Introduction
Cancer remains one of Australia’s most significant public health challenges, with nearly 1 million new cancer diagnoses recorded over the past decade (Australian Institute of Health and Welfare [AIHW] 2025). Aroud 3% of Australian population were living within five years of a cancer diagnosis in 2022 (World Health Organization [WHO] 2022). National expenditure on cancer care exceeded $18 billion between 2022 and 2023 in Australia (Queensland Government 2025a), reflecting expanding demand for treatment advances and longer duration of survival. Although the cancer survival rate has increased to 72% due to early cancer screening programs and treatment, 95% of the increase in cancer incidence that occurred between 2000 and 2025 was largely attributed to population ageing and growth (AIHW 2025). Queensland, a state located in the northeast of Australia, experiences a particularly high cancer incidence rate, approximately 13% above the national average, due to increased smoking rate, obesity, and prolonged sun exposure (Queensland Government 2025a; 2025b). Despite substantial government investment in prevention initiatives, early detection, treatment infrastructure, and supportive cancer services, cancer treatment outcomes remain poorer in regional, rural, and remote (RRR) areas, with individuals’ diagnoses occurring at an advanced stage, where access to specialized services is limited (Queensland Government 2025c).
Family members/friends of the individual with cancer who volunteer to provide care for their loved ones are often referred to as unpaid cancer caregivers (UCCs) (Papadakos et al. Reference Papadakos, Samoil and Umakanthan2022). UCCs play an increasingly essential role as cancer care shifts from inpatient facilities to outpatient and home-based care (Kent et al. Reference Kent, Lee and Asad2023; Dave et al. Reference Dave, Friedman and Miller-Sonet2024; Furtado et al. Reference Furtado, Davis and Groarke2024). UCCs undertake a wide range of responsibilities across the cancer trajectory, including assisting with treatment decisions, medication management, nutrition, mobility, and toileting during early stages, to symptom management and end-of-life care in advanced stages, as well as providing continuous emotional support (Li et al. Reference Li, Sun and Zhang2022; Papadakos et al. Reference Papadakos, Samoil and Umakanthan2022; Dave et al. Reference Dave, Friedman and Miller-Sonet2024). The cancer caregiving experience for UCCs living in RRR areas is considerably exacerbated by long travel distances, reduced availability of oncology and supportive services, and potential financial burden associated with treatment-related expenses (Hallgren et al. Reference Hallgren, Hastert and Carnahan2020). Importantly, many UCCs assume demanding responsibilities without any previous experience of caregiving or education (Mollica et al. Reference Mollica, Litzelman and Rowland2017; Bining et al. Reference Bining, Wasserman and Brahim2022), increasing their risk of anxiety, depression, emotional distress, and social isolation (Darley et al. Reference Darley, Coughlan and Furlong2021; Li et al. Reference Li, Sun and Zhang2022).
A growing body of research shows that emotional distress, including anxiety, depression, loneliness, stress, and burden arising from caregiving responsibilities, adversely affects UCCs’ well-being (Trevino et al. Reference Trevino, Prigerson and Maciejewski2018; Wang et al. Reference Wang, Chung and Tan2018; Zhang et al. Reference Zhang, Zhang and Liu2023). Previous research in various contexts suggested that negative experiences of cancer caregiving are associated with unmet supportive care needs throughout the cancer trajectory (Denham et al. Reference Denham, Wynne and Baker2020; Wang Reference Wang2021; Bining et al. Reference Bining, Wasserman and Brahim2022). Unmet needs are considered perceived needs for additional support that UCCs have not adequately received (Yarosh et al. Reference Yarosh, Nichols and Hirschey2025). Although unmet needs may be defined differently across contexts, a systematic review conducted by Wang et al. (Reference Wang, Chung and Tan2018) identified unmet needs among UCCs across several domains, including physical, emotional, spiritual, healthcare service access, informational, social, and practical needs. Several studies have explored how multidisciplinary healthcare staff can assist UCCs in addressing their needs once unmet needs have been identified (Wang et al. Reference Wang, Chung and Tan2018; Ringborg et al. Reference Ringborg, Schandl and Wengström2022; Langegård et al. Reference Langegård, Cajander and Ahmad2023), there is limited understanding of how unmet needs are experienced by UCCs residing in RRR settings, where geographic isolation, cultural context, and restricted service availability create unique challenges (Too et al. Reference Too, Lelei and Adam2023).
