Objectives Use of the term ‘patient’ has been recently debated, compared with alternatives including ‘consumer’ and ‘client’. This scoping study aimed to provide an integrated view of preferred labels across healthcare contexts and countries to clarify labelling preferences of individuals accessing healthcare.
Design Scoping study.
Data sources A preliminary literature search using GoogleScholar, Medline, Embase and PsycINFO found 43 key papers discussing terminology for labelling individuals accessing healthcare services. We then used citation chaining with PubMed and GoogleScholar to identify studies discussing term preferences among healthcare recipients.
Eligibility criteria No date limits were applied, and all healthcare settings were considered. Primary research studies examining terminology preferences of individuals accessing healthcare, published in peer-reviewed journals were eligible.
Data extraction and synthesis All authors extracted data regarding preferred term and study characteristics, and assessed reporting quality of the studies using criteria relevant to our design.
Results We identified 1565 articles, of which 47 met inclusion criteria. Six articles that examined preference for personal address (eg, first name) were excluded. Of the remaining 41 studies, 33 examined generic terms (‘patient’, ‘client’, ‘consumer’) and 8 focused on cancer survivorship. Of the 33 examining generic terms, 27 reported a preference for ‘patient’ and four for ‘client’. Samples preferring ‘client’ were typically based in mental health settings and conducted in the USA. Of the eight cancer survivorship studies, five found a preference for ‘survivor’, and three ‘someone who had had cancer’.
Conclusions Overall, healthcare recipients appear to prefer the term ‘patient’, with few preferring ‘consumer’. Within general clinical and research contexts, it therefore seems appropriate to continue using the label ‘patient’ in the absence of knowledge about an individual’s preferences. Reasons for preferences (eg, familiarity, social identity) and the implications of labelling for healthcare have not been investigated adequately, necessitating future empirical (including qualitative) research.
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: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This is the first study to review patient preference for terminology across a range of healthcare settings.
Because of the range of healthcare setting included, we could compare preferences between these settings, providing a comprehensive overview of the literature.
The heterogeneity of study designs and quantitative methods precluded the possibility of conducting quantitative synthesis (meta-analysis).
An individual who sought the assistance of a healthcare professional was once almost universally described as a ‘patient’. Various sociopolitical forces, including the commodification of healthcare and the gradual shift from the paternalistic to the patient-centred model of healthcare, have prompted debate about the use of alternative labels to describe ‘patients’. Whereas the term ‘patient’ may summon an image of a passive individual awaiting treatment, terms such as ‘client’, ‘consumer’, ‘customer’ and ‘service user’ may be more empowering, implying greater equality between participants in personal healthcare decision-making.1
Yet alternative terms to ‘patient’ have not gained mainstream acceptance, although ‘client’ has had some dominance in social work2 and psychotherapy, at least among those from humanistic traditions,3 likely stemming from Rogers' client-centred work.4 Many commentaries in peer-reviewed journals and other media have addressed this issue,1–3 5–13 most arguing for retention of ‘patient’, but with some exceptions.1 8 The arguments against continued use of ‘patient’ are that its etymology implies suffering and passivity, it is characterised by an unequal relationship and a new term would more accurately reflect today’s person-centred healthcare practice.1 Proponents for retention argue that a term’s etymology need not govern its current usage,9 other terms introduce other problems,7 the therapeutic relationship is not inherently one of equality3 and the status quo need not change given the absence of a widely acceptable alternative. Others have argued that neither ‘patient’ nor its alternatives are satisfactory.10 Preferences within the cancer survivorship context have also been questioned, with uncertainty regarding the relative merits of labelling people ‘cancer survivors’, ‘a person with cancer’ or ‘participants’ within the research context.
