Introduction

Haematological cancers include leukaemias, lymphomas and myeloma and collectively account for 6.5% of all cancers worldwide [1]. The sub-types of haematological cancers are diverse, comprising of incurable types, highly aggressive forms or slowly progressing cancers [2]. Treatments may be intensive and include chemotherapy, radiotherapy and bone marrow transplantation [2]. The National Cancer Institute (NCI) and National Coalition for Cancer Survivorship (NCCS) define a cancer survivor as someone diagnosed with cancer, from the time of diagnosis until the end of that person’s life [3, 4]. While the cancer types and treatments vary, similar experiences that haematological cancer survivors face include an increased risk of infection, prolonged fatigue, extended recovery trajectories [5] and long-term follow-up [2]. Long-term effects of haematological cancer and/or treatment may include cognitive deterioration and adverse effects on employment, psychological wellbeing and social functioning [6]. Given the burden haematological cancer survivors experience, understanding in which areas survivors need further assistance is important for informing where services direct additional support and resources for this population.

Supportive care needs assessment identifies the physical, informational, emotional, psychological, social, spiritual and practical needs that cancer survivors require help with [7]. Systematic reviews found that little research has examined haematological cancer survivors’ unmet supportive care needs [8, 9] and that shortcomings of these studies included inadequate sample sizes and using only one or two treatment centres to recruit survivors which may limit the generalisability of findings [9]. Australian and UK studies with a heterogenous group of haematological cancer survivors [10] or multiple myeloma survivors [11] reported that accessible hospital parking [11], acquiring life and/or travel insurance [11], handling the fear of recurrence [10, 11] and having access to an ongoing case manager [10, 11] were common unmet needs. In two other studies, haematological cancer survivors reported that dealing with feeling tired was the most common unmet need [12, 13].

Haematological cancer survivors who reside in rural areas may have unique or additional unmet needs compared to their urban counterparts, because there are fewer cancer services in rural areas [14]. Challenges faced by rural cancer survivors include travelling considerable distances to access cancer treatment, the financial cost of travel and/or staying away from home, separation from family, difficulty accessing specialists, lack of coordination between specialists and rural health professionals and difficulty accessing follow-up and support services [14, 15]. Although some studies discovered that compared to urban cancer survivors rural cancer survivors were more likely to report unmet needs for some domains such as physical/daily living [16, 17], other research has found no disparities in the unmet needs of rural compared to urban survivors [18]. However, this prior research surveyed those diagnosed with breast cancer [16], gynaecological cancer [17] or any type of cancer [18]. None of these studies focussed on a haematological cancer survivor population.

The Survivor Unmet Needs Survey (SUNS) has been found to be a psychometrically robust measure among those with various cancer types [19] and haematological cancer survivors [20]. The SUNS measures unmet needs across the following five domains: information needs, financial concerns, access and continuity of care, relationships and emotional health [19]. Given that the SUNS provides a detailed list of needs and has demonstrated acceptable psychometric properties, the SUNS is an appropriate tool for comparing the unmet needs between rural survivors and urban survivors.

There is a lack of evidence that examines the unmet needs of haematological cancer survivors residing in rural compared to urban areas. Such information could be used to inform the delivery of cancer services to these groups. Therefore, this study aimed to examine the following:

  1. 1)

    The most common “high/very high” unmet supportive care needs among rural versus urban haematological cancer survivors;

  2. 2)

    mean and median scores of rural versus urban haematological cancer survivors for each of the five SUNS’ domains and

  3. 3)

    socio-demographic, rurality, cancer history and psychological characteristics associated with unmet need scores for each of the SUNS’ domains.

Methods

Sample

Survivors were eligible to participate if they met the following inclusion criteria: a diagnosis of haematological cancer including leukaemia, lymphoma and myeloma in the past 7 years; aged 18–80 years at enrolment; and sufficient English to complete the survey. This study used the NCI and NCCS definition of a cancer survivor as someone diagnosed with cancer, from the time of diagnosis until the end of that person’s life [3, 4].

Procedure

Between September 2010 and September 2013, haematological cancer survivors were recruited via five Australian state cancer registries (registries 1, 2, 3, 4 and 5). Residential postcodes were used to stratify the sample, with postcodes classified as major city/inner regional or outer regional/remote/very remote based on the Accessibility and Remoteness Index of Australia (ARIA+) [21]. All eligible haematological cancer survivors residing in an outer regional/remote/very remote location were approached due to limited numbers while survivors living in a major city/inner regional location were randomly selected.