Peer support has emerged as a promising and flexible model for addressing informational, practical, and emotional needs among UCCs (Benson et al. Reference Benson, Oliver and Washington2020). Peers, defined as part of the social support of individuals with lived experience of cancer caregiving, including family members, volunteers, and other UCCs (Amonoo et al. Reference Amonoo, Guo and Keane2024; Sezgin and Bektas Reference Sezgin and Bektas2025), provide empathy, shared understanding, and contextualized guidance that can complement healthcare services (Ziegler et al. Reference Ziegler, Hill and Lieske2022). Peer support can be delivered through group programs, one-on-one interactions, online platforms, telephone calls, and face-to-face meetings (Jablotschkin et al. Reference Jablotschkin, Binkowski and Markovits Hoopii2022). Such flexibility is particularly valuable in RRR areas, where caregivers often face long travel distances, limited access to specialized care, and substantial time constraints due to intensive caregiving responsibilities (Andersen et al. Reference Andersen, Nielsen and Danbjørg2019; Yanai et al. Reference Yanai, Makihara and Matsunaga2022). Evidence suggests that peer support can reduce isolation and enhance coping capacity (Gunn et al. Reference Gunn, Spronk and Fletcher2022; Winter et al. Reference Winter, McKenzie and Spence2023). Furthermore, the flexibility offered by peer support can give UCCs a sense of belonging to a community, leading to feeling less isolated and emotionally distressed (Myrhøj et al. Reference Myrhøj, Nielsen and Visler2025). While peer support is considered a valuable source of support, the impact and outcomes of such support in improving UCCs’ well-being may vary depending on factors such as place of residence, common language, culture, cancer stage, UCCs’ mental health status, and the extent of compatibility between UCCs and their peers (Shi et al. Reference Shi, Chen and Su2019; Breuning et al. Reference Breuning, Mählmann and Kerek-Bodden2024). This study aims to examine the unmet supportive care needs of UCCs living in RRR Queensland and to explore how perceived peer support is associated with caregiver distress and burden and address this gap in a cross-sectional study by employing several instruments.
Method
Study design and sample size
This quantitative cross-sectional study was conducted at a single point in time using an online survey to identify the relationship between unmet supportive care needs, perceived peer support, and well-being among UCCs living in RRR areas of Queensland, Australia. The sample size was estimated a priori for structural equation modeling (SEM) using the a priori sample size calculator for structural equation models. The calculation was based on an anticipated effect size of 0.30, desired statistical power of 0.90, probability level of 0.05, 7 latent variables, and 32 observed variables/items, as reflected in the proposed SEM. This yielded a recommended minimum sample size of 210 participants. A total of 211 participants completed the survey. Responses were screened for duplicate entries and inconsistencies, and 202 eligible participants were retained for the final analysis.
Setting and timeline
The survey was conducted among UCCs living in RRR areas of Queensland, Australia, which are defined by the Modified Monash Model (MMM), an Australian geographical classification used to identify metropolitan, rural, remote, or very remote areas in Australia by assigning residential postcodes in each state (Australian Government 2025). The MMM classification includes 7 categories. MMM1 refers to metropolitan cities, which include approximately 70% of Australia’s population. MMM2 includes regional areas located within 20 km of major towns with populations of more than 50,000. Rural areas (MMM3–MMM5) are defined as towns with populations between 15,000 and 50,000 and at least 10 km from a regional town. Remote areas are defined as MMM6–MMM7 and include locations with populations of less than 1,000 that are separated from the mainland (Australian Government 2025). The MMM category is essential for identifying shortages of healthcare professionals, guiding government funding for healthcare services, and ensuring that medical resources are accessible across RRR areas (Australian Government 2025). The MMM map is shown in Figure 1.
Modified Monash Model (MMM) 2023 classification map of Australia, with Queensland highlighted as the study setting. Adapted from the Australian Government (2025) Department of Health, Disability and Ageing.
Participants’ eligibility criteria and recruitment
Participants were invited to take part in the survey if they met the following eligibility criteria:
1. Providing unpaid care to a family member who has been diagnosed with advanced cancer, including stage III and IV cancer (Deshmukh et al. Reference Deshmukh, Dhar, Thomas, Badwe, Gupta, Shrikhande and Laskar2024);
2. UCCs must be adults aged over 18 years old;
3. The ability to write, read, and communicate in English;
4. Providing care for an individual with advanced cancer for at least a minimum duration of 2 months to ensure that UCCs had caregiving experience;
5. Living in an RRR area of Queensland, Australia (MMM2–MMM7).
Recruitment strategies included displaying study flyers in the outpatient clinic of a regional-based hospital, a major public hospital which is often visited by UCCs and their loved ones for treatment. Furthermore, social platforms, including community pages or groups, were used to promote the flyer. Cancer organizations were contacted to support the promotion of the flyer through community communication channels. As a substantial part of the recruitment process was conducted on the social media platforms, the sample size may overrepresent participants who were more digitally literate, educated, or already connected with cancer organizations.