Despite this, a paucity of research has investigated what terms of reference ‘patients’ themselves prefer. Nor have there been attempts to comprehensively synthesise the literature; only commentaries citing isolated empirical studies. Furthermore, empirical studies of labelling preferences have spanned a range of healthcare contexts, including social work, mental health, occupational therapy and cancer treatment, and patient preferences may differ across contexts. The fact that multiple articles addressing the issue of labelling are titled ‘What’s in a name?'2 14–17 and fail to cite one another highlights the narrow confines within which this research has been conducted.
Debate has also centred on whether terminology dictates the behaviour of actors involved, and therefore the quality of the healthcare.6 8 Such conclusions appear based predominantly on the authors’ clinical experience with minimal reliance on empirical data (an exception is Goodyear and Parish18). Consequently, the effect of terminology in healthcare settings remains unknown. Labelling can generate stigma (both positive and negative), as suggested in a range of settings (eg, criminology, mental health). This in turn may give rise to congruent behaviours, consistent with labelling theory. Exploration of labelling preferences thus has important practical and behavioural implications. Widespread use of the term ‘patient’ may impede progress towards empowerment and shared decision-making; widespread adoption of the terms ‘client’ or ‘consumer’ may accelerate the commodification of healthcare. At present, evidence to support these or other possibilities is lacking.
Given the absence of resolution regarding healthcare recipients’ preferred labels and the increasing prevalence of ‘consumer’ representative groups, an integrated review of studies is necessary to understand labelling preferences and inform consideration of appropriate labelling. A previous review conducted in the mental health context observed a preference for ‘patient’,19 but no review that covers other healthcare contexts has been published. Therefore, we aimed to fill this gap and conducted a scoping review of the relevant literature to examine preferred terms of healthcare recipients across a range of healthcare contexts, with clinical and theoretical applications.
Data sources and searches
Initially, we planned to conduct a systematic search of Medline, Embase and PsycINFO, with search strategy based on a scoping search for relevant studies of terminology preferences of healthcare users and research participants. This scoping search (no date limits applied) involved (1) searching both Google and GoogleScholar using keywords patient/consumer, patient/client and patient/client/consumer and examining the first 100 matches for each, and (2) searching Medline, Embase and PsycINFO using patient AND client AND consumer AND preference (the restrictive nature of the latter search was chosen because of the very large number of results produced by less restrictive searches). We identified 43 papers on the general topic of terminology for individuals seeking healthcare (including commentaries as well as empirical studies on term preferences, only the latter of which are relevant to the present review). A comprehensive systematic search strategy (Medline, Embase, PsycINFO) from inception to 28 February 2018 developed using these keywords and Medical Subject Heading (MeSH) terms used in these papers identified an infeasible 652 006 records in Medline alone, owing to keywords and MeSH terms being varied and inconsistently applied. A random sample of 2000 records revealed no studies relevant to terminology preferences, and variants of the search strategy yielded similar results. Furthermore, some of the 43 papers were not indexed in Medline, Embase or PsycINFO. Due to the infeasibility of traditional systematic review methodology, we applied a targeted citation chain method20 21 to conduct a robust scoping review.22 23 We searched PubMed and GoogleScholar identifying articles from the reference lists of the initial 43 eligible source papers and their citing articles to identify additional literature on 28 February 2018 (updated December 2018). This collectively amassed a comprehensive, multidisciplinary collection of articles relevant to this topic. We included articles published in peer-reviewed journals discussing the terminology preferences of individuals accessing healthcare. We excluded grey literature (books, blogs, theses), discussion and commentaries to ensure conclusions were based on high-quality original research studies. No date restrictions were applied. Because of the importance of semantics in this study, only articles written in English were included. The results of this review are reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses24 statement, figure 1 and online supplementary appendix 1.
Supplementary file 1
All authors independently considered the potential eligibility of titles and abstracts generated by the search strategy. Full-text articles were obtained unless two or more reviewers deemed an abstract ineligible, with 10% of excluded records cross-checked. Each full-text report was assessed independently for final study inclusion. Disagreements about full-text inclusion were resolved by consensus. Meta-analysis was precluded by heterogeneity in outcome variables and labels employed. Four authors (DSJC, RM-B, ST, ZS) extracted data from an approximately equal number of articles, and all authors cross-checked the extraction for accuracy and consistency. Disagreements were resolved by consensus.
