Introduction

Currently, over 32 million people globally live with cancer [1] with incidence predicted to increase by 70% over the next two decades [2]. As health care systems struggle to cope with the mounting cancer burden and treatments have shifted from inpatient to outpatient modalities [3], care for cancer survivors is increasingly dependent upon relatives, friends and partners as informal caregivers (here on referred to as support persons) [4]. Support persons serve a crucial role in assisting patients during treatment, managing both disease and treatment-related side-effects, providing emotional and financial support, helping with activities of daily living and taking on additional family responsibilities [5, 6]. National survey data from the USA indicates that on average, support persons provide care for 8.3 h a day [6] addressing more than 50% of the care needs of cancer survivors [7]. In 2009, over 65 million people in the USA served as support persons for ill relatives [8], while in Australia, the number is estimated to be 2.7 million and rising [9].

Caring for a person with cancer places enormous financial, physical and emotional burden on those taking on the role [10,11,12,13,14] Cancer support persons report conflict and strain in their relationships [12], sleep deprivation and fatigue [15], diminished physical health [11, 12], heightened responsibility and fear [3], sexual dysfunction [16], and legal and financial hardship [17]. Estimations of the direct costs to support persons including hours spent having to take sick leave, holiday and unpaid leave as well as direct expenses, suggest an economic burden of $14,060US per year per cancer diagnosis [18]. An increasing worldwide incidence [19], lower survival rates [19] and aggressive treatment regimes that are often on-going [20] mean that survivors of haematological cancer are a unique and vulnerable population. The role of support persons for this population may be especially critical.

To enable health care providers to address the needs of support persons, it is important to understand and assess their unmet supportive care needs. Unmet needs have been defined as the difference between services or support necessary to deal with particular issues and the services or support actually received [21]. They include informational, physical, social, emotional, spiritual and practical needs [22]. Limited research has explored the supportive care needs of support persons of haematological cancer survivors [3], with some evidence suggesting that they may have greater or specific needs that differ from support persons for other cancer patient populations. For example, a recent Australian study of newly diagnosed cancer patients and their support persons found that carers of haematological cancer patients reported significantly greater unmet needs compared with support persons of people diagnosed with solid tumours [14].

For cancer survivors (and their support persons) living in rural areas, there are additional burdens of greater distances to treatment centres, less access to health care providers and support groups, the need for relocation and associated financial costs [23]. To our knowledge, no studies have compared the unmet needs of support persons living in rural areas with those in urban areas. Further, no studies have investigated the unmet needs of support persons of haematological cancer survivors living in rural areas.

The Institute of Medicine (IOM) and other international cancer organisations advocate that formal health care providers have a responsibility to prepare support persons for their role and help them to manage their own well-being [24, 25]. It is imperative then that the specific unmet needs of support persons of all types of haematological cancer survivors are identified and understood, including those who are most vulnerable, so that appropriate and targeted interventions are provided for support persons to address their needs. This exploratory study thus aimed to compare support persons of haematological cancer survivors living in rural and urban areas in regard to the type, prevalence and factors associated with reporting unmet needs and identify services perceived as helpful in reducing impact of unmet needs.

Methods

Study design

A cross-sectional, population-based sample of adult support persons were recruited as part of a larger, national study on unmet needs and psychological well-being of haematological cancer survivors and their support persons in Australia.

Participants

Study participants were adult support persons nominated by each haematological cancer survivor and defined as ‘someone who has helped you the most during your cancer journey.’ Cancer survivors were adults aged between 18 and 80 years and diagnosed with an ICD-10 or ICD-0-3 (M) defined haematological cancer (including leukaemia, non-Hodgkin lymphoma, myeloma and other blood cancers) recruited from five Australian state population-based cancer registries. There are eight state and territory cancer registries across Australia that serve a vital role in monitoring cancer incidence, mortality and linkage with patterns of care [26].

Procedure

Eligible cancer survivors were identified and recruited from each registry according to their specified registry recruitment procedures as described elsewhere [27]. All recruited cancer survivors were mailed a study package containing an invitation letter from the registry, a pamphlet explaining the purpose of the cancer registry, a study information statement, survey and reply-paid envelope. A second survey and reply-paid envelope were also included for survivors to pass on to their nominated support person. Non-responding survivors were sent a reminder letter and second study package after 3 weeks, with a follow-up telephone reminder at 6 weeks.

Measures

Socio-demographic-related items

The following demographic data were collected from support persons: gender, age, residential postcode, marital status, nationality, indigenous status, education, employment status, relationship to survivor and if living with the cancer survivor. Support persons were also asked about their own, if any, cancer history and non-cancer-related health problems. Residential postcodes were used to classify support persons as ‘rural’ or ‘urban’ based on the Accessibility and Remoteness Index of Australia (ARIA+) [28].

