Introduction

Evidence suggests people with cancer use urgent and emergency care (UEC) services more than the general population [1,2,3], often presenting with ‘high acuity, high symptom burden, and frequent need for admission’ ([4] , p.9). UEC use by people with cancer is expected to rise in tandem with an ageing population and increased survival [5]. Reviews of research have identified factors associated with UEC use by people with cancer (e.g., symptom clusters, demographic factors) [4, 6,7,8] and interventions which aim to prevent emergency care use by this population [9, 10]. Psychosocial factors that might influence usage, and preferences and experiences of informal caregivers, remain neglected areas of research [11]. To our knowledge, there is no overview of studies that has focussed on how people with cancer make decisions about which UEC service to access when acutely unwell. Understanding which services and patient groups have been studied, and what factors influence their choices, is needed to inform ways in which we might support patients to effectively access UEC for complications of cancer and its treatment.

Navigating UEC systems can be difficult for people with cancer [12]. The concept of ‘work’ has been used to understand the decisions, activities, and tasks undertaken by patients and informal caregivers to manage illness, the consequences of these actions, and the challenges encountered [13]. In the Model of Urgent Care Help-seeking, Turnbull et al. [14] describe how decisions about UEC service use are, in part, the product of ‘navigation work’ undertaken to make sense of the availability, accessibility, and acceptability of services. Turnbull et al.’s model [15] describes how this work is either undertaken alone (individual level navigation work) or with the support of others, such as family or friends (social network level navigation work). The model also describes how time of day and social contexts influence this workload (socio-temporal navigation work) [15]. To understand the extent to which international research has studied this phenomenon, this scoping review aimed to achieve the following: (1) describe research undertaken about choices made by people with cancer about routes to access UEC, (2) characterise how people with cancer decide to use which service in terms of ‘navigation work’, and (3) identify priorities for research.

Methods

Scoping reviews aim to map the breadth and depth of research in a field to answer a broad research question [16]. In contrast to systematic reviews, scoping reviews aim to ‘extract the essence of a diverse body of evidence’ ([17], p.1398) and can be used to clarify key characteristics of a concept and identify knowledge gaps in an emerging field [18]. Building upon preliminary findings [19], this review aimed to answer the question as follows: what research describes how adults with cancer and their informal caregivers navigate urgent and emergency care? We followed Arksey and O’Malley’s [20] framework, which comprises five stages: (1) identifying a research question, (2) identifying relevant studies, (3) study selection, (4) charting the data, and (5) collating, summarising, and reporting results. PRISMA guidelines [21] for scoping reviews were followed.

Identifying relevant studies

We retrieved literature in two stages: (1) electronic database searching and (2) ‘berrypicking’. Our review focused on research published since 2000 due to changes in UEC delivery models that have taken place internationally since the end of the 1990s [22].

Electronic database search

Search terms were developed from subject headings (e.g. MeSH terms) and informed by key publications, such as Mills et al. [6] and Turnbull et al. [14], to capture research describing decisions to use different UEC services. Search terms and subject headings were divided into 4 categories: (1) population (people with cancer or their informal caregivers); (2) concept (help-seeking decisions); (3) context (UEC services); and (4) study (qualitative methods). Electronic databases were searched from January 2000 to May 2021 and included the following: AMED, CINAHL, MEDLINE, PsycInfo (via EBSCO), and Embase (via Ovid). Searches were undertaken June-July 2021. An example electronic database search strategy is displayed in Additional File 1.

Berrypicking

We drew upon Booth et al.’s [23] cluster searching method, applying ‘berrypicking’ strategies to search for potentially relevant papers co-located with or related to papers included via electronic database searching. Techniques reflected four of Bates’ [24] berrypicking strategies: (1) footnote chasing, (2) citation searching, (3) author searching, and (4) area scanning. The procedures followed for each strategy are displayed in Additional File 2.

Study selection

We used pre-specified eligibility criteria to screen papers for relevance (Table 1). Titles and abstracts of retrieved papers were screened by the principal reviewer (J. D.); where potentially relevant, citations were imported into EndNote™ 20 (Clarivate™) and de-duplicated. Full-text papers that met the eligibility criteria were included for data extraction. Uncertainty about whether a paper met the criteria was resolved by discussion with the review team (R. W. and A. R.). The process is displayed in the PRISMA flowchart [25] in Fig. 1.