Data collection
Data collection occurred from October 2025 to December 2025. The survey was administered online via the UniSQ Survey Tool. Participants accessed the survey by scanning a QR code or clicking the link provided in the study flyer. Prior to commencing the survey, participants were asked to read the participant information sheet and consent form, both embedded within the UniSQ Survey Tool platform, to ensure they understood the aims of the study. After providing informed consent, participants proceeded to the survey, which took approximately 20 minutes to complete. Before accessing the main survey, participants were asked to confirm the eligibility criteria and then start the survey. Following data collection, responses were assessed for duplicate entries and invalid responses; 202 participants were included in the final analysis.
Survey instruments
After refining and assessing the reliability of the tools, unmet needs of UCCs living in RRR areas, and how the availability of peer support impacts their well-being, the following tools were used:
• Comprehensive Needs Assessment Tool for Cancer Caregivers (CNAT-C);
• Multidimensional Scale of Perceived Social Support (MSPSS);
• Depression Anxiety and Stress Scale (DASS-21);
• Caregiver’s Burden Scale in End-of-Life Care modified to Caregivers’ Distress Scale (CDS); and
• Spiritual Needs Questionnaire (SpNQ).
MSPSS was used as a tool to assess perception of UCCs for availability of peer support from their family, friends, and other resources. To improve the content coverage of the Religious and Spiritual Support (RSS) subscale and capture broader spiritual and religious support needs among UCCs living in RRR areas, the CNAT-C RSS subscale was modified by incorporating3 items from the SpNQ instrument (items 3, 21, and 23). Preliminary internal consistency for RSS subscale in the pilot study was strong (Cronbach’s α = 0.85). In addition, 2 CNAT-C information items (I4 and I7) were removed due to conceptual redundancy and overlap with HFS5. Such an adaptation modified the structure of the original CNAT-C; therefore, subscale findings were interpreted cautiously and were not considered directly comparable with studies using the original CNAT-C instrument. Thus, a modified 42-item version of the CNAT-C was used, and the original 41-item CNAT-C was adapted by removing 2 items and incorporating 3 SpNQ items into the RSS subscale. As a result, SpNQ was not analyzed as a separate instrument. Prior to the main quantitative survey, the instrument set was pilot tested among 16 UCCs residing in RRR areas of Queensland to assess feasibility, clarity, acceptability, and preliminary internal consistency. Corrected item-total correlations and Cronbach’s alpha were examined, and minor rewording was made to improve clarity and conceptual alignment before the main survey. Although all DASS-21 items remained unchanged during the pilot study and were incorporated in the survey, due to a data management error, 1 DASS-21 item (D7) was missed in the final quantitative dataset. Therefore, the DASS-related analyses were based on the remaining 20 available items and were interpreted as a DASS-21-derived overall distress score rather than the standard 21-item DASS-21 total score. The detailed summary of instruments is indicated in Table 1.
Description, scoring, and psychometric properties of instruments used in survey
Table 1 Long description
Five instruments are listed with columns for instrument name, number of items used in the study and post pilot study, scoring, psychometric properties, and reference. MSPSS has 12 items, uses a 7 point Likert scale from 1 very strongly disagree to 7 very strongly agree, and reports Cronbach alpha 0.90. CNAT C excluding RSS items has 37 items, uses a 4 point Likert scale from 0 no need to 3 high need, and reports Cronbach alpha 0.81. DASS 21 has 21 items, uses a 4 point Likert scale from 0 not at all to 3 most of the time, and reports Cronbach alpha 0.88. CBS EOLC modified to CDS has 9 items pre pilot and 10 items post pilot after adding one item, uses a 4 point Likert scale from 0 never to 3 very often, and reports Cronbach alpha 0.89. Adapted RSS and SpNQ items has 5 items, with RSS1 and RSS2 from CNAT C and RSS3 to RSS5 from SpNQ items 3, 21 and 23, uses a 4 point Likert scale from 0 no need to 3 high need, and reports Cronbach alpha 0.85.
MSPSS = Multidimensional Scale of Perceived Social Support, CNAT-C = Comprehensive Needs Assessment Tool for Cancer Caregivers, DASS = Depression Anxiety and Stress Scale, CBS-EOLC = Caregiver’s Burden Scale in End-of-Life Care, CDS = Caregivers’ Distress Scale, RSS = Religious and Spiritual Support, /SpNQ = Spiritual Needs Questionnaire.
