Cancer is increasingly recognised as a chronic illness, with the number of people living with a history of the disease expected to triple to 75 million people worldwide in 2030 [
]. While most survivors adjust well over time [
], a minority are at risk of adverse physical [
], psychological [
] and social [
] effects which may emerge soon after diagnosis and treatment, or in the ensuing years. Detailed knowledge about the issues faced by survivors, their care and support needs, and the extent to which these are met by current services is critical to guiding where to focus limited healthcare resources in order to deliver care that is responsive to the needs of the growing population of cancer survivors.
There are a number of different approaches for more fully understanding survivors’ cancer experiences and quantifying their outcomes including assessment of quality of life, satisfaction with health care, and needs assessment [
]. Needs assessment not only identifies needs and their importance as perceived by the survivor, but also the extent to which they are met [
]. The key strength of this approach is that it enables resources to be focused on the issues that survivors have expressed they want addressed in order to achieve optimal wellbeing.
Increasing interest in the application of needs assessment to cancer care has resulted in the development of a number of valid and reliable cancer-specific tools assessing a comprehensive range of needs [
] and a growing literature describing their administration across a variety of settings, stages in the cancer journey, and populations. A recent systematic review found that while the prevalence of unmet need among cancer survivors varied from 30% to 50% across studies, it is typically highest in the psychological, health information, and physical aspects of daily living domains [
]. While evidence about the factors that influence survivors’ unmet needs is inconsistent, a number of studies have found that those who are not in remission [
], are psychologically distressed [
] and geographically isolated [
] are more likely to report unmet needs. However, the literature is plagued by a lack of consistency in the methods used to measure, classify and report unmet needs, making it difficult to compare between studies and to generalise findings [
The seminal publication,
From Cancer Patient to Cancer Survivor: Lost in Transition
], focused the attention of the cancer control community on the survivorship stage of the cancer trajectory with a series of recommendations to accelerate progress in this area, including the need for large-scale studies using valid and reliable measures with diverse cancer populations to be conducted as a priority. Furthermore, a recent review [
] identified unmet supportive care needs as one of four main gaps in knowledge about the problems faced by adult cancer survivors. To guide care planning and help inform future health service delivery, the current study aimed to (1) describe the prevalence of adult cancer survivors’ supportive care needs, overall and by cancer type, at six months post-diagnosis; (2) identify the most prevalent items of moderate to high level unmet need and (3) identify the individual, disease, health behaviour, psychological and social factors associated with survivors reporting moderate to high level unmet psychological, health systems and information, physical and daily living, patient care and support, and sexuality needs.
This paper is based on Time 1 (T1) data collected at six months post-diagnosis from survivors participating in the population-based longitudinal
Cancer Survival Study
(CSS). The study protocol and aspects of the study findings have been reported in detail elsewhere [
]. While the term cancer ‘survivor’ has varied definitions [
], this study considers ‘survivor’ to encompass anyone diagnosed with cancer, from the time of diagnosis to the end of life [
]. This paper focuses on survivors in the late treatment to early survivorship phase of the cancer continuum.
Participants & procedures
The sample was recruited from new notifications to the two largest state-based cancer registries in Australia which together account for 60% of all new cancer cases diagnosed [
]. Eligibility was restricted to those who were (1) diagnosed in the previous six months with their first primary cancer of one of the top eight incident cancer types in Australia (prostate, colorectal, female breast, lung, melanoma, non-Hodgkin’s lymphoma, leukaemia, head & neck); (2) aged between 18 and 80 years and living in the state of New South Wales (NSW) or Victoria (VIC) at diagnosis; (3) considered by their physician to be aware of their diagnosis, physically and mentally capable of participating in the study, and sufficiently proficient in English to complete a questionnaire; and (4) alive.
The recruitment and survey methodology have been described in detail previously [
]. Briefly, eligible potential participants whose physician had given active (NSW) or passive (VIC) consent for them to be contacted about the study received a mailed package from the registries. Eligible survivors who agreed to the registries passing on their contact details to the researchers were sent a self-administered questionnaire to complete. Non-responders were sent a reminder questionnaire three weeks later and received a reminder phone call after a further three weeks. A three week interval was used to allow adequate time for survivors to receive, respond to and return the mailed questionnaire prior to receiving a reminder. Return of the questionnaire to the research team indicated voluntary consent to participate. The Human Research Ethics Committees of the University of Newcastle (H-199-1101), Cancer Institute NSW and Cancer Council Victoria approved the study.
Data were collected by self-administered questionnaire with additional clinical information obtained from the Cancer Registries for each participant.
