Article Text

The unmet needs of partners and caregivers of adults diagnosed with cancer: a systematic review
  1. Sylvie D Lambert1,
  2. James D Harrison2,
  3. Ellen Smith3,
  4. Billie Bonevski4,
  5. Mariko Carey4,
  6. Catalina Lawsin5,
  7. Chris Paul4 and
  8. Afaf Girgis1
  1. 1Translational Cancer Research Unit, Ingham Institute for Applied Medical Research, South Western Sydney Clinical School, The University of New South Wales, Sydney, Australia
  2. 2Surgical Outcomes Research Centre (SOuRCe), Sydney Local Health District and Sydney School of Public Health, Royal Prince Alfred Hospital, The University of Sydney, Sydney, Australia
  3. 3School of Psychology, Faculty of Science and Information Technology, The University of Newcastle, Newcastle, Australia
  4. 4Priority Research Centre for Health Behaviour, School of Medicine and Public Health, The University of Newcastle, Newcastle, Australia
  5. 5School of Psychology, The University of Sydney, Sydney, Australia
  6. 6School of Medicine and Public Health, The University of Newcastle, Newcastle, Australia
  1. Correspondence to Dr Sylvie Lambert, Ingham Institute for Applied Medical Research, South Western Sydney Clinical School, The University of New South Wales, PO Box 3151 (Westfield), Liverpool, NSW 2170, Australia; sylvie.d.lambert{at}gmail.com

Abstract

Objectives The recognition that a partner or caregiver is typically the patient's primary support person and is also deeply affected by the cancer diagnosis has prompted efforts to document their unmet supportive care needs. This review aimed to: (1) quantify the prevalence of unmet needs reported by partners and caregivers, (2) categorise their unmet needs by domain and (3) identify the main variables associated with reporting more unmet needs.

Methods Manuscripts were identified through systematically searching electronic databases, checking the reference lists of retrieved publications, online searching of key journals and contacting researchers in this field.

Results Unmet need items across 29 manuscripts were clustered into six domains: comprehensive cancer care (prevalence 1.1%–96%), emotional and psychological (3%–93.2%), partner or caregiver impact and daily activities (2.8%–79%), relationship (3.7% and 58%), information (2.2%–86%) and spiritual (2%–43%). Studies of caregivers of palliative care or terminal patients often reported a higher prevalence of unmet needs than studies of caregivers of cancer survivors. Variables associated with higher unmet needs included being female, not being the spouse of the patient, having lower social support or reporting distress.

Conclusions Despite the ability to classify unmet needs within broad domains, quantification of unmet needs was challenging. This was mainly due to the diversity in methods used across studies (eg, different measures, variability in conceptualisation of unmet needs, etc). Rigorous, context-specific, longitudinal studies that use validated measures are needed to benefit future intervention research.

Statistics from Altmetric.com

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.

Introduction

Individuals diagnosed with cancer report needs for emotional support, social and spiritual care, advice about managing symptoms and side effects, and high quality information.1 ,2 These aspects of care are typically classified under the umbrella term ‘supportive cancer care’. In light of the accumulating evidence reinforcing the value of supportive care in helping patients and partners and caregivers (P&Cs) cope with their needs and in reducing psychological distress,3 supportive care is recommended for all people with cancer and their P&Cs from the time of diagnosis.1 ,3

One prerequisite to the co-ordinated delivery of supportive care includes the comprehensive assessment of the extent to which patients' and P&Cs' needs are met.1 A need is typically labelled as ‘unmet’ when services required to deal with the particular issue are not received.4 From a healthcare delivery perspective, although ‘needs’ identify problems, ‘unmet needs’ provide additional information on where support deficiencies lie. This information can then be used to prioritise resources where unmet needs are high.4 The concept of ‘unmet needs’ has inherent appeal for policy makers, clinicians and researchers, reflected by the steady increase in the literature associated with this concept over the past decade.5 ,6

While much of the focus in supportive cancer care has been on assisting patients, there is growing recognition that P&Cs have their own unique needs.1 ,7,,9 P&Cs often see themselves as jointly ‘fighting the illness’ with patients10 and face multiple ‘ripple effects’ of the diagnosis and treatment.8 ,9 ,11,,13 For many, coping with cancer challenges can be physically, socially and psychologically challenging.7,,9 ,12 ,13 Some of P&Cs' most prominent supportive care needs mirror those identified by patients, and there is increasing evidence that many of these also remain unmet.14,,20 This is concerning, as unmet needs not only compromise P&Cs' quality of life14 ,21,,23 but also adversely impact on patients' distress.22 Hence, both patients' and caregivers' illness adjustment may be optimised if P&Cs' unmet needs are addressed.23 This systematic review aimed to: (1) quantify the prevalence of unmet needs reported by the P&Cs of adults with cancer; (2) categorise P&Cs' unmet needs by domain; and (3) identify the main variables associated with reporting unmet needs.