Data extraction and quality assessment
The reporting quality of the reviewed studies was assessed by all authors using criteria based on selected criteria from the Strengthening the Reporting of Observational Studies in Epidemiology statement for observational and cross-sectional designs25 (online supplementary appendix 2). Reporting quality assessments were conducted by a second reviewer for 10% of articles. Results of all articles were interpreted, however articles receiving a low-quality score (<25%) were excluded from secondary analyses to determine whether such studies influenced the results.
Supplementary file 2
Data synthesis and analysis
For each study we examined preference, ranking or rating of each term, and determined which was most preferred. We summarised these data taking into account the specific healthcare context in which each study was conducted, and how the question was asked (including specific phrasing, if provided). We also considered year of publication and country in which the study was conducted, as well as other variables potentially predictive of preference.
Patient and public involvement
As this is a literature review, no individuals other than the authors were directly involved in this study.
Our search strategy identified a further 1522 (in addition to the original 43 articles, for a total of 1565) articles. Of these, 47 met the criteria for full-text review (figure 1). The studies spanned 13 countries (20 UK, 11 USA, 6 Australia, 3 Ireland, 2 Canada and 1 each from Croatia, Israel, Italy, Korea, New Zealand, Norway, Poland and Trinidad and Tobago), published between 1990 and 2015, and covered a range of healthcare settings. Methods of preference elicitation varied across studies. Some studies asked participants how they would like to be labelled, others asked participants what they thought a particular group should be called, many studies were ambiguous on this point, and other studies did not specify the question posed. Most of the studies asked participants to nominate one or more preferred terms; fewer asked participants to rank26–30 or provide ratings for each term.28 31–33
Reporting quality ratings of included studies
Quality ratings varied from 21.4% to 100% across studies (mean=73.9%, SD=20.0). Two studies failed to meet the quality threshold of >25%: Cybulska34 and Probert,35 both of which were letters to editors that reported preferences for the term ‘patient’. Excluding these studies made no difference to the overall results of this review.
Summary of included studies
Eligible studies were categorised into three broad groupings. The first group (33 studies; see table 1) compared generic terms like ‘patient’, ‘client’ and ‘consumer’ across various healthcare contexts. The second group (eight studies; see table 1) focused on people with a prior cancer diagnosis and examined survivorship identity. Labelling preference options included ‘survivor’, ‘someone who has had cancer’, ‘conqueror’, ‘patient’ and ‘victim’. The third group (six studies)28 36–40 focused on forms of personal address, for example, ‘first name’, ‘last name’ and ‘title’. One of these studies41 examined preferences for both first/last name and patient/client/service user, and so is also one of the 33 in the first group. Only the first and second groups addressed the specific research question of preferred terminology in healthcare settings, hence results from only these groups (n=41) are explored in detail (see table 2). It is worth noting only that six of the seven studies28 36–41 examining forms of personal address indicated a preference for first name. The survivor identity studies are also reported separately given the specificity of this context, and their exclusion of certain key comparison terms (ie, terms like ‘consumer’ and ‘client’ were not given as options). The specific terms compared in each study are shown in table 2.
Of the 33 studies that solicited preferred terms for groups of healthcare recipients, ‘patient’ was most preferred in 27 studies. ‘Client’ was most preferred in four studies. One study42 showed differences in preferences based on the type of health professional, whereas another study16 found that preferred term changed after terms were defined, discussed further below. A study43 of women attending an antenatal clinic found ‘mother to be’ was preferred if the context was an information brochure, ‘pregnant woman’ was preferred for a medical journal or obstetrician talking to a midwife, and ‘patient’ was preferred when the respondent was talking to another woman. No other term was most highly rated in any of the studies, even though ‘customer’ (13 studies) and ‘consumer’ (15 studies) were commonly presented to respondents as options.