Support Person Unmet Needs Survey

The 78-item Support Person Unmet Needs Survey (SPUNS) assesses the level of unmet need experienced by support persons over the past month across six domains: Information and Relationship Needs (27 items), Work and Financial Needs (8 items), Needs for Access and Continuity of Care (9 items), Personal Needs (14 items), Emotional Needs (16 items) and Needs relating to the Future (4 items). Each item is scored from zero to four, with zero representing ‘no unmet need’ and four representing a ‘very high unmet need’. A total unmet needs score was calculated by summing responses to the 78-item SPUNS. A total SPUNS score of zero corresponded with reporting no unmet needs. Six domain scores were calculated by summing responses for domain-specific items and dividing by the number of non-missing responses for that domain. The SPUNS has demonstrated satisfactory levels of reliability and construct validity among cancer survivor support persons [29].

Psychological well-being

The 21-item Depression, Anxiety and Stress Scale (DASS 21) was used to measure self-reported anxiety (e.g. I found it hard to wind down), depression (e.g. I felt I had nothing to look forward to) and stress (e.g. I experienced breathing difficulty) over the past week. Participants responded to each item on a 4-point scale from zero to three, with ‘zero’ representing ‘did not apply to me at all’ and ‘three’ representing a ‘applied to me very much, or most of the time’. Three (3) sub-scales are assessed with seven items and domain scores are calculated by summing all subscale items and multiplying by two. DASS-21 cut-off scores of ‘10 or higher’ for depression, ‘8 or higher’ for anxiety and ‘15 or higher’ for stress were used to indicate clinical levels of symptoms [30]. Scores were only calculated for participants who responded to six or more items in each DASS-21 sub-scale. The DASS 21 has demonstrated acceptable levels of reliability and validity [31].

Additional adverse impacts

Seven items relating to other adverse impacts associated with providing care for a cancer survivor were asked of participants. These included Have you had to relocate as a result of the person you support having blood cancer? (Yes, temporarily; Yes, permanently; No). Those who responded ‘Yes, temporarily’ were asked to indicate how much time they spent in temporary accommodation. Those who responded ‘Yes, permanently’ were asked: Where did you relocate to? (Different place in the same town or city; Another town or city; Inter-state: Over-seas). All participants were asked about time spent travelling to access treatment services (5 response options); whether they had experienced any of seven different financial impacts (e.g. ‘Had less income’, ‘Had difficulty paying bills’); and preferred services perceived as helpful in reducing financial impacts.

Data analysis

Each categorical variable was summarised using frequencies and percentages with comparisons between urban and rural groups using chi-square/Fisher’s exact tests. Continuous variables were summarised using means, standard deviations, medians, min/max and interquartile ranges. Comparisons between urban and rural were conducted using t tests or Kruskal-Wallis tests. The mean, standard deviation, median and first and third quartile for each of the six SPUNS domains and total SPUNS scores were calculated. A conservative method of only utilising data from participants who completed more than 70% of items was adopted [32]. The percentage of support persons who reported a ‘high/very high’ level of unmet need was calculated for each of the 78 SPUNS items. Multiple unmet needs were defined as six or more ‘high/very high’ unmet needs, a similar cut-point used in previous research [16]. Multiple logistic regression analyses were conducted to identify factors associated with reporting multiple ‘high/very high’ unmet needs. All sociodemographic variables, the three DASS subscales and all additional adverse impact variables were included in logistic regression analyses. Variables with a p value of ≤ 0.2 on univariate analyses were included in the final multiple regression analyses. Complete case analysis was conducted. Statistical analyses were performed using SAS v9.4.

Results

Study sample

Of 4299 eligible haematological cancer survivors invited to participate in the larger national study, 1511 (35%) returned a completed survey. Of these, 1058 (92%) indicated they had planned to pass a survey on to their support person with 1004 (95%) support persons (792 urban and 193 rural) returning a completed survey. The majority of support persons were female (67%), aged over 60 years (56%) and the spouse or partner (84%) of a haematological cancer survivor. Rural support persons were less likely to have post-secondary school education (p = 0.006) and more likely to be have been born in Australia (p = 0.001) compared to urban participants (see Table 1). Twelve percent of support persons (n = 118) reported having a previous diagnosis of cancer themselves with the most common being breast (26%) and 49% (n = 487) indicated they experienced other non-cancer health-related problems including arthritis (74%) and vision or hearing difficulties (545). Rates of reported health problems did not differ significantly between urban and rural support persons.