Registry 1 recruited eligible survivors by mailing the questionnaire package directly to survivors. Non-respondents were mailed another questionnaire package after 4 weeks. Registries 2, 3 and 4 sought passive consent from the survivor’s treating clinician. These cancer registries contacted the clinician via mail to seek clinician consent to contact the survivor. If the registries received consent from the clinician to contact the survivor or received no reply from the clinician within 4 weeks, the registries contacted the survivor by mail and sought the survivor’s permission to provide their contact details to the research team. Registry 5 used active clinician consent to recruit eligible survivors. This registry obtained written consent from the treating clinician before contacting the survivor via mail to ask if they would give the registry permission to provide their contact information to the researchers. Survivors from registries 2, 3, 4 and 5 who agreed to being contacted by the research team were mailed a letter, survey, consent form to have their demographic and disease details released from the registry and envelope for return of the survey. Non-respondents were mailed another questionnaire package 4 weeks later. After a further 4 weeks, a reminder telephone call was made to those who did not respond.

The University of Newcastle Human Research Ethics Committee and ethics committees for the cancer registries granted ethical approval.

Measures

Survivor Unmet Needs Survey

The 89-item SUNS measures the unmet needs experienced over the last month. The SUNS consists of five domains: information needs (8 items), financial concerns (11 items), access and continuity of care (22 items), relationships (15 items) and emotional health (33 items) [19]. The response options are no unmet need, low unmet need, moderate unmet need, high unmet need and very high unmet need. The SUNS has acceptable face validity and content validity with haematological cancer survivors that has been demonstrated through semi-structured interviews with 17 haematological cancer survivors [20]. Furthermore, the SUNS has acceptable construct validity (5-factor structure, all factor loadings > 0.30) and convergent validity (all five SUNS domains had correlation coefficients from 0.44 to 0.73 with all three subscales of the Depression Anxiety and Stress Scale) with haematological cancer survivors [20].

Depression, Anxiety and Stress Scale

The Depression, Anxiety and Stress Scale (DASS-21) contains three subscales measuring depression (7 items), anxiety (7 items) and stress (7 items) [22].

Socio-demographic characteristics

Consenting survivors had age, sex and residential postcode obtained from cancer registry records. Marital status, highest level of education and employment status were self-reported.

Cancer-related characteristics

Cancer type and date of diagnosis were extracted from cancer registry records for consenting survivors. Survivors were asked about whether they had private health insurance, the treatments that they had ever or were currently receiving and time taken to travel to treatment.

Statistical analysis

SAS software version 9.4 was used to perform statistical analyses. Means, standard deviations or medians (continuous items) or frequencies and percentages (categorical items) were calculated. The five SUNS’ domain scores were determined by adding the scores for all questions in that domain and dividing by the number of non-missing responses in that domain. A response of no unmet needs was scored 0, low unmet need a 1, moderate unmet need a 2, high unmet need a 3 and very high unmet need a 4. Data from survivors who completed > 70% of questions per domain were included in the analysis. As per the scoring manual, DASS-21 subscale scores were determined by adding all questions in a subscale and multiplying by two [23]. Cut-off scores used to classify each condition were depression (> 9), anxiety (> 7) and stress (> 14) [23]. Participants with more than one item missing from the depression, anxiety and stress subscales were not included in the analysis of that subscale.

Quantile regression was used to examine socio-demographics, rurality, cancer history and psychological characteristics associated with survivors’ unmet need scores for each SUNS domain. Results are presented for analysis of the median (the 50% quantile). Parameter estimates and 95% confidence intervals were calculated. Adjusted Wald p values for the overall effect of each variable have also been presented. Characteristics in the regression models were sex, age, education level, employment status, rurality, cancer type, time since diagnosis, time to travel to treatment, private health insurance, depression, anxiety and stress.

Results

Sample

Of 4299 eligible haematological cancer survivors identified via the five registries, 2287 survivors gave approval to be mailed the survey and 1511 returned a completed survey (66.1% of survivors mailed survey; 35.1% of eligible survivors). The ARIA+ was used to classify respondents as rural survivors (i.e. resided in outer regional/remote/very remote locations, ARIA+ > 2.4) or as urban survivors (i.e. resided in major cities/inner regional areas, ARIA+ 0–2.4) [21]. A visual illustration of the ARIA+ classifications can be found at http://www.spatialonline.com.au/ARIA_2011/ [24] where outer regional, remote and very remote locations in Australia are shaded in green, light blue and dark blue, respectively and major cities and inner regional areas are shaded in red and orange, respectively. Registry data in relation to rurality was obtained for 1417 respondents, with 1145 classified as urban survivors and 272 as rural survivors.