Statistical analysis
After data collection, the data were de-identified and coded into the Statistical Package for the Social Sciences (SPSS) version 31.0 for data analysis. IBM SPSS AMOS was also used for SEM. The participants’ characteristics were summarized using frequencies. The scale reliability was examined using Cronbach’s alpha. Associations between key variables were assessed using Pearson correlations. Also, multiple linear regression was performed to assess independent variables. SEM was conducted in AMOS to test the hypothesized relationships among CNAT-C, MSPSS, DASS-20, and CDS.
Results
Participants’ characteristics
A total of 202 UCCs participated in the survey. The results showed that most UCCs were aged 30–39 years (39.1%) and between 18 and 29 years (34.7%); 54.5% were male. Participants primarily resided in rural (42.1%), regional (36.1%), and remote (21.8%) areas. Over half of the participants were married (65.8%) and were primarily caring for either a spouse/partner (29.2%) or a parent (28.7%). Approximately one-third of UCCs reported that the person they cared for had a cancer diagnosis of either 2–3 years (37.6%) or 3–5 years (32.7%), and just under half reported that they had been providing care for their loved ones for 2–3 years (41.1%). The most frequent cancer type was breast cancer (21.3%) and lung and upper respiratory tract cancer (such as laryngeal cancer) (17.3%). Most participants held either a bachelor’s degree (39.6%) or a certificate/diploma (32.7%). Most reported being employed either full-time (39.1%) or part-time (39.1%). The participant characteristics are summarized in Table 2, and the detailed description of participant characteristics is provided in Supplementary Material 1.
Participant characteristics (n = 202)
Table 2 Long description
Participant characteristics are reported as counts and percentages for age in years, gender, relationship to the person with cancer, marital status, cancer type, residency, and employment. Age groups were 18 years and under, 3 participants, 1.5 percent; 19 to 29, 70, 34.7 percent; 30 to 39, 79, 39.1 percent; 40 to 49, 41, 20.3 percent; 50 to 59, 2, 1.0 percent; and over 60, 7, 3.5 percent. Gender counts were female, 92, 45.5 percent. Relationship categories were spouse or partner, 59, 29.2 percent; parent, 58, 28.7 percent; adult child, 35, 17.3 percent; sibling, 21, 10.4 percent; friend, 15, 7.4 percent; and other relatives, 14, 6.9 percent. Marital status was single, 45, 22.3 percent; married, 133, 65.8 percent; widowed, 7, 3.5 percent; divorced, 10, 5.0 percent; separated, 6, 3.0 percent; and other, 1, 0.5 percent. Cancer type was breast, 43, 21.3 percent; lung and upper respiratory tract, 35, 17.3 percent; gi, 30, 14.9 percent; haematologic, 29, 14.4 percent; urological, 21, 10.4 percent; gynaecologic, 11, 5.4 percent; skin, 13, 6.4 percent; neuroendocrine, 7, 3.5 percent; and other, 13, 6.4 percent. Residency was regional, 73, 36.1 percent; rural, 85, 42.1 percent; and remote, 44, 21.8 percent. Employment was full time, 79, 39.1 percent; part time, 79, 39.1 percent; casual, 11, 5.4 percent; self employed, 20, 9.9 percent; unemployed, 11, 5.4 percent; retired, 1, 0.5 percent; and other, 1, 0.5 percent.
Descriptive statistics
While missing data were minimal, ranging between 98.5% and 100%, scale analyses were based on complete cases, and listwise deletion was applied for reliability, correlation, and regression analyses. The number of valid cases varied across analyses. The summary is demonstrated in Table 3.
Descriptive statistics of key study variables among unpaid cancer caregivers (n = 202)
Table 3 Long description
Descriptive statistics are reported for four scales, DASS minus 20, CDS, CNAT-C, and MSPSS, with columns for N, Minimum, Maximum, Mean, and Std. Dev. DASS minus 20 has N 202, minimum 0.25, maximum 2.95, mean 1.649, and standard deviation 0.593. CDS has N 202, minimum 1.08, maximum 3.92, mean 2.664, and standard deviation 0.6182. CNAT-C has N 202, minimum 0.56, maximum 2.81, mean 1.945, and standard deviation 0.481. MSPSS has N 202, minimum 1.17, maximum 7.00, mean 4.634, and standard deviation 1.298.
DASS = Depression Anxiety and Stress Scale, CDS = Caregivers’ Distress Scale, CNAT-C = Comprehensive Needs Assessment Tool for Cancer Caregivers, MSPSS = Multidimensional Scale of Perceived Social Support.