Table 1 Demographic characteristics of rural and urban support persons of haematological cancer survivors
Full size table

Prevalence of unmet needs

Overall 18% (n = 172) of support persons expressed no unmet needs with 82% (n = 802) reported having at least one unmet need. Two-thirds (66%) of support persons had at least one ‘moderate, high or very high’ unmet supportive care need with a median number of 5. Twenty-four percent (n = 182) of participants identified having 6 or more ‘high/very high’ unmet needs. Of 1004 participants, 996 (99%) completed more than 70% of SPUNS items with the mean total unmet needs score for participants (n = 996) found to be 47 (SD = 56) with a median of 26.0 (Q1 = 4.0; Q3 = 72). There were no significant differences between rural and urban support persons in regard to the prevalence of multiple unmet needs, median number of moderate/high/very high needs, mean total unmet needs scores and each of the six mean domain scores (as shown in Table 2). However, a significantly higher proportion of rural support persons (76%) had at least one ‘moderate/high/very high’ unmet need compared with urban support persons (64%, χ2 = 8.72, p = 0.003).

Table 2 Total and mean domain unmet needs scores of urban and rural support persons
Full size table

Top ten ‘high/very high’ unmet needs

All four items in Needs relating to the Future were ranked in the top ten most frequently reported ‘high/very high’ unmet needs (see Table 3) with three of these—‘Dealing with worrying about the future of the person I support’ (19%), ‘Dealing with not knowing what lies in the future’ (19%) and ‘Dealing with worry about the cancer getting worse’ (14%) ranked one, two and three, respectively. Five items from the Personal Needs domain were also listed in the top ten unmet needs. Rural and urban participants differed in regard to only one of the top ten needs—‘Finding information about what type of financial help is available and how to obtain it’ with significantly more rural support persons (20%) reporting this as a concern compared to their urban (13%, p = 0.008) counterparts.

Table 3 Top ten most frequently reported ‘high or very high’ unmet needs in urban and rural support persons
Full size table

Factors associated with multiple ‘high/very high’ unmet needs

All of the variables tested in univariate analyses were found to have p values of 0.2 or less and hence were subsequently included in the multivariate analysis with the exception of two variables—‘number of health conditions’ and ‘had to sell an asset’. Results of multiple logistic regression analysis found that support persons who had ‘Difficulty paying bills’ (AOR = 3.02, 95%CI 1.42:6.45, p = 0.0042), scored ≥ 10 on the DASS-21 depression subscale (AOR = 5.91, 95%CI 3.37:10.36, p < 0.0001) and scored ≥ 14 on the DASS-21 stress subscale (AOR = 4.71, 95%CI 2.64:8.40, p < 0.0001) had higher odds of reporting multiple ‘high/very high’ unmet needs. Living in a rural area and other sociodemographic variables were not significant predictors of multiple high unmet needs (see Table 4).

Table 4 Multiple logistic regression analysis of variables associated with support persons reporting multiple ‘high or very high’ unmet needs
Full size table

Additional adverse impacts

A significantly higher proportion of participants living in rural areas were forced to relocate temporarily in comparison with those living in urban areas (35 vs 8%, p < 0.0001). For the majority of these (89%), relocation was to a different town or city. In contrast, 25% of urban participants forced to relocate did so within the same city versus only 1.5% of rural support persons (p < 0.0001). The mean and median length of time spent in temporary accommodation for urban support persons was 125 and 77 days, respectively, with no significant differences between rural and urban. Results indicated large differences in travel time from place of living to cancer treatment centres. Forty percent of rural participants had to travel between 2 and 5 h compared with only 5% of urban support persons, while travel time for the majority of those in urban areas (81%) was less than 1 h compared with 35% in rural areas (p < 0.0001).

When asked about financial impacts that occurred as a result of supporting the person with haematological cancer, significantly more participants from rural areas reported that they had to take time off work (52 vs 43%, p = 0.023), had less income (37 vs 27%, p = 0.009), had difficulty paying bills (22 vs 13%, p = 0.002), had trouble meeting day-to-day expenses (15 vs 8%, p = 0.011) and were forced to use up their savings (28 vs 16%, p = 0.0003) (see Table 5).

Table 5 Additional adverse impacts associated with being a support person for a haematological cancer survivor in rural and urban areas
Full size table

A higher proportion of rural support persons (54%) compared with urban (20%), indicated that getting treatment close to home would have been helpful in reducing the financial impact they experienced (p < 0.0001). Similarly, significantly more rural participants (26%) compared with urban (16%, p = 0.0013) indicated that access to free transport and financial assistance (31 vs 18%, p < 0.0001) would have been helpful (see Table 6).