Methods

Methodological framework

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement was used.24

Search methods

Papers were primarily identified through a systematic search of the electronic Medline, CINAHL, CANCERLIT, AMED, Embase and PsycINFO databases. Secondary strategies included: (1) checking the reference lists of manuscripts retrieved; (2) online searching of key journals (Cancer, Cancer Nursing, Journal of Cancer Survivorship, Journal of Psychosocial Oncology, Oncology Nursing Forum, Patient Education and Counselling, Psycho-Oncology and Supportive Care in Cancer); (3) contacting researchers known to conduct work in this area; and (4) using the ‘find similar’ function in the databases for seminal manuscripts.

Search terms used

The ‘unmet needs’ search terms were modelled on those used by Harrison et al6 and included: need* assess*, need* survey, need* instrument, need* question*, unmet need*, health need*, info* need*, physical need*, emot* need*, psycho* need*, support* care need*, perceived need*, OR need satisfaction. The ‘partners/caregivers’ search terms were similar to those used by Northouse et al25 and included: support person, wife, wives, husband(s), close relative(s), next of kin(s), significant other(s), couple(s), partner(s), spouse(s), carer(s), family, families, relative(s), OR caregiver(s). ‘Cancer’ search terms used were: neoplasm, oncology, tumor, OR cancer. The ‘unmet needs’, ‘partners/caregivers’ and ‘cancer’ search terms were then combined using the AND function and the search was limited to ‘all adults’ and 1995 onwards (the last previous review was conducted in 1993).26 The same search terms were used across the databases; however, if a database specific MeSH relevant term was identified, it was then added. The full electronic search strategy used for one database (PsycINFO) is provided in supplementary online appendix 1.

Inclusion/exclusion criteria

Manuscripts considered for this review had to be published in a peer-reviewed journal and report primary data on the unmet supportive care needs of P&Cs of adults with cancer, from diagnosis to palliative care (excluding the bereavement phase). Studies published in either English or French were included. Qualitative studies were excluded, as it was often unclear whether a reported need was met or not. Studies of professional or paid carers were also excluded.

Study selection and data extraction

For all databases, two reviewers (SL and ES) independently ran the search terms, screened the titles, read the abstracts, obtained full texts and extracted the data using a data collection form based on the Cochrane Library's recommendations for data extraction.27 At each stage of the review process, disagreements were discussed between reviewers. For each manuscript, a third reviewer corroborated the data extracted and its inclusion in the review. Where results were unclear, authors were contacted for clarification.

Data analysis

Results across studies were pooled and the prevalence of unmet needs compiled. When reported, the average number of unmet needs was retained. To facilitate comparison across studies, and where appropriate, averages were standardised from 0 to 100. Most studies reviewed reported on specific unmet need items and to facilitate reporting, comparable need items were combined and clustered in domains. The labelling of these domains was guided by studies identified in this review.20 ,28

Results

Overview of studies reviewed

The search identified 3228 potentially relevant titles of which 3199 were rejected (see figure 1). Details of the 29 manuscripts retained are shown in online supplementary table S1. In almost three quarters of the studies, participants were the partner or spouse of the individual diagnosed16,,20 ,23 ,28,,40 and studies typically included a predominance of female P&Cs.14 ,17 ,22 ,23 ,29 ,31 ,32 ,35 ,36 ,38 ,39 ,41 ,42 When reported, the average age of participants ranged from 47 to 61 years16 ,17 ,20 ,22 ,28,,31 ,33,,36 ,43,,45 and time since diagnosis from 2 to 71 months.17 ,22 ,28,,31 ,37 Most studies included the P&Cs of individuals diagnosed with tumours from a range of sites and/or at various stages along the illness trajectory,16 ,17 ,19,,23 ,28 ,29 ,31 ,35,,37 ,41 ,42 ,45 which in turn limited the interpretation of findings along tumour specific variables.

Figure 1

PRISMA flow diagram of search process.