Table 1 Eligibility criteria
Full size table
Fig. 1
figure 1

PRISMA flowchart

Full size image

Quality appraisal

As is convention in scoping reviews [18], and to ensure findings represented the breadth of research literature, no papers were excluded on the grounds of quality.

Data extraction and analysis

Data were charted using a template developed a priori and with reference to the Model of Urgent Care Help-seeking [15]. A two-stage approach to collating and summarising data was undertaken: (1) descriptive mapping of study, participant, and service characteristics and (2) directed qualitative content analysis of data amenable to interpretation as navigation work.

Stage 1: mapping study, participant, and service characteristics

Data extracted and analysed in stage 1 included the following: publication details (e.g. country of origin); study design and methods (e.g. theories enrolled); patient participant characteristics (e.g. cancer stage); informal caregiver characteristics (e.g. relationship to patient participant); and services studied (e.g. ambulance).

Stage 2: content analysis of author and participant quotations

We used qualitative content analysis [26] to characterise the work involved in navigating access to UEC services. Author and participant quotations amenable to interpretation as ‘navigation work’ were extracted from the ‘Results’ sections of papers and treated as data. When studies also focussed on help-seeking for other conditions, data were extracted only if clearly relevant to people with cancer. Directed qualitative content analysis was undertaken following the method described by Assarroudi et al. [27]. We theoretically defined coding rules and categories with reference to Turnbull et al. [14] and piloted the coding matrix on data extracted from five papers which studied five different UEC services. Data were managed in Excel® (Microsoft®) and analysed by the principal reviewer (J. D.); coding decisions were discussed in data analysis meetings with the review team (R. W. and A. R.). Subcategories were developed by comparing and interpreting similarities and differences in meaning across groups of codes [27] (in pursuit of the latent content [28]).

Results

Searches identified 21,723 potentially relevant papers. Stage 1 (electronic database searching) identified 19,561 citations, and stage 2 (berrypicking) identified 2162 citations. After de-duplication and screening, 90 full-text papers were assessed for relevance, of which 18 studies [29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46] met the eligibility criteria and were included for data extraction (Table 2).

Table 2 Summary of included studies
Full size table

Characteristics of studies, participants, and services

Study, participant, and service characteristics are mapped in Figs. 2 and 3.

Fig. 2
figure 2

Map of participant characteristics

Full size image
Fig. 3
figure 3

Map of service characteristics

Full size image

Study characteristics

The majority of studies were conducted in Europe (n = 12/18) [30, 32,33,34,35, 37,38,39, 41,42,43, 46] and North America (n = 5/18) [29, 31, 40, 44, 45], with one [36] originating from Australia. Most European studies (n = 10/12) [32, 33, 35, 37,38,39, 41,42,43, 46] were conducted in the United Kingdom (UK); two studies [30, 34] originated from Denmark. North American studies were conducted in the United States (US) (n = 3/18) [31, 44, 45] and Canada (n = 2/18) [29, 40]. Most studies (n = 14/18) [29,30,31,32,33,34,35,36,37,38,39,40,41,42] were published after 2014. Almost all reported qualitative studies (n = 16/18) [29,30,31, 33,34,35,36,37,38,39, 41,42,43,44,45,46] with one [34] reporting a secondary analysis of another included paper by the same authors. The remaining papers (n = 2/18) [32, 40] used mixed methods. Most qualitative studies (n = 14/16) used qualitative description; one paper reported a phenomenological study [30], whereas the other [37] reported a constructivist grounded theory approach. Of the mixed methods studies, one described using a concurrent triangulation design [40], while the other [32] did not make the design explicit. Both used qualitative descriptive methods in the qualitative component [32, 40]. Many of the papers (n = 13/18) did not explicitly enrol theory; theories and conceptual frameworks enrolled by the remaining (n = 5/18) studies were the following: ‘Model of Healthcare Utilization’ [29], ‘Burden of Treatment Theory’ [33], ‘Cumulative Complexity Model’ [35], ‘Model of Emergency Department Use’ [38], and the ‘Utilization Process Model’ [40].