Descriptive analysis of UCCs’ unmet needs from CNAT-C subscales ranged from 1.85 (0.67) to 2.04 (0.61) (N = 202). The highest mean score was reported for practical support (PS), followed by hospital facilities and services and healthcare staff. Table 4 shows the complete descriptive analysis of the unmet needs of participants.
Descriptive statistics of CNAT-C subscales representing unmet supportive care needs among unpaid cancer caregivers (n = 202)
Table 4 Long description
Descriptive statistics summarize CNAT-C subscale scores for unmet supportive care needs, with columns for N, Min, Max, Mean, and Std. Dev. Each subscale has N 202, and all Max values are 3.00. HPP, Health and Psychological Problems, has Min 0.17, Mean 1.893, and Std. Dev. 0.645. FSS, Family or Social Support, has Min 0.00, Mean 1.882, and Std. Dev. 0.6341. HCS, Health Care Staff, has Min 0.00, Mean 1.983, and Std. Dev. 0.589; Information has Min 0.17, Mean 1.976, and Std. Dev. 0.608. RSS, Religious or Spiritual Support, has Min 0.00, Mean 1.851, and Std. Dev. 0.666; HFS, Hospital Facilities and Services, has Min 0.33, Mean 1.990, and Std. Dev. 0.592. PS, Practical Support, has Min 0.50, Mean 2.042, and Std. Dev. 0.613.
Reliability
Reliability analysis was performed after listwise deletion within each scale. Internal consistency was excellent across all scales: MSPSS (12 items; α = 0.902; n = 202), CDS (10 items; α = 0.806; n = 201), and CNAT-C (42 items; α = 0.924; n = 199). All items remained unchanged and were not removed. Despite missing one of the DASS-21 items due to a data management error, the reliability was excellent (20 items; α = 0.899; n = 200) with the remaining 20 items in the data analysis.
Bivariate correlations
Correlations were performed among participants with complete data (listwise deletion; N = 194). Pearson’s correlations indicated significant positive associations among the variables (N = 194, all p < 0.001). Similarly, DASS-20 Total was also positively correlated with the CNAT-C Total (r = 0.397), MSPSS Total (r = 0.393), and CDS Total (r = 0.555). Moreover, CNAT-C Total was positively correlated with the MSPSS Total (r = 0.321) and CDS Total (r = 0.495). MSPSS Total also indicated a positive correlation with CDS Total (r = 0.234). Table 5 shows correlations among unmet needs, perceived social support, distress, and caregiver burden.
Pearson correlations among unmet needs, perceived social support, distress, and caregiver burden (N = 194)
Table 5 Long description
Pearson correlation matrix compares DASS minus 20, CNAT-C, MSPSS, and CDS, with dashes on the diagonal. DASS minus 20 correlates with CNAT-C at 0.397 with double asterisks, with MSPSS at 0.393 with double asterisks, and with CDS at 0.555 with double asterisks. CNAT-C correlates with MSPSS at 0.321 with double asterisks and with CDS at 0.495 with double asterisks. MSPSS correlates with CDS at 0.234 with double asterisks. All listed off diagonal correlations are positive, ranging from 0.234 to 0.555, and each is marked with double asterisks.
DASS = Depression Anxiety and Stress Scale, CNAT-C = Comprehensive Needs Assessment Tool for Cancer Caregivers, MSPSS = Multidimensional Scale of Perceived Social Support, CDS = Caregivers’ Distress Scale.
Pearson’s r (2-tailed).
** p < 0.001.
Regression analysis
Two multiple linear regression analyses were conducted using the Enter method to examine whether unmet needs (CNAT-C) and perceived peer support (MSPSS) predicted well-being, including distress (DASS-20) and caregiver burden (CDS) (N = 194). The model predicting DASS-20 was significant, explaining 23.6% of the variance. Both unmet needs and perceived peer support were significant predictors of well-being. The model predicting caregiver distress (CDS) explained 25.1% of the variance; unmet needs were a significant predictor, whereas perceived peer support was not. Full regression results are presented in Table 6.
Multiple regression models predicting distress and caregiver burden among unpaid cancer caregivers (n = 194)
Table 6 Long description
Multiple regression results are reported for two outcomes, DASS 20 Total and CDS Total, with columns for unstandardized coefficient with standard error, standardized coefficient beta, and p value. For DASS 20 Total, the constant is 0.319 with standard error 0.179 and p value 0.076. For DASS 20 Total, CNAT-C Total has B 0.370 with standard error 0.082, beta 0.301, and p value less than 0.001; MSPSS Total has B 0.136 with standard error 0.031, beta 0.296, and p value less than 0.001. For CDS Total, the constant is 1.308 with standard error 0.186 and p value less than 0.001. For CDS Total, CNAT-C Total has B 0.603 with standard error 0.085, beta 0.468, and p value less than 0.001; MSPSS Total has B 0.041 with standard error 0.032, beta 0.084, and p value 0.203.