Overall prevalence of unmet needs

The reported mean number of unmet needs ranged between 1.3 and 16 (range measured: 17–67).14 ,17 ,18 ,22 ,30 ,35 ,41 ,45 ,46 Standardisation revealed that, on average, 5%–47% of needs remained unmet, with several studies finding that up to 10% of assessed needs were unmet.17 ,18 ,22 ,30 ,35 Higher levels of unmet needs were experienced by the P&Cs of individuals diagnosed with a brain tumour14 or those in the palliative care phase.46

Eight studies reported that 15.5%–67.9%17 ,19 ,22 ,23 ,28 ,30 ,31 ,35 of P&Cs experienced at least one unmet need. Kim et al23 reported the upper limit of 67.9% among the P&Cs of individuals with colorectal cancer 2 months after diagnosis. In this study, the overall prevalence of reporting unmet needs was found to decrease as time since diagnosis increased. Other studies used a higher cut-off for their reporting. For instance, Soothill et al19 found that 28% of P&Cs identified three or more significant unmet needs. Similarly, Girgis et al17 found that 36% of P&Cs reported three or more unmet needs at 6 months after diagnosis, with a significant decrease to 17.5% at 24 months.17 Ten or more unmet needs were reported by 17% of P&Cs in a study by Hodgkinson et al28 and by 10% of P&Cs in the Hwang et al32 study.

Prevalence of unmet needs by domain

Comprehensive cancer care unmet needs

The prevalence of unmet comprehensive cancer care needs ranged from 1.1% to 96% across 22 studies, with almost half reporting that up to a fifth of P&Cs report an unmet need in this domain.17 ,19 ,20 ,22 ,23 ,29,,32 Seven categories of unmet needs were identified and are shown in table 1. Fewer P&Cs of people with head and neck cancer29 or cancer survivors17 identified requiring help with these needs. Comprehensive cancer care needs were also mostly met in the Soothill et al19 ,20 studies among a mixed group of patients. Conversely, a higher prevalence of unmet needs was found in studies among the P&Cs of patients hospitalised,38 in the acute diagnostic and treatment phases33 ,40 and in the palliative care phase.36 ,41 ,46 ,47

Table 1

Summary of domains and prevalence ranges for subcategories of unmet needs

Emotional and psychological unmet needs

The prevalence of emotional and psychological unmet needs ranged from 3% to 93.2% across 15 studies, with most of these identifying that up to a third of P&Cs endorsed a given psychological unmet need item.16 ,17 ,19 ,20 ,22 ,23 ,28 ,30 ,31 Emotional and psychological unmet needs across studies are represented by the five categories shown in table 1. Across most of these unmet needs categories, a lower prevalence was reported by the P&Cs of cancer survivors.17 ,22 ,28 Needs related to dealing with negative feelings and getting emotional support for one's self remained largely unmet for the P&Cs of individuals in the palliative care phase.46 A high prevalence of unmet needs related to helping P&Cs provide emotional support to patients was reported among the wives of men with prostate cancer (53%–59%) in the acute phases of the illness33 ,40 and among a mixed group of P&Cs (50.7%).37 Of note, in the study by Kim et al23 across all three cohorts (2–60 months after diagnosis), psychological unmet needs were most prevalent.

Partner or caregiver impact and daily activities unmet needs

Across 22 studies, between 2.8% and 79% of P&Cs reported unmet partner or caregiver impact and/or daily activities needs. Half of these found that up to approximately a third of P&Cs identified unmet needs in this domain.16 ,17 ,19 ,20 ,22 ,23 ,28,,32 Common categories of unmet needs in this domain are listed in table 1. Overall, a lower prevalence of unmet needs was reported in studies involving the P&Cs of cancer survivors.17 ,22 ,23 ,28 A higher proportion of P&Cs of individuals in the palliative care phase endorsed unmet needs in this domain.36 ,46

Relationship unmet needs

Eleven studies found that between 3.7% and 58% of P&Cs reported unmet needs related to their relationships or interactions with the patient or others. Most studies found that up to a quarter of P&Cs reported relationship unmet needs.17 ,22 ,23 ,28 ,30 ,31 The most prominent areas of relationship unmet needs are shown in table 1. Overall, the pattern of prevalence was consistent with the other domains, where studies of the P&Cs of cancer survivors reported lower unmet needs17 ,22 ,28 ,30 than P&Cs in the palliative care phase.46

Information unmet needs

Twenty studies found 2.2%–86% of P&Cs experienced unmet information needs, with almost half of these stating that an unmet information need was reported by up to a third of P&Cs (table 1).14 ,16 ,17 ,19 ,20 ,23 ,29 ,32 Illness and medical unmet information needs were notably lower in studies that involved the P&Cs of cancer survivors.17 ,29 A high prevalence of information unmet needs was found among P&Cs in the early37 ,40 or palliative phase.46

Spirituality unmet needs

Spirituality needs were less often documented (2%–43%), but a common unmet need related to feeling there is hope for the future (table 1). Typically, studies identified that spiritual needs remained unmet for less than 20% of P&Cs.19 ,20 ,22 ,28 ,29 ,32 ,33 ,40

Variables associated with partners' and caregivers' unmet needs

Relationships between the level of unmet needs and demographics, psychosocial variables, illness variables and/or healthcare context are summarised in table 2.