Survivor characteristics and treatment factors

Table 1 outlines the characteristics of rural and urban survivors and the total sample. The mean age of participants at diagnosis was 58 years and the majority were male (56.9%), married or living with a partner (78.1%) and resided in an urban location (80.8%). Most survivors were diagnosed with non-Hodgkin lymphoma (56.3%), were diagnosed more than 3 years ago (58.8%) and had received chemotherapy (82.3%). For 22.9% of survivors, it took more than 1 h to travel to treatment. More rural residents (49.2%) than urban residents (16.7%) travelled more than an hour to treatment.

Table 1 Characteristics of rural and urban survivors and the total sample
Full size table

Most common high/very high unmet needs nominated by rural and urban survivors

Table 2 lists the most common high/very high unmet needs for rural and urban survivors separately. Seven of the most prevalent high/very high unmet needs were common to both urban and rural survivors. “Dealing with feeling tired” was the most common “high/very high” unmet need for the rural (15.2%) and urban (15.5%) survivors. “Coping with having a bad memory or lack of focus” was the second highest high/very high unmet need for urban survivors (13.1%) and the third highest for rural survivors (11.8%). “Having access to cancer services close to my home” was the second highest high/very high unmet need for rural survivors (14.3%), but was not listed among the most common high/very high unmet needs for urban survivors (6.5%). Two other items, “Dealing with feeling guilty about what I have put others through” (11.1%) and “Finding car parking that I can afford at the hospital or clinic” (11.0%) were also listed in the most common unmet needs for rural survivors only. Most items nominated as high/very high unmet needs by rural and urban survivors were from the emotional health domain.

Table 2 Most common high/very high unmet needs identified by rural and urban survivors
Full size table

SUNS domain scores

Table 3 outlines the mean and median scores for the five domains for the rural survivors and urban survivors, respectively. The mean and median domain scores were skewed towards no unmet needs/low unmet needs. The emotional health domain had the highest mean score for rural respondents (mean = 0.66, SD = 0.84) and urban participants (mean = 0.73, SD = 0.92). This aligns with the results described in Table 2 that demonstrated items from the emotional health domain were most commonly nominated as areas of unmet need. In contrast, the access and continuity of care domain had the lowest mean unmet need score for rural survivors (mean = 0.39, SD = 0.60) and urban survivors (mean = 0.37, SD = 0.64).

Table 3 Mean and median SUNS domain scores
Full size table

Factors associated with unmet needs in haematological cancer survivors

Characteristics associated with each of the SUNS’ five domains are outlined in Table 4.

Table 4 Characteristics associated with unmet need domain scores
Full size table

Information needs

Being employed, aged 18–39 years, diagnosed in the last 24 months, being depressed, anxious or stressed were significantly associated with an increased unmet information needs domain score.

Financial concerns

Taking more than 1 h to travel to treatment, having depression, anxiety or stress were significantly associated with a higher unmet financial concerns domain score.

Access and continuity of care

Being diagnosed in the last 24 months, taking more than 1 h to travel to treatment and being depressed, anxious or stressed were significantly associated with an increased unmet access and continuity of care domain score.

Relationships

Characteristics significantly associated with an increased unmet relationships domain score were having a university education, being aged 18–39 years and having depression, anxiety or stress.

Emotional health

Being aged 18–39 years, having been diagnosed in the last 24 months, a myeloma diagnosis (as compared to non-Hodgkin lymphoma) and being depressed, anxious or stressed were significantly associated with a higher unmet emotional health domain score. In contrast, being a rural resident was associated with a decreased unmet emotional health domain score.

Discussion

This study compared the high/very high unmet supportive care needs of rural haematological cancer survivors to their urban counterparts. Although < 20% of respondents indicated high/very high unmet needs on any individual item, “dealing with feeling tired” was the most common unmet need for both rural and urban survivors which is consistent to prior research with haematological cancer survivors [12, 13]. A systematic review has reported that aerobic exercise in supervised environments is effective in reducing cancer-related fatigue [25], which suggests that health care providers could routinely offer and encourage haematological cancer survivors to participate in exercise programs in an effort to help manage fatigue. However, access to appropriate and effective programs may be variable.