Cancer survivorship identity
Of the eight cancer survivorship identity studies, ‘survivor’ (five studies) and some variant of ‘someone who had had cancer’ (three studies) were preferred. When these options were provided, ‘patient’ was preferred only to ‘victim’.
Personal address versus general reference
Four studies asked how participants would prefer to be addressed in person, and the remaining 37 asked about a general preference, either in relation to how the individual would like to be referred to (24, including all 8 cancer survivorship studies), how people in general should be referred to (1), and the other 12 studies were unclear, either because the questions were ambiguous or were not quoted in the article.
For the 41 studies examining preferred terminology (as opposed to personal address with first name, last name, etc), the four studies that exhibited a preference for ‘client’ were conducted either in mental health settings (three studies) or among welfare service users (one study). In an additional study,42 ‘patient’ was preferred by respondents if told the health professional was a general practitioner, psychiatrist or community psychiatric nurse, but preference for ‘patient’ and ‘client’ were similar for psychologists and occupational therapists, and ‘client’ was preferred for social workers. One further study (also in mental health)16 required participants to state a preference a second time after having the meanings of the terms explained to them; after this explanation, ‘patient’ was preferred. Also noteworthy is that two of the four ‘client’-preferring studies were conducted in the USA, compared with two of 26 ‘patient’ preferring studies. No associations were found between preference and study characteristics in the survivorship identity studies. Year of publication did not appear to differ between studies exhibiting preference for ‘patient’ (median 2002, range 1990–2015) and those exhibiting a preference for ‘client’ (median 2004, range 1996–2008). Preferred term also did not appear to differ depending on whether respondents were inpatients, outpatients or other.
Correlates of terminology preferences
Many studies examined associations between preference and other variables. Synthesis of these results was difficult due to heterogeneity in study designs, measurement of variables and analytic methods. There was some evidence that older respondents tended to prefer ‘patient’ to other terms,14 42 44–46 although many studies found no association and some found that preference for ‘patient’ decreased with age.16 47 48 Three studies suggested that women have stronger preference for business-oriented terms, (eg, consumer, client) than men.16 48 49 In the survivorship studies, preference for ‘survivor’ was positively associated with psychological variables (eg, positive affect,50 benefit finding,51 lower rumination52), prior history of cancer,53 longer time since diagnosis54 and cancer treatment.52 55
This scoping review, which spanned several countries and healthcare contexts, found that most individuals receiving healthcare preferred the term ‘patient’ over alternative terms. ‘Consumer’ was not the overall preferred term in any study. Studies conducted in the cancer survivorship context indicated that individuals preferred the term ‘survivor’ and ‘someone who had had cancer’ over ‘patient’ and ‘victim’. Predictors of terminology preferences varied across studies and contexts.
These results suggest that the preference for the term ‘patient’ has persisted over time despite the movement towards person-centred care, shared decision-making and the commodification of healthcare, which might lead us to expect a growing preference for terms like ‘consumer’ and ‘client’. Although a large body of research demonstrates that preferences are complex and determined by both affective and cognitive factors,56 57 a robust and replicated empirical result within social psychology is the mere-exposure effect, in which an entity becomes preferred by an individual solely because the individual has had greater exposure to it.58 59 Thus, this strong preference for ‘patient’ may arise because of its familiarity to individuals seeking healthcare. One possible explanation for the mere-exposure effect is that repeated pairing of a stimulus (eg, the label ‘patient’) in the absence of an aversive outcome results in a classically conditioned response.59 Hence, switching to a less familiar term may introduce uncertainty, for example, with a change in terminology signifying a potentially undesirable change in healthcare delivery or approaches.
Similarly, Hoyt3 described the common usage of ‘client’ within humanistic approaches to healthcare (prompted by Rogers' client-centred approach to psychotherapy4), whereas ‘patient’ appears more popular in medical and psychoanalytic disciplines. Consistent with this and the findings of Dickens and Picchioni,19 the four reviewed studies indicating a preference for ‘client’ were set in either mental health contexts or among welfare service users (although several other mental health studies did indicate a preference for ‘patient’). Preferences are thus likely to be both contextually and individually determined.