Table 2

Summary of findings of common variables examined for their association with unmet needs

Caregiver demographics

Overall, findings concerning the impact of P&Cs' demographic characteristics on unmet needs remain equivocal. In several studies, age,19 ,22 ,30 ,31 ,33 ,37 ,44 ,47 sex,19 ,37 ,45 ,47 education22 33 ,37 ,40 ,47 and employment19 ,23 ,30 ,37 ,44 ,45 ,47 were not significantly associated with unmet needs. When a significant relationship was reported, the direction of that relationship varied across studies (see table 2).23 ,28 ,36 ,38 ,40 ,41 ,43,,45 Of note, five studies reported that caregivers who were not the patient's partner experienced higher unmet needs than those who were the partner.19 ,36 ,38 ,41 ,47

Patient or caregiver psychosocial variables

The relationships between elevated partner or caregiver distress,22 ,45 anxiety14 ,17 ,28 ,30 ,31 ,35 ,46 ,41 and/or depression14 ,17 ,28 ,30 ,35 ,46 and higher unmet needs were confirmed across several studies (see table 2). Two studies found a positive association between patients' and P&Cs' unmet needs.22 ,31 Additional variables associated with higher unmet needs included low social support,17 ,19 low relationship satisfaction,22 and having caring responsibilities.17 ,19 ,32 ,46 Of note, Kim et al23 and Girgis et al17 found that the associations between psychosocial variables and unmet needs varied across time.

Caregiver or patient health/illness variables

Numerous studies did not support a significant relationship between caregiver or patient health/illness variables and unmet needs (table 2).14 ,19 ,22 ,28 ,30 ,31 ,37 ,40 ,43 ,45 A few studies have suggested higher unmet needs among the P&Cs of individuals with advanced cancer.17 ,19 ,34 ,41

Healthcare context and care variables

Several studies have suggested a relationship between some cancer care variables and unmet needs (table 2).21 ,32 ,34 ,47 Examples include specialised versus acute cancer care units21 ,47 and source of caregiver information.34

Discussion

This systematic review summarised P&Cs' unmet needs across 29 studies and identified a subgroup of P&Cs at risk of reporting more unmet needs than their peers.

Prevalence of unmet needs – is experiencing one unmet need significant?

A popular indicator to gauge the overall prevalence of unmet needs was the proportion of P&Cs reporting at least one unmet need.17 ,19 ,22 ,23 ,28 ,30 ,31 ,35 It is difficult to determine the significance of experiencing one unmet need, particularly as to date there has been no attempt at quantifying the clinical significance of a given unmet need.48 However, as many of the top ranking unmet needs across studies pertained to coping with psychological or emotional distress,16 ,17 ,19 ,20 ,23 ,46 it is foreseeable that experiencing any of these, even if just one, could adversely impact clinical outcomes.

Unmet needs domains

Across the six unmet needs domains identified, it appeared that an unmet need item is typically endorsed by up to a third of P&Cs, with a higher proportion of P&Cs of individuals in the palliative care phases reporting unmet needs.36 ,41 ,46 ,47 Fewer P&Cs of cancer survivors identified unmet needs,17 ,22 ,23 ,28 a finding consistent with the Harrison et al6 review of patients' needs. Unmet needs in the comprehensive cancer domain reflected P&Cs' need to be aware of the resources available to them, to have a sympathetic relationship with healthcare professionals, and to be considered an integral part of the health team. Studies of emotional and psychological unmet needs highlighted P&Cs' need for assistance to manage their distress or negative feelings and reduce stress in the patient's life. Interestingly, in a recent study by Deshield et al,49 80% of caregivers reported they helped patients deal with their emotions and about a third reported this to be the most difficult aspect of caregiving. Issues of curtailing the impact of caregiving on lifestyle and schedule and performing illness management tasks were raised within the partner or caregiver impact and daily activities domain. In the relationship domain, communicating with the patient and sexuality and intimacy were prominent unmet needs. Several studies found many P&Cs do not get the information they needed, including information about what to expect and the illness and treatment. Overall, spirituality needs were assessed in fewer studies; however, when these were examined, ‘feeling hope for the future’ was a prominent unmet need.