Our study also found that for both rural and urban survivors, the emotional health domain had the highest mean unmet need score while the access and continuity of care domain had the lowest unmet need score. This is similar to Canadian research with rural survivors with various cancer types that also found that the access and continuity of care domain had the lowest mean unmet need score while emotional health was ranked second after information as an unmet need [26]. However, even though in our study access and continuity of care had the lowest mean domain score, access to cancer services close to home was the second highest unmet need for rural survivors. To address emotional health unmet needs among haematological cancer survivors, health care providers should undertake distress screening regularly and ensure referral to effective and accessible psychosocial support. It may also be helpful to explore whether current forms of emotional and psychosocial support are well suited to the needs of haematological cancer survivors and accessible to survivors during and beyond the active treatment phase.

Rurality was associated with the emotional health domain only, with rural haematological cancer survivors reporting lower unmet emotional health scores than urban participants. There is evidence that factors protective of emotional health in rural cancer survivors but not urban cancer survivors include active coping and positive reinterpretation [27] and the belief that a person’s health is decided by powerful others [28]. However, similar to research with survivors diagnosed with any type of cancer [18], haematological cancer survivors’ scores on information, financial, access and continuity of care and relationships unmet needs did not differ by rurality. Despite the additional transport and distance-related challenges faced by rural survivors diagnosed with haematological cancer, these challenges do not appear to result in poorer overall emotional health and unmet need.

Despite the lack of differences in the unmet need domain scores between rural and urban survivors, haematological cancer survivors who travelled for more than 1 h to treatment had increased unmet need scores in the financial concerns and access and continuity of care domains. Telemedicine via video-conferencing has been successfully used in Australia to provide specialist consultations and cancer treatment [29] and may help to reduce the unmet access and continuity of care needs of those who travel long distances for treatment by allowing them to receive treatments closer to home. Given the widespread availability of video communication software such as Skype and FaceTime, such platforms could also be used to provide care to survivors who travel more than 1 h to treatment. Furthermore, cancer organisations, such as Cancer Councils, provide online peer support groups facilitated by health care professionals which may also assist in addressing the unmet access and continuity of care needs of haematological cancer survivors who travel long distances to treatment. Better coordination between cancer specialists and general practitioners may also help reduce this unmet need in those who travel more than 1 h to treatment. Despite the availability of financial support and accommodation for haematological cancer survivors via organisations such as the Leukaemia Foundation of Australia, those who travelled for more than 1 h to treatment had increased unmet financial concerns. It is unclear from our data whether this reflects that survivors were unaware of these resources, could not access them or were able to access these but found them insufficient to alleviate the financial burden of travel. A recent study of cancer patients indicated that 74% did not use financial assistance, and more than one third of these patients did not know that financial assistance was available [30]. Future research should examine haematological cancer survivors’ knowledge and ability to access financial support and also assess what type and level of financial assistance would be sufficient and beneficial to survivors, particularly those who travel long distances for treatment.

Associations were found between depression, anxiety and stress and increased unmet need scores across all five domains. These findings are similar to research with multiple myeloma patients [11] and other cancer types [31, 32] that found that there was a relationship between depression and anxiety with unmet needs. Implementing strategies such as additional assistance from health care providers, cancer organisations and peer support groups to address information, financial, access and continuity of care, relationships and emotional unmet needs among haematological cancer survivors with depression, anxiety or stress may be beneficial. Furthermore, consistent with prior research with haematological cancer survivors [10, 13, 33] and other cancer survivors [16, 34], younger age was associated with survivors reporting increased unmet need scores for information needs, relationships and emotional health. A diagnosis of haematological cancer 1–24 months ago was also associated with increased unmet need scores on three of the five domains, specifically, information, access and continuity of care and emotional health indicating that this subgroup of survivors may benefit from additional support in these areas.

To our knowledge, this study of over 1500 participants is the largest to investigate the unmet supportive care needs of a heterogeneous group of haematological cancer survivors. The strengths of this study include using rigorous sampling techniques to recruit a national sample of survivors from five Australian state cancer registries and the use of an unmet need measure shown to have acceptable reliability and validity with haematological [20] and other cancer survivors [19]. The study limitations include that although a heterogeneous sample of haematological cancer survivors was surveyed, because more than half were diagnosed with non-Hodgkin lymphoma and other haematological cancer types were represented in smaller numbers, the findings may have limited generalisability to all haematological cancer types.

While overall haematological cancer survivors reported low levels of unmet needs, particular subgroups, for example those with depression, anxiety, stress, younger survivors and those diagnosed 1–24 months ago, were identified as a priority for targeted service provision. In contrast, generally, unmet needs did not differ for rural survivors compared to urban survivors, indicating that rural survivors (most diagnosed > 3 years ago) have managed their cancer and treatment as well as their urban counterparts.