CNAT-C = Comprehensive Needs Assessment Tool for Cancer Caregivers, MSPSS = Multidimensional Scale of Perceived Social Support, CDS = Caregivers’ Distress Scale, B = unstandardized coefficient, SE = standard error, β = standardized coefficient. Predictors were entered simultaneously (Enter method). Two-tailed tests.
Model fit (DASS_Total): R 2 = 0.236; Adjusted R 2 = 0.228; F(2,191) = 29.489; p < 0.001.
Model fit (CDS_Total): R 2 = 0.251; Adjusted R 2 = 0.243; F(2,191) = 32.027; p < 0.001.
Structural equation modelling
The SEM was performed in SPSS AMOS (version 31) using maximum likelihood estimation (N = 199). Overall model fit was good, based on standard fit indices (χ 2 (445) = 577.353, p < 0.001; Chi-square Minimum Discrepancy divided by Degrees of Freedom (CMIN/DF) = 1.297; Comparative Fit Index (CFI) = 0.947; Tucker–Lewis Index (TLI) = 0.940; Incremental Fit Index (IFI) = 0.948; Root Mean Square Error of Approximation (RMSEA) = 0.039 [90% CI: 0.029–0.047], p-value for Close Fit (PCLOSE = 0.985). Bollen–Stine bootstrapping (1,000 samples) was also performed to support the overall fit (p = 0.428). The final SEM is presented in Figure 2.
Simplified structural equation model of relationships between unmet supportive care needs, perceived social support, psychological distress, and caregiver burden among unpaid cancer caregivers. This simplified figure is for presentation purposes only and does not represent a re-estimated model.
Figure 2 Long description
A structural equation model with various nodes and directional arrows indicating relationships. Key nodes include zA, zS, zD, DASS, CDS, MSPSS and CNATC. Each node is connected by arrows, showing the flow of influence or correlation. For example, zA is connected to nodes A1 through A7 and zS is connected to nodes S1 through S7. The node DASS is central, receiving inputs from zA, zS and zD and influencing other nodes like CDS and MSPSS. The node CDS is linked to CDSCS and CDSES, while MSPSS connects to MSPSS-SO, MSPSS-Fam and MSPSS-FH. CNATC is connected to CNATC-HPP, CNATC-FSS, CNATC-HCS, CNATC-RSS, CNATC-HSF and CNATC-PS. Arrows indicate the direction of relationships, with some nodes having multiple inputs and outputs, illustrating complex interactions within the model.
The final model showed that greater unmet needs (CNAT-C) significantly predicted distress (DASS-20) (β = 0.379, p < 0.001) and higher caregiver burden (CDS) (β = 0.551, p < 0.001), indicating that higher unmet needs are associated with higher distress and caregiver burden and, therefore, poor well-being. Moreover, MSPSS also significantly predicted DASS-20 (β = 0.229, p = 0.002), suggesting UCCs’ perceived peer support is linked to higher distress. However, there was no direct path from MSPSS to CDS, as the inclusion of a direct path did not improve the model fit and suggested overlap with DASS-20. Additionally, the model demonstrated moderate explanatory power, indicating 25.2% of the variance in DASS-20 (R 2 = 0.252) and 30.4% of the variance in CDS (R 2 = 0.304). Simplified conceptual path diagram showing the main relationships tested in the structural equation model. The detailed AMOS output, including indicators, error terms, and covariance paths, is provided in Supplementary Material 2.
Discussion
The present study found that UCCs living in underserved areas in Queensland face further unmet needs in PS, followed by needs related to hospital facilities/services, and healthcare staff. In addition, greater unmet supportive care needs were associated with higher levels of distress and caregiver burden, indicating poorer caregiver well-being.