Partners and caregivers at risk of experiencing unmet needs

As indicated in table 2, the subgroup of P&Cs at risk of experiencing more unmet needs included those who were female, were not the spouse or not living with the patient, had lower social support, and reported distress, anxiety and/or depression. Findings on gender and social support mirror those of studies on predictors of P&Cs' anxiety and depression.12 ,13 ,50 ,51 Of note, many of the studies reviewed corroborated that illness variables were not associated with level of unmet needs. However, these conclusions should be interpreted with caution, as many of these variables were reported in mostly cross-sectional studies. Hence, causality or lack thereof cannot be inferred from these data.

Implications for service delivery

The key categories of unmet needs identified in this review provide an evidence base to guide the design and implementation of supportive care services, particularly towards those unmet needs that are amenable to change and vulnerable partner/caregiver subgroups. For instance, top ranking psychological unmet needs across a number of studies included helping P&Cs manage their own emotional distress as well as the patients' distress, which in turn suggests the need for coping skills training interventions.

In addition, the findings of this review emphasised that a multidisciplinary approach to supportive care for P&Cs is needed. Although the concept of multidisciplinary care is well-recognised for oncology patients, it is questionable whether this model of care has reached P&Cs in the same way.19

However, beyond the issues of resource allocation, P&Cs often do not access services, even when these are available.52 Hence, this suggests that healthcare professionals might first have to reassure P&Cs that the support they provide to patients is considerable and their own needs are legitimate.

Methodological issues of studies reviewed

Response rates were not consistently reported and when they were, these were often low. Many authors acknowledged that clinicians hesitated to refer the P&Cs of acutely ill patients; this selection bias might have led to underestimating unmet needs. In addition, there is an under-inclusion of individuals from culturally and linguistically diverse (CALD) backgrounds. As few longitudinal studies were found,17 little is still known about how unmet needs change over time. Furthermore, studies gave little consideration to the location of treatment or availability of supportive care services, which in turn might have influenced the prevalence of unmet needs reported.

One of the main challenges of this review was the quantification of unmet needs to obtain a useful depiction of P&Cs' experiences, a challenge also raised by Harrison et al,6 and can be attributed to (1) the wide variety of unmet needs measures used and (2) differences in the definition and analysis of unmet needs. For some studies a need was labelled unmet only if it was both important and not satisfied. Even when comparable measures were used, some studies considered ‘moderate or high’ unmet needs,17 while others considered ‘low, moderate or high’ unmet needs.28 In addition, the timing of assessments varied dramatically and for some measures, items were confusing and elusive for reviewers, which raises concerns about how these were interpreted by P&Cs.

Future research implications

Longitudinal studies recruiting P&Cs in the early postdiagnosis phase to map their unmet needs as they confront different milestones along the illness trajectory are required. Moreover, studies among non-spouse, younger and/or CALD caregivers are also needed. Another research area under-explored is the influence of the availability of supportive care services on unmet needs. In addition, unmet needs measures require further analysis, including whether items are comprehensive and understood by P&Cs in the intended manner. Although a number of short measures are available to screen for patients' unmet needs (eg, SCNS-ST9),53 most caregiver measures included more than 30 items and comparable brief measures to facilitate integration in clinical practice are needed. This review also highlighted a demand for the development and evaluation of multidisciplinary interventions that could include both patients and their P&Cs, given the overlap and interdependence in unmet needs.

Limitations of the review

It is possible that some eligible papers were missed, as papers published in languages other than English or French and those that were not published in peer-reviewed journals were not included. Although our review was from 1995 and there is a 2-year gap between the 1993 review by Laizner et al26 and this one, it is unlikely that publications have been missed since most studies retrieved were published after 2000.

Conclusions

This systematic review identified 29 studies that quantified P&Cs' unmet needs. The prevalence of unmet needs across six domains varied greatly (range 1%–96%) and was particularly notable among the P&Cs of individuals in the palliative care phase. Furthermore, caregivers who were female, were not the spouse and not living with the patient, had lower social support, or reported higher distress seemed more at risk. Overall, findings can be used to provide an evidence base for decision-making regarding allocation of healthcare resources towards those subgroups more at risk of experiencing unmet needs and the content of psychosocial interventions to target the key areas of unmet needs.

References

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.

    Files in this Data Supplement:

    • Web Only 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.
    • Web Only 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.
    • Web Only 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.

Footnotes

  • Funding Dr Sylvie Lambert is supported by a National Health and Medical Research Council Training Fellowship, Dr Mariko Carey by a Hunter Medical Research Institute Post-doctoral Fellowship, Dr Billie Bonevski by a Cancer Institute New South Wales Career Development Fellowship and Professor Afaf Girgis by a Cancer Institute New South Wales grant.

  • Competing interests None.

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