Participant characteristics

Characteristics of 339 people with cancer who had used UEC were described by the 18 studies [29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46]. Type of cancer was specified for over two-thirds of patient participants (n = 230/339) [29,30,31,32,33,34,35,36,37,38,39,40,41,42]. Most patient participants had solid cancers (n = 192/230) [29,30,31,32,33,34,35,36,37,38,39,40, 42]; over one-third (n = 67/192) [29, 31, 32, 35, 36, 38, 39] had lung cancer, with breast (n = 32/192) [29,30,31, 34, 37, 40, 42] and lower gastrointestinal tract (n = 19/192) [29, 30, 34, 35, 40, 42] cancers being the next most common diseases. Only 38 patient participants across five studies [29, 36, 38, 40, 41] had haematological malignancies. Cancer stage was made explicit for all participants in 11 studies [30, 34,35,36,37,38,39, 43,44,45,46] representing approximately two-thirds of patient participants (n = 233/339); almost all (n = 217/233) had advanced cancer [29,30,31, 33,34,35,36,37,38,39, 42,43,44,45,46]. A minority were described as receiving anticancer treatment (n = 161/339) by half of studies included [29, 31, 32, 35, 37, 39,40,41, 43]. Treatment provided was specified for more than half (n = 101/161) of these participants [29, 32, 35, 37, 39, 41, 43]; however, only five studies [31, 32, 37, 40, 41] explicitly focussed on patients’ decisions and experiences during anticancer treatment. Where specified, chemotherapy was the dominant (n = 81/101) treatment modality [29, 35, 37, 39, 41, 43]; 13 patient participants had received immunotherapy [32], and seven had received radiotherapy [29]. For one paper [35], data extraction about anticancer treatment modality was limited by the way participant characteristics were reported. One study [40] recruited participants (n = 5/339) with comorbidities; however, this was not the focus.

Most studies (n = 14/18) [29, 30, 33,34,35,36,37,38,39, 41,42,43, 45, 46] recruited informal caregivers (n = 166). For three papers [33, 39, 45], it was largely unclear whether participants were informal caregivers of people with cancer or other diseases; this limited data extraction from these papers. Four studies did not recruit informal caregivers [31, 32, 40, 44]. Informal caregivers’ relationships to patient participants were specified for fewer than half (n = 76/166) of caregivers about whom data were extracted [29, 30, 33, 34, 36, 37, 39, 41, 43, 45, 46]. The most frequently reported relationship to patients was spouses or partners (n = 55/76) [29, 30, 33, 34, 36, 37, 39, 41, 43, 45, 46], followed by grown-up children (n = 8/76), [29, 30, 34, 36, 43], and parents (n = 5/76) [29, 36, 37]. There were no dedicated studies of the help-seeking experiences of informal caregivers.

Service characteristics

Most studies (n = 12/18) [29,30,31, 33, 34, 36, 38, 40, 42, 44,45,46] focussed on decisions to use a single UEC service. The majority (n = 8/12) [29, 31, 33, 36, 38, 40, 44, 45] studied help-seeking from emergency departments (ED); a smaller number focussed on specialist emergency helplines (n = 2/12) [30, 34] and out-of-hours primary care services (n = 2/12) [42, 46]. Two studies [35, 39] focussed on the events leading up to emergency hospital admission from various routes; however, most participants had used a single service: a specialist emergency helpline [35] or an ED [39]. Three studies [32, 37, 41] focussed on the decision to report symptoms suggestive of anticancer treatment toxicity without specific focus on individual UEC services. Of UEC services used by patients in these studies, most had contacted [32, 41], or discussed decisions with reference to [37], a specialist emergency helpline. None focussed on non-use of services, and only one paper [43] studied help-seeking from a UEC system.

Directed qualitative content analysis

Six subcategories reflecting the navigation work undertaken at individual (‘Subcategory 1: making decisions with, and seeking help from, specialists’, ‘Subcategory 2: seeking safety’, ‘Subcategory 3: positioning to access the desired treatment’, ‘Subcategory 4: negotiating tortuous pathways to help’), social network (‘Subcategory 5: making decisions in collaboration with caregivers’), and socio-temporal (‘Subcategory 6: managing isolation from services and social networks’) levels were identified. Included papers have been cited as source documents from which data were extracted, condensed, and interpreted to develop groups of codes for each subcategory.