Evidence also indicates that preferences are constructed in the elicitation process.60 That is, stated preferences are sensitive to methods through which they are obtained. Individual preferences may thus (at least in part) be determined by the framing of questions and may function to elicit a preference for those who might otherwise be impartial. Heterogeneity was apparent in the phrasing of questions in the reviewed studies, the most notable difference being whether the question asked for preferences of general reference or direct address (discussed further below). Questions were often vague or ambiguous in this respect; with such crude data, we can only speculate about the processes participants used in reporting preferences. The study by Thalitaya et al 16 is notable because spontaneous preference for ‘client’ was replaced with ‘patient’ once definitions of each were provided to participants. Individuals may thus intuitively prefer familiar terms, but alter their preferences when challenged to consider more deeply, an effect consistent with dual-processes (ie, intuitive vs deliberative) theory.61 62
Preference for terminology may also be indicative of something more stable and fundamental, such as the individual’s identity or perceived social role. ‘Client’ was preferred over ‘patient’ by welfare service users63 and individuals seeing social workers, but not occupational therapists.42 Welfare and social work may not be perceived as traditional physical healthcare contexts, so individuals seeing professionals in these fields may not consider themselves ‘patients’. The same may apply to mental health settings. This suggests the possibility that the same individual in different healthcare contexts (eg, traditional physical health vs mental health, acute vs chronic care) may represent different social roles. Role theory64 argues that social roles are guided by social norms, are often reciprocal, as in the dyad of individual seeking healthcare and their health professional65 and an individual may have multiple roles in different settings. Biddle66 argued that preferences are reflective of role expectations. Similarly, according to social identity theory,67 an individual’s sense of identity is at least partly determined by their group membership and may vary across settings (eg, the healthcare context).68 Individuals consulting with occupational therapists, like those who have ceased active treatment for cancer, may no longer consider themselves ‘patients’ and may more readily identify with some other label.
Parsons’ concept of the ‘sick role’,69 in which illness is considered sanctioned social deviance entailing certain rights and obligations, is particularly relevant. The emergence of the patient-centred model of care and consumerism in healthcare accompanied a decline in popularity of the ‘sick role’ concept. Because the concept of ‘patient’ is more congruent with the sick role than ‘client’ and ‘consumer’, one might expect preference for the term ‘patient’ to have declined over time. That our results provided no evidence for this may be explained by what role theory (specifically, interactional role theory) has to say about changing roles.64 Social pressures can lead to changes in roles, but certain conditions are required to facilitate such change, including a unified desire for change among the actors. Perhaps preference for ‘patient’ has resisted change because the push for change has come from sources other than the individuals seeking healthcare themselves.
Despite the rise of ‘consumer representative networks’, the negligible preference for the term ‘consumer’ (despite 15 studies providing this term as an option) may reflect a misalignment of the consumer movement with the desires of most individuals seeking healthcare. It would be instructive to examine preferences in patient advocacy settings (both in research and clinical applications) to determine if the term ‘consumer’ is in fact more acceptable within these networks. Furthermore, although labelling preferences were relatively clear across studies, the practical significance of how much terminology and its meaning matters to respondents is unclear. Only Mueser et al 70 addressed this, finding that approximately one-fifth of respondents selected ‘does not matter’ (although note that their question does not distinguish between personal address and general reference). Future research should explore this issue by asking respondents about the importance and meaning assigned to labels, and whether labelling influences quality of care provided.
Preferences may also vary cross-culturally. For example, the proportion of US-based studies exhibiting a preference for ‘client’ over ‘patient’ was larger than for other countries. Although this conclusion requires caution given most US-based studies reviewed were from mental health contexts, it suggests possible cultural differences in terminology preferences, warranting further investigation.