Table 6 Services identified as being helpful in reducing financial impact of supporting a person with haematological cancer
Full size table

Discussion

This study importantly builds onto the sparse literature on the unmet needs of support persons of haematological cancer survivors and is the first nation-wide study to measure unmet needs in this population. Further, it is the first study to directly compare the unmet needs of support persons in rural and urban areas in Australia. Findings indicate that 66% reported at least one ‘moderate to very high’ unmet supportive care need. A quarter of the sample (24%) reported six or more ‘high/very high’ unmet needs. These figures are similar to those reported in previous studies [16, 33, 34] but higher compared to rates of 15 and 14% reported by Heckel et al. [14] and Soothill et al. [35], respectively. The disparity is likely due to a higher cut-point (i.e. 10 or more) used to define multiple unmet needs and differing study populations. Our results appear to give to weight to previous findings [14, 34] that supporting a person diagnosed with haematological cancer correlates with a high level of unmet needs. Given the importance of support persons for the functioning and well-being of survivors, failure to address these needs is likely to have implications for health care costs and productivity.

The most commonly reported unmet needs in the present study were in the domains of information, personal and emotional, and needs relating to the future. With regard to the later, concerns about what the future holds for the cancer survivor and managing those concerns were the most frequently reported needs by all support persons. This finding is consistent with previous research that has identified managing worries about the future and the possibility of cancer returning as overriding concerns for support persons [16, 33, 34]. A number of other unmet needs found to be frequently reported by participants in past studies, including accessibility of car parking [33], the impact of caring on the support person’s life [16] and reducing stress in the cancer survivor’s life [16, 34] were supported by the findings here also. Understanding the specific needs experienced by support persons of haematological cancer survivors can assist with offering specific services to address these needs, such as counselling around concerns for the future.

No differences were found between rural and urban support persons in regard to the proportions who had some and/or multiple high unmet needs though a significantly higher proportion of rural (76%) compared with urban (64%) had at least one moderate or higher level of unmet need. There were some significant differences between rural and urban participants in regard to specific unmet needs, particularly with regard to financial needs. Support persons from rural areas were more likely to report trouble with finding information about financial assistance, obtaining such support and being able to meet costs. As this is the first study to investigate the unmet needs of rural and urban support persons, there are no previous findings with which to compare our results. Not surprising though, our results appear to mirror the findings of Paul et al.’s study [36] on haematological cancer survivors in non-metropolitan areas in one Australian state who were also found to be more likely to report financial problems in comparison to their urban counterparts.

While living in a rural area was found to not be significantly associated with having multiple unmet needs in the present study, having financial impacts was. This suggests that differences in unmet needs of support persons are not necessarily based on the ‘rurality’ of where support persons reside, but rather related to financial stability. This is likely linked with the greater financial burden imposed by further distances to travel for treatment and relocation costs. One in three support persons from rural areas were forced to relocate temporarily to a different town while the cancer survivor underwent treatment and 40% reported extensive travel times (between 2 and 5 h). Significantly higher proportions of rural support persons also reported financial impacts including having to take time off work, having less income and using up savings. These findings again appear to mirror the problems experienced by cancer survivors living in rural areas as found by Paul et al. [36]. They suggest that cancer treatment and support services should target rural cancer survivors by proactively providing information about available travel subsidies and financial assistance [36]. We recommend this be extended to support persons in rural areas also with rural participants in this study indicating that financial assistance, free transport and treatment services closer to home would be beneficial in relieving the financial burden.

There are some limitations to the present study. First, the low response rate by cancer survivors (35%) may limit generalisability of our findings though this rate is comparable with other studies on haematological cancer survivors [14]. Additionally, the sociodemographic characteristics of our national sample of support persons closely correspond with participants in previous studies [14]. Second, due to the cross-sectional study design, it was not possible to explore causality nor how unmet needs may change over time.

Despite these limitations, the present study advances previous literature in this field. While past studies have used small samples of one or more haematological cancer sub-types, we have utilised a large, population-based sample inclusive of several sub-types of haematological cancers. Measurement of unmet needs was via a standardised, validated tool (the SPUNS) that addressed unmet needs across six domains. This was the first study to compare the unmet needs of rural and urban support persons in Australia and makes a substantial contribution to the paucity of research on the unmet needs of support persons of haematological cancer survivors.

Conclusion

Support persons of haematological cancer survivors appear to be greatly impacted by their role, reporting multiple unmet needs. Support persons living in rural areas are particularly vulnerable to financial-related unmet needs. Findings highlight the importance of greater efforts to ensure that health care providers and our models of health care service adequately address their needs. Future translational research should investigate the efficacy and adoption of strategies for addressing unmet needs of those people who are supporting survivors of haematological cancer with particular regard for strategies that address concerns about the future, and alleviate the financial burden imposed on those living in rural areas. Such strategies may include prioritising rural patients’ preferences regarding the timing of treatment and their access to financial support.