Subcategory 1: making decisions with, and seeking help from, specialists

People preferred to be guided by specialists whom they knew and trusted [31, 35, 37, 38, 43]. Familiar clinicians from cancer and palliative care were consulted as part of the decision to use UEC [31, 38, 40, 45]. People with cancer were directed to UEC on the advice of specialists whom they contacted in the first instance [29, 31, 33, 35, 38, 40, 45] or used prior verbal and written instructions from their cancer centre to choose which service to use [30, 34, 36, 38, 39]:

All these instructions are basically on the card I’ve got. It tells you to go to emergency once your temperature gets up over 38. (Patient participant; ED [36], p.442)

They often considered specialists the only safe option from whom to seek help [30, 33, 34, 43], with a perceived lack of cancer-related expertise deterring many from using non-specialist UEC [31, 46]. Familiarity with their specialist team meant that some deferred seeking help until routine clinic appointments, even when specialist emergency helplines were available [32, 37]. Ultimately, attending the ED was a last resort when they failed to reach specialists for advice [29, 40], when community care was deemed unviable [36, 38, 39, 44], or when cancer-specific UEC was unavailable [29, 33, 36].

Subcategory 2: seeking safety

Fear and anxiety (provoked by uncontrolled symptoms or uncertainty about which course of action to take) influenced which services were selected [31, 40, 46]. People with cancer chose to attend ED when faced with severe symptoms [33, 40, 44], and many were motivated by the comfort and safety they found in the hospital environment [31, 36, 38, 46]:

Here [in the hospital], I’m safe [if] something happens. (Patient participant; ED [31], p.e1295)

Feeling safe at, and trust in, the cancer centre where they received treatment provoked help-seeking from UEC aligned to these institutions [36, 38]. Services were chosen to share the burden of responsibility for self-management [34, 38], and people sought reassurance from specialists to feel safe when working to manage deteriorating symptoms at home [30, 31, 34].

Subcategory 3: positioning to access the desired treatment

People with cancer sought help from services with clear objectives in mind; services were selected to support symptom self-management [31, 32, 34, 35] or, when faced with deterioration, to provide rapid symptom control [29, 38, 39, 42]. Previous experiences of cancer-related UEC influenced re-use (or not) of services [37, 39, 40, 42, 46]. Symptoms and previous supportive treatment guided service reselection with the intention of receiving specific interventions again [29, 45]. Seeking help from services that involved multiple stages and re-explanation to multiple professionals was burdensome and avoided where possible [33, 38, 42, 43, 46]:

… they would have to go through someone else to go through someone else (Researcher: Mmm) do you know what … I wouldn’t want anything like that. (Patient participant; ED [38] , p.7)

People bypassed parts of the UEC system that would result in unnecessary work, either because utilising alternative services was perceived as futile [38, 46] or to seek help from specialists via unofficial channels [38, 43]. Indeed, they positioned their help-seeking to achieve onward care as much as initial symptom control; people with cancer chose to attend ED as a way of brokering access not only to a hospital admission [29, 36, 38] but also to cancer wards [29, 36].

Subcategory 4: negotiating tortuous pathways to help

Action was taken following careful appraisal of a range of options [33, 38, 39, 43]. For many, the path to obtaining help was convoluted and characterised by clusters of escalating contacts with (often multiple) services [33, 34, 36, 38, 43]:

I phoned the daughter up and told her what had happened … she said ‘oh, hold on I'll be round’ … so she said ‘I better ring up, 111, just to get a bit of advice’, so, she phoned them … they said ‘well we think he better go to the local hospital’, so that's how they got the ambulance (Patient participant; ED [33] , p.4)

Contacting one service often resulted in being given more navigation work to do, either by being asked to select another service from which to obtain onward treatment [35, 39, 45] or to ‘phone back in the morning’ ([32] , p.7) if telephone-delivered services could not manage the problem out-of-hours [32, 46]. Over time, many people with cancer had to learn how UEC services were organised to negotiate their way through the system effectively [33, 46].