The sociodemographic profile of participants in the current study varies from that reported in previous research. Studies conducted by Stiller et al. (Reference Stiller, Goodwin and Crawford-Williams2021) and Goodwin et al. (Reference Goodwin, Crawford-Williams and Ireland2022) in Australia found that UCCs who participated in paper-based or telephone surveys were more likely to be female, have low to middle levels of education, and be middle-aged. However, a recent mixed-methods study conducted by Yuen et al. (Reference Yuen, Goldsmith and Wilson2024) showed that UCCs who were invited to participate via email were predominantly female, highly educated, had easy access to the internet, lived in metropolitan areas, and had already been connected with cancer organizations in Australia. The findings from the current study may suggest that female caregivers living in RRR are less likely to have digital literacy and preoccupied with caregiving responsibilities. While previous Australian studies have reported that caregivers tend to be older (Stiller et al. Reference Stiller, Goodwin and Crawford-Williams2021; Gunn et al. Reference Gunn, Spronk and Fletcher2022), the findings of the present study contrast with this pattern, with a greater proportion of younger participants. However, these findings are consistent with more recent Australian studies (Yuen et al. Reference Yuen, Goldsmith and Wilson2024; Panahi et al. Reference Panahi, Romaniuk and Jahan2026), suggesting that as the older population grows, younger generations are likely to take on caregiving roles.
International research reported that the prominence of practical and service-related unmet needs likely reflects the increasing complexity of caregiving as cancer progresses to advanced stages (Hart et al. Reference Hart, Crawford-Williams and Crichton2022; Too et al. Reference Too, Lelei and Adam2023), requiring more PS. While previous findings showed that some practical caregiving tasks, such as shopping, can be shared among social networks to meet such needs in RRR areas in the absence of healthcare services, many practical tasks, such as wound care, accessing mobility equipment, and learning safe manual handling with equipment (Strohbuecker et al. Reference Strohbuecker, Gaertner and Stock2011), cannot be sufficiently addressed by peers without adequate instruction from healthcare staff. Therefore, healthcare staff and services are essential in RRR areas to ensure that practical needs are effectively met (Miranda et al. Reference Miranda, Musher and Badr2025), as well as to enable UCCs to visit or contact them in emergencies when support provided at home becomes unmanageable and requires hospital admission (Northouse et al. Reference Northouse, Mood and Schafenacker2013).
Our findings demonstrated that higher unmet needs of UCCs are associated with poor well-being, which is consistent with the findings conducted by Aziza et al. (Reference Aziza, Wang and Huang2018) and Lambert et al. (Reference Lambert, Hulbert-Williams and Belzile2018), which can be further intensified during the end stages of cancer (Yang et al. Reference Yang, Lee and Kuparasundram2021). The overall poor well-being of UCCs can be associated with coping with the imminent death of their loved ones, managing potential financial challenges, and the need to provide care for their loved ones for longer hours, with limited high-quality local services (Gunn et al. Reference Gunn, Spronk and Fletcher2022; Johnston et al. Reference Johnston, Craig and Stiller2023). Additionally, cultural and social norms – such as UCCs’ perception as being self-reliant – among those living in RRR areas may contribute to reluctance to seek support from healthcare staff, even when services are available, leaving their needs unmet (Ugalde et al. Reference Ugalde, Blaschke and Boltong2019; Vayro et al. Reference Vayro, Brownlow and Ireland2020, Reference Vayro, Brownlow and Ireland2021).
The findings from this study indicate that higher perceived peer support is associated with lower well-being and higher unmet needs. The findings are similar to Xu et al. (Reference Xu, Weng and Chen2024), who found that UCCs providing care for individuals with a near-death stage experience more anxiety, depression, and stress; hence, support from other social networks increases to alleviate emotional distress arising from anticipated bereavement. Therefore, poor well-being is more likely to be linked to a poor cancer prognosis, impending death of individuals with cancer, and unmet needs during the advanced stages of cancer, while peers attempt to enhance the support they provide to UCCs (Choi et al. Reference Choi, Vatovec and Reblin2025).
Moreover, despite positive perceptions of the availability of peer support in addressing UCCs’ unmet needs and improving well-being in RRR areas (Henning-Smith et al. Reference Henning-Smith, Moscovice and Kozhimannil2019; Chua et al. Reference Chua, Nguyen and Puah2022), many studies reported that the perception of availability of peer support does not necessarily address UCCs’ needs and enhance their well-being (Kirby et al. Reference Kirby, Barlow and Saurman2016; Stanek et al. Reference Stanek, Olsavsky and Hill2024; Breivik et al. Reference Breivik, Ervik and Kitzmüller2025). Similarly, a qualitative study reported that although UCCs in RRR areas perceived peer support, some UCCs experienced emotional distress after hearing peers share negative cancer caregiving experiences (Perera et al. Reference Perera, O’Callaghan and Ugalde2021).