For research reporting purposes, whether preferences vary when directly addressing an individual as opposed to referring to the individual or labelling collectively is unclear. For example, in some studies ‘first name’ was nominated by participants in open-ended ‘other’ response options, implying that the participant interpreted the question as asking about personal address. Further, labelling context may impact preferences (as described above). Clearly, researchers do not intend that health professionals personally address an individual in a clinical setting as ‘patient’ instead of their name, or that published articles name each participant rather than using a collective label. Some studies were clear regarding the context in which terms would be used, either by making this explicit in the question (eg, "I would rather be addressed by a general practitioner as…"42) or providing alternative contexts. A notable example of the latter is Batra and Lilford,43 who found that preferred term depended on whether it would be used in an information brochure or medical journal, or in conversation involving health professionals or the respondents themselves.
The present review represents a critical first step in examining and delineating labelling preferences, and provides a basis for future research regarding cross-cultural and contextual applicability of labelling preferences, as well as the meaning assigned and importance of terminology preferences. Further research, particularly qualitative research, is warranted. The question of why certain individuals prefer certain terms has received very little attention (although Dickens et al 27 and Simmons et al 29 reported some qualitative explanation for preference). A more difficult but important question is whether labelling has behavioural implications for how healthcare recipients are treated and the quality of care received. There is evidence in the mental health setting that labelling behaviour as ‘deviant’ can have negative impact on the labelled individuals.71 Experimental approaches to this issue may have utility, with evidence suggesting that labels attached to individuals may foster attitudinal biases.18 If, for example, the term ‘consumer’ was to gain wider usage, would this prompt a further move towards treating consultations with a healthcare professional more like service encounters,72 73 and if so what are the further implications? Alternatively, would rejection of all terms but ‘patient’ return healthcare to the paternalistic model? With data currently unavailable, we can only speculate on such possibilities, necessitating further research building on the present study.
Limitations of the present review include heterogeneity in the study designs and outcome variables, which precluded a systematic review and meta-analysis. We also chose not to exclude studies conducted with non-English-speaking participants, introducing potential issues regarding comparability between translated labels. We deemed, however, that the utility of the information provided by these studies outweighed any problems of translation and meaning, particularly given that such problems are less likely for single terms than phrases, and that differences across English-speaking countries may be as large as language differences. Strengths of this review include the breadth of healthcare contexts and study designs included, increasing ecological validity.
In conclusion, the findings of this research may be applied in both clinical and research contexts through continued use of the term ‘patient’, when knowledge about a particular individual’s preferred label is lacking. We speculate that preference is partially determined by familiarity, social identity, the context of the role (eg, specific healthcare setting) and the preference elicitation method. These possible factors entail specific testable hypotheses that can be subjected to empirical examination, and the behavioural implications of labels used is crucial to determining whether labelling impacts the healthcare of individuals.
The authors would like to thank Dr Yulia Ulyannikova, librarian at the University of Sydney, for her time and effort in assisting with the literature search.
This manuscript is dedicated to our colleague the late Dr Anna-Lena López and her family.
Contributors DSJC contributed to the conception and design of the study, literature searches, decisions regarding inclusion and exclusion, quality assessments and the drafting of the manuscript. RM-B contributed to the conception and design of the study, literature searches, decisions regarding inclusion and exclusion, quality assessments and the drafting of the manuscript. ST contributed to the conception and design of the study, literature searches, decisions regarding inclusion and exclusion, quality assessments and the drafting of the manuscript. ZS contributed to the conception and design of the study, literature searches, decisions regarding inclusion and exclusion, quality assessments and the drafting of the manuscript. A-LL contributed to the conception and design of the study, literature searches, decisions regarding inclusion and exclusion, quality assessments and the drafting of the manuscript.
Funding This research was supported by the School of Psychology, University of Sydney.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Extra data are available by emailing Daniel Costa, firstname.lastname@example.org.
Patient consent for publication Not required.
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.