Subcategory 5: making decisions in collaboration with caregivers

People with cancer and their informal caregivers often worked together [31, 37, 39] but deciding when to use, and initiating contact with, UEC was often led by caregivers [31, 33, 38, 39, 42, 43, 46]. Navigation work was sometimes a collaborative effort between patients, their caregivers, and community [38] or pre-hospital [39] practitioners. In the event of serious illness or brisk deterioration, decisions were taken by caregivers on patients’ behalf [34, 35, 46]. However, decisions about the need to use UEC often precipitated patient-caregiver conflict; caregivers applied increasing pressure if patients minimised symptoms or delayed seeking help [31, 37, 38]. Indeed, caregivers sometimes overrode patients’ decisions by initiating contact with UEC services against their wishes if perceived to be making an unwise choice [31, 41, 43]:

The sicker he gets the less he wants any intervention, but I now know that I have to quickly overrule him. (Caregiver participant; specialist emergency helpline [41], p.2690)

Subcategory 6: managing isolation from services and social networks

Deciding which service to use was influenced by the isolation people with cancer felt when facing deterioration alone [30, 43]. Distant family members were contacted and converged on patients to reduce the burden of accessing UEC [31, 33]. For caregivers however, the potential negative impact on patients that travelling to obtain help may incur was carefully considered when deciding which service to use but also constituted a burden [34, 43, 46]:

She’s suffering the pain and I am stressing, and I am thinking do I get in the car? Do I create more pain? You know, it’s all these sorts of things that go through your mind. (Caregiver participant; out-of-hours service [43], p.174-175)

Unfamiliarity with out-of-hours systems made decision-making harder; the unavailability of specialists made deciding which service to use more difficult [40, 43]. During this period, people weighed up whether to wait until in-hours services resumed [40, 46] and accessed emergency care overnight and at weekends when no alternative remained [29, 40].

Discussion

This scoping review aimed to map what qualitative research exists about decisions to use UEC by people with cancer. It identified a small body of research, homogenous in terms of the following: (1) design and methods used, (2) patient populations recruited, and (3) services studied. The review also sought to characterise the ‘navigation work’ undertaken by people with cancer and their informal caregivers. It has also identified areas that would benefit from further research.

The focus of research to date has been about help-seeking by people with advanced cancer, not people receiving anticancer treatment. A minority of studies [32, 37, 41] focussed on and were explicit about the anticancer treatment received for all participants, the majority of whom were treated with chemotherapy. This is significant as the processes of care associated with planning and delivering anticancer treatment, and symptoms and trajectories of complications, differ by treatment modality. Our understanding of how these factors might influence decisions to use UEC (or not) is limited by the evidence. Furthermore, there is a surprising lack of focus on the sequelae of advanced cancer known to necessitate emergency care. For example, we could only identify one instance [45] where a patient was explicitly described as seeking help for metastatic spinal cord compression. No studies focussed on the decisions to use UEC by people with cancer who were multi-morbid. Evidence suggests this population perceive healthcare systems to be fragmented and difficult to navigate [47] and experience convoluted pathways through acute care [48]. Further research that focuses on the decisions and experiences of these patient groups is needed.

Individual level navigation work

Findings from the content analysis suggest navigating UEC entails different types of work for people with cancer compared to other groups of patients (‘Subcategory 1: making decisions with, and seeking help from, specialists’, ‘Subcategory 2: seeking safety’, ‘Subcategory 3: positioning to access the desired treatment’, ‘Subcategory 4: negotiating tortuous pathways to help’). Our findings show people with cancer prefer to seek help from cancer and palliative care services to whom they are known and work hard to access these via UEC. This contrasts with previous research [49] which has shown people with long-term conditions (excluding cancer) select services for their perceived technological capability rather than care delivered by familiar clinicians. Studies continue to demonstrate the importance of trusting relationships with care providers to people with cancer [50,51,52], and the findings of our review suggest the concept of ‘being in safe hands’ ([53], p.1538) during scheduled cancer treatment may influence which UEC service is selected when acutely unwell. In comparison with the Model of Urgent Care Help-seeking [15], our findings suggest acceptability of services is more important than their availability, with acceptability contingent on feeling safe at, and trust in, services’ ability to meet the unique needs of people with cancer.