Supportive care needs
were measured by the 34-item Supportive Care Needs Survey (SCNS-SF34) which assesses cancer-specific perceived needs across five factor analytically derived domains: psychological (10 items), health systems and information (11 items), patient care and support (5 items), physical and daily living (5 items), and sexuality (3 items) [
]. For each item, respondents indicate their level of need for help over the last month as a result of having cancer on a five point Likert scale with the following response options: 1 = no need, not applicable; 2 = no need, satisfied; 3 = low need; 4 = moderate need; and 5 = high need. For each domain, survivors were categorised as having a ‘moderate to high’ level of need if they selected response options 4 or 5 to at least one item in the domain or ‘no to low’ need if they selected response options 1, 2 or 3 to all items in the domain [
]. The SCNS-SF34 has high internal consistency with Cronbach’s alpha of at least 0.86 for each subscale, and is moderately correlated with the Hospital Anxiety and Depression Scale, Distress Thermometer and Quality of Life Questionnaire-Core 30 (QLQ-C30) [
]. Furthermore, cancer patients have reported a preference for the SCNS-SF34 over the QLQ-C30, Functional Assessment of Cancer Therapy-General and Kingston Needs Assessment–Cancer as a strategy for conveying their needs to health care providers [
Age at diagnosis and sex were obtained from the cancer registry. Current marital status, highest level of education completed, health insurance coverage, current employment situation, geographical location, size of household, and presence of physical co-morbidities were obtained by questionnaire.
Disease and treatment
Primary cancer type and spread of disease at diagnosis were obtained from the cancer registry, with survivors’ cancer categorised as ‘early/less progressed’ (in-situ or localised; grade 1 or 2; T1 or T2), ‘late/more progressed’ (invasion of adjacent organs, regional nodes or distant metastases; grade 3 or 4; not T1) or ‘not applicable’ (haematological cancers). Extent of disease at six months post-diagnosis, and cancer treatments received in the last month were obtained by questionnaire.
Seven questionnaire items adapted from existing measures assessed health behaviors: two items assessed smoking behavior, with participants classified as ‘current’, ‘former’ or ‘never smoker’ [
]; two items assessed alcohol consumption [
] and participants who consumed more than two standard drinks on any day were classified as being at ‘increased lifetime risk of harm’ from alcohol related injury or disease [
]; and three items assessed physical activity [
] with participants classified as ‘sufficiently active’ (at least 150 minutes of physical activity per week), ‘insufficiently active’ (participating in some activity but not enough in total time) or ‘sedentary’ (no physical activity) [
Two questionnaire items assessed treatment for mental health illness (eg. depression, anxiety, schizophrenia) before and since the cancer diagnosis. Coping was assessed by the 21-item Mini Mental Adjustment to Cancer Scale (Mini-MAC) which measures five cancer-specific coping strategies: helplessness-hopelessness, anxious preoccupation, fighting spirit, cognitive avoidance and fatalism [
]. The mini-MAC has demonstrated reliability with Cronbach alpha coefficients for each subscale ranging from 0.62–0.88. Raw scores for each subscale were standardised from 0 to 100 [
] and survivors who scored in the top 16% of each distribution were classified as a ‘case’ on that specific coping strategy [
Social support was assessed by the MOS Social Support Survey (MOS − SSS) which measures four domains of functional support: emotional/informational, tangible, affectionate, and positive social interaction [
]. Raw subscale scores were standardised from 0 to 100 and survivors who scored in the bottom one-third of each distribution were classified as having ‘low’ availability of that particular type of social support (Sherbourne, personal communication). The survey has high internal consistency with alpha coefficients exceeding 0.91 for each subscale and demonstrated validity with the chronic illness population [
Due to small numbers, data from survivors diagnosed with non-Hodgkin’s lymphoma or leukaemia were combined and categorised as ‘haematological’ cancer. The proportion of survivors who reported either ‘no needs’ (ie. selected response option 1 or 2 to all 34 items), ‘low needs’ (ie. selected response option 3 to at least one item, but did not select response option 4 or 5 to any items) and ‘moderate to high needs’ (ie. selected response option 4 or 5 to at least one item) was calculated overall and by cancer type, with 95% confidence intervals. The association between cancer type with reporting ‘no needs’, ‘low needs’ and ‘moderate to high needs’ was examined using chi-square analyses. For each domain, the proportion of survivors who reported ‘moderate to high needs’ versus ‘low or no needs’ was calculated with 95% confidence intervals. The proportion of survivors who endorsed each SCNS-SF34 item at either a ‘moderate’ or ‘high’ level was calculated with 95% confidence intervals and the ten most prevalent items and their corresponding domain identified. Chi-square analyses examined the association between survivors’ individual, disease, health behaviour, psychological and social characteristics with ‘moderate to high needs’ versus ‘low or no needs’ for each domain. Multiple logistic regression analyses were then conducted to examine factors associated with ‘moderate to high needs’ while adjusting for potential confounders. Variables with a p-value ≤0.2 on univariate analyses were included in a backward logistic regression model for each domain. Variables were removed from the model if they had a p-value <0.1 on the likelihood ratio test; those with a p-value ≤0.05 were considered statistically significant.
