to culture. This may be a result of empathy being traditionally defined by clinicians and from Western countries.61 There are linkages between the concept of empathy and culture,61 62 and articles about empathy research from a wide range of countries have been included in this review (despite articles in English only being included), which suggests that future research should consider cultural influences and differences when crafting a clear definition of both empathy and empathetic communication. We argue that culture is critically important in healthcare and future definitional work ought to be conducted in partnership with communities.

Nearly a third of the unique definitions split empathy into cognitive and affective components. While other components such as behavioural empathy were less commonly described,24 46 features of these and cognitive/affective empathy were often described inexplicitly. Few studies suggested that empathy was a linear process, which suggests that this would likely not be part of a clear definition compared with the different components of empathy described. These issues need to be explored before research tools can be developed, tested and replicated within clinical environments.

One study in this review cited a 1994 article when describing two definitions of empathy.44 The authors describe ‘empathic communication’ as developed by their experiences with a physician–patient communication workshop and mention that ‘complete communication of this sort includes appreciation of the patient’s feelings, support of his or her past actions, and, often, promise of help in the future’ in reference to physician–patient interactions.63 If this was to be considered a definition of ‘empathetic communication’, it shares the importance of understanding another’s feelings with many of the definitions found in this review,23–29 32–39 43 44 46–48 though none mention supporting another’s actions or promising help, which may be due to this definition being developed specifically for the physician–patient relationship. They also mention that they define empathetic communication as a cognitive skill, that is, ‘an ability to take another’s viewpoint, infer his feelings and put oneself in his shoes’ rather than an affective one—’an ability to construct for oneself another’s emotional experience’.63 However, putting ‘oneself in another person’s shoes’ was also described as emotional/affective empathy by two definitions found in this review,34 46 which suggests that cognitive and emotional/affective empathy are possibly often used interchangeably.

Several of the definitions of empathy found in this review mention that empathy involves conveying an understanding of both cognitive and affective empathy, but a precise definition of affective empathy appears to be unclear. This is relatively unsurprising given the inconsistencies seen between the identified definitions. Other reviews examining how empathy is characterised in literature since 200164–66 also obtained similar results to our findings in regards to developing a robust research platform for future research. The authors identified that researchers tended to identify empathy as a multidimensional concept that aligns with many definitions we identified, and they further identified multiple inconsistencies between definitions. They discussed the benefits of avoiding using words such as ‘empathy’ and instead describing what an author means by empathy directly, which could possibly avoid confusion and debate regarding its definition; the authors of this paper concur with the use of language that directly describes what is meant and is being measured.

Practice implications

The numerous differences between empathy definitions could potentially lead researchers to base their study methods using definitions of empathy that widely differ from one another, leading to potential inconsistencies within empathy research. For example, a qualitative study found in this review coded instances of empathy according to a linear model of empathy,48 while two other qualitative studies that were included coded instances of empathy as components of multifaceted non-linear definitions of empathy,24 34 thus highlighting incompatibility. This could ultimately make it harder to synthesise empathy research findings into delivering effective, therapeutic care using empathetic communication. A concept analysis would lead to the identification of key components and a definition of ‘empathetic communication’ from which tools could be developed for use in research.


This scoping review has some methodological limitations. For example, the method requires that a rigorous search strategy be employed so that other researchers can reimplement this particular review. In doing so, researchers are required to identify search terms and discipline boundaries that restrict the type of articles that will be identified. In this instance, the terms ‘empathy’ and ‘empathetic communication’ formed the basis of the inclusion criteria; as a result, sibling concepts were not explored. Equally, research that used measurement tools specifically designed to measure empathy were not included in the review because empathy or empathetic communication was not research objective. For example, researchers examining healthcare interactions using the CARE measurement tool widely used to measure empathy, were missing from this review if the research focused was not specific to empathy. This is a limitation of scoping review design and highlights the need for consistency of definitions, and measurement thereof, in clinical research.


No precise definition of ‘empathetic communication’ within the hospital setting is identifiable, and there is a considerable lack of a consistent and clear definition of empathy. For example, the role of communication in empathy varies between being described as an ability to communicate and also as a dynamic process. However, among the 33 unique definitions of empathy found in this review, common themes arose—that communication is an important part of empathy, that empathy is not a linear process, and that different components such as cognitive and emotional empathy exist. These findings can potentially pave the way for future research to develop a consistent definition of empathy and empathetic communication for use in clinical settings. The findings of the review highlight that there is additional work needed to define ‘empathetic communication’ and associated behaviours that would lead to the development of observable clinically focused measurement tools for use in research; the first step being a concept analysis of the term.

Data availability statement

No data are available.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.


We would like to acknowledge Rosemary Rowe, Subject Librarian, Victoria University of Wellington, for her guidance in completing the literature search and retrieving included articles.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Twitter @KarenMcBride-Hen1, @caz_hales, @Helen_Rook

  • Contributors HR, CH and KM-H led the conceptualisation and design of this work. Data extraction was conducted by JH-T. HR, CH, KM-H and JH-T contributed to the screening and selection of papers. Paper drafting was conducted by JH-T and KM-H. All authors made substantial contributions to the critical revision of the work and approved the final manuscript. CH and HR are acting as guarantors.

  • Funding This work was supported by a Victoria University of Wellington University Research Fund Grant Number [223854].

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.