In addition, findings from recent studies suggest that as perceived support increases, family conflict between peers and UCCs may arise, including difficulties in fairly sharing caregiving responsibilities, differing expectations about the caregiving role, disagreements regarding end-of-life decisions, and unwanted help that increases emotional distress rather than relieving, resulting in unmet needs and poor well-being (Benson et al. Reference Benson, Parker Oliver and Demiris2019, Reference Benson, Washington and Landon2023; Wright et al. Reference Wright, Bylund and Bagautdinova2025). Thus, not only the availability and perception of peer support but also the quality of the support received by UCCs should be taken into consideration, where support is often provided by peers rather than healthcare services in RRR areas.
Thus, this cross-sectional study highlights that UCCs reported the highest unmet needs in PS and healthcare services. The results showed that although UCCs reported high levels of perceived peer support, this does not necessarily lead to improved well-being and may instead be associated with greater caregiver burden and increased distress. Therefore, addressing both the availability and quality of support is essential to effectively meet supportive care needs and improve well-being.
Implications for practice
The high demand for practical and healthcare services among UCCs living in RRR areas suggests that specific healthcare services and staff support should be designed, implemented, and strengthened to facilitate easy access to oncology-based regional healthcare staff, referral pathways, assistance with practical tasks, and ongoing support from the initial appointment through to the bereavement period. It is worth noting that peer support programs should also focus on the quality and compatibility of peer support with UCCs’ cultural needs to effectively address their supportive needs and improve their well-being during such difficult situations. Such flexibility is particularly essential for other geographically RRR areas where UCCs may experience similar challenges related to distance, limited-service access, and reliance on home-based care, while their cultural, social, and religious needs and circumstances may vary across different settings.
Implication for future studies
Future research should incorporate qualitative methods to explore UCCs’ lived experiences of support in RRR areas, including what forms of support are most helpful, how support is accessed, and what barriers prevent caregivers from receiving timely and appropriate assistance. Longitudinal studies are also needed to clarify whether perceived peer support functions as a protective factor over time or whether caregivers report higher support primarily during periods of heightened distress. In addition, future work should examine subgroup differences (e.g., gender, relationship to the patient, cancer stage, and remoteness) to identify tailored strategies to reduce caregiver distress and address unmet needs in RRR contexts.
Strengths and limitations
This cross-sectional study explored not only unmet needs but also employed several validated instruments to assess well-being and perceived support from UCCs living in underserved areas, providing a broader understanding of UCCs residing in RRR areas rather than focusing solely on their unmet needs. Moreover, the use of SEM as an advanced analytical approach enabled researchers to have a more comprehensive examination of the relationships among variables. In addition, the participation of UCCs from different RRR areas (MMM2–MMM7) strengthened the contextual relevance for planning healthcare services for UCCs residing in such areas. However, due to the cross-sectional design, we were unable to explore barriers and facilitators in delivering high-quality peer support and assessed the perceptions of peer support solely. Moreover, as this study was conducted via an online survey rather than a paper-based survey, participants with low digital literacy or limited internet access might be underrepresented. Although participants were recruited from RRR areas of Queensland, this study was unable to distinguish between subgroups across identifying potential differences in unmet supportive care needs, perceived peer support, and well-being among RRR areas.
Conclusion
This cross-sectional study identified PS and healthcare service-related needs as the most prominent unmet supportive care needs among UCCs living in RRR areas. Higher unmet needs were associated with poorer caregiver well-being, including greater psychological and caregiver distress. Higher perceived peer support was also associated with poorer well-being, suggesting that support may be mobilized in response to distress and that the quality and appropriateness of support, not only its perceived availability, are important considerations in caregiver support initiatives in RRR settings.
Supplementary material
The supplementary material for this article can be found at 10.1017/S1478951526103940.
Data availability statement
The data that support the findings of this study are not publicly available due to ethical and confidentiality restrictions. Data may be made available from the corresponding author upon reasonable request and subject to approval from the UniSQ Human Research Ethics Committee.
Acknowledgments
The authors would like to thank Darling Downs Health and Toowoomba-based hospitals for facilitating recruitment. We also thank the Prostate Cancer Foundation of Australia and the Leukaemia Foundation for their collaboration in promoting the study flyer through newsletters, community communication channels, and social media. We are grateful to Facebook community administrators who kindly shared the flyer within Queensland’s unpaid cancer caregiver communities. Finally, we sincerely thank all Queensland participants for their enthusiasm and willingness to take part in the study.
Funding
No funding was provided for this cross-sectional study.
Competing interests
The authors report no conflict of interest in this cross-sectional study.
Ethical approval
This study was approved by the University of Southern Queensland (UniSQ) Human Research Ethics Committee (HREC), under ethics number ETH2025-0803. Participation was voluntary, and participants could withdraw from the survey at any point prior to submission without providing a reason.
Open access