Similarly to Turnbull et al. [14], decisions about which service to use were ‘recursive’ (i.e. shaped by prior help-seeking experiences). Findings from our review suggest decisions are informed by experiences since a diagnosis of cancer; however, recent evidence [54, 55] suggests frequent ED attendances prior to a cancer diagnosis are associated with increased ED use following diagnosis. Further qualitative work is needed to explore which experiences, both prior to and following cancer diagnosis, influence decisions about future UEC use and why. In contrast to the Model of Urgent Care Help-seeking [15], our findings show decisions made by people with cancer are predicated on detailed appraisals of their clinical need rather than convenience. Seeking help from the ‘right’ service constitutes hard work; people with cancer are keen to avoid, but often face, burdensome access procedures, which they attempt to circumvent by bypassing parts of the system. This is significant as international evidence has largely described ‘bypassing’ in terms of avoiding primary care when choosing secondary care [56,57,58] and suggests systems, and the work required to navigate them, may be more complex for people with cancer. However, our understanding of how people with cancer make sense of the UEC landscape is limited by a literature that has largely studied services in isolation.

Social network level navigation work

Our review shows how informal caregivers perform a dual role in help-seeking (‘Subcategory 5: making decisions in collaboration with caregivers’). Caregivers not only support people with cancer to navigate UEC but often take charge of this workload. This finding contrasts with those of Turnbull et al. [15] who describe how navigation work by the general population predominantly takes place at an individual level, with relational network members consulted as part of, but not responsible for, decision-making. It also contrasts with findings from studies of help-seeking for complications of long-term conditions in which caregivers are largely described as fulfilling encouraging and facilitative roles [59,60,61]; acting without patients’ permission was rarer [62]. It is therefore significant that this phenomenon has been captured by the small body of research identified by this review. Evidence suggests informal caregivers of people with cancer act as both ‘advocate’ and ‘protector’ ([63], p.803), assuming responsibility for decisions about [64] and negotiating access to care [65] on patients’ behalf. However, support for informal caregivers to learn how to navigate cancer health systems is lacking [66]; further, dedicated exploration in relation to UEC is needed.

Socio-temporal contexts

Navigation work by people with cancer also appears to be influenced by both time of day and social contexts (‘Subcategory 6: managing isolation from services and social networks’). Unfamiliarity with the UEC system out of hours and unavailability of usual support structures made deciding whether to seek help, and which service to use, more difficult. Difficulty relating to prior information at the point of deterioration confounded decision-making, a finding echoed by wider literature [12, 67]. A decade ago, a report from the UK [68] highlighted the need to embed personalised contingency planning into the routine care of people at risk of complications of cancer and its treatment. To the best of our knowledge, how people with cancer and their informal caregivers prepare for complications that necessitate contact with UEC remains an area that has been unexplored by dedicated research. Our findings reinforce the need to understand how preparing for acute illness might influence the work of navigating UEC.

Recommendations for research

This scoping review has identified a small body of literature and significant knowledge gaps. Although policy and practice would benefit from high-quality, theory-led studies that explore any aspect of UEC use by people with cancer, based on findings from this review, we have outlined five areas which should be prioritised in Table 3.

Table 3 Research priorities
Full size table

Limitations

Scoping reviews are time-consuming, resource-intensive projects and involve collating and interpreting large volumes of data [69]. For pragmatic reasons, the process of identifying, selecting, and extracting data from studies was undertaken by a single reviewer (J. D.). Multiple reviewers may have mitigated against potentially relevant papers and data being excluded, a commonly reported limitation of scoping reviews [70]. Time and funding constraints precluded translation of articles not published in English, which may also have resulted in relevant articles being excluded. The relatively small number of, and the richness of data amenable to interpretation as navigation work from, studies published outside of the UK precluded a comparative analysis by country. In addition, no studies from low- or middle-income countries were eligible for inclusion. As such, our findings and research recommendations may not be relevant to all international contexts.

Conclusion

This review scoped research studies pertaining to decisions about using UEC by people with cancer and their informal caregivers that have used qualitative and mixed methods. It has identified and summarised a small amount of narrowly focussed research, namely studies of decisions to use EDs by people with advanced cancer who are not receiving anticancer treatment. This review suggests navigating UEC constitutes hard work and is qualitatively different for people with cancer. Our understanding about ways in which we might prepare and support patients to effectively access these services is constrained by the evidence base. There is a need to progress research in this area.