The registries were required to recruit a quota of 1660 eligible survivors who consented to being contacted about the study. Based on previous experience [
], we estimated that 80% of survivors would return a completed survey, resulting in a sample size of approximately 1320 at T1. Assuming a prevalence of moderate to high needs of 20%, a sample of this size would allow the proportion of survivors with unmet needs to be estimated with 95% confidence intervals within ± 3%, and provide 90% power to detect differences of 7% between categories of study factors associated with moderate to high needs at the 5% significance level.
This study found that six months after a cancer diagnosis, about one-third (37%) of survivors reported one or more items of moderate or high level unmet need, while almost two thirds (63%) reported either no or low level unmet needs. The most commonly reported moderate to high level unmet needs were from the psychological and physical and daily living domains. This is consistent with other recent needs assessments conducted with samples of cancer survivors at the end of treatment [
], in early phases of survivorship [
] and in long-term survivorship [
]. However, previous studies [
] found between 43%-60% of survivors reported at least one moderate or high level unmet need, compared to 37% of survivors in this study. Similarly, unlike earlier studies which found the most prevalent item of moderate or high unmet need occurred among 27-40% of recent survivors [
], the most commonly reported item of unmet need in this study was endorsed by only 15% of survivors.
The prevalence of unmet need reported by survivors in this study is clearly lower than previously reported, despite using the same validated instrument, and classification of unmet need. This may be because earlier studies of cancer survivors diagnosed with a diversity of cancer sites did not use population-based samples [
] and are therefore more susceptible to selection bias. In contrast, we used the two largest state-based cancer registries in Australia to assemble a population-based sample of survivors in the very early stages of cancer survivorship. Given that the study sample is generally representative of the source population, we are confident in our findings that most survivors’ supportive care needs, as measured by the SCNS-SF34, are relatively well met.
Due to the size and composition of the study sample, we were able to directly compare the prevalence of supportive care needs between seven common cancer types in Australia [
]. This bivariable analysis revealed significant variation across cancer types, with particularly low levels of unmet need reported by survivors of melanoma, 65% of whom reported no items of unmet need. This is fitting with our anecdotal experience whereby participants who were survivors of melanoma often questioned the legitimacy of their contribution to the study as they perceived themselves to have suffered less than survivors of other cancer types, and therefore less deserving of attention. Australia has the world’s highest incidence rate of melanoma; it is typically identified at early stages when simple treatment such as surgery will achieve a good prognosis [
]. It is possible that the omission of melanoma survivors from the sample composition of previous studies [
] may have contributed to their higher prevalence of unmet need compared to this study. In contrast, the highest levels of unmet need were reported by survivors of lung cancer, with 60% reporting at least one item of moderate or high level need. Given the high level of burden associated with lung cancer in terms of poor prognosis, treatment side effects and declining physical health, this finding is not surprising.
Subgroups of survivors with domain-specific and widespread unmet needs were identified. After adjusting for a comprehensive range of individual, disease, health behaviour, psychological and social factors, cancer type was found to be significantly associated with moderate to high level unmet physical and daily living, and sexuality needs only. In particular, survivors of lung cancer had the highest odds of reporting unmet physical and daily living needs, while survivors of prostate cancer had extremely high odds of reporting unmet sexuality needs. These findings suggest that the type of unmet need experienced by survivors does not routinely differ between cancer types. Rather, the notion of cancer site- specific unmet needs appears to apply to only a few explicit dimensions of unmet need.
Consistent with previous studies, not being in remission was associated with unmet health system and information, and patient care and support needs; this is not surprising given this subgroup of survivors is likely to be receiving intermittent treatment and symptom management. While almost three quarters of survivors reported not receiving any active treatment in the last month, we did not assess if participants had completed all active treatments given the changeable and uncertain nature of adjuvant treatment regimes. While each treatment was considered separately, having received chemotherapy in the last month was the only treatment associated with higher odds of reporting unmet needs. Interestingly, physical activity was the only health behaviour associated with unmet needs, with sedentary survivors reporting higher odds of unmet psychological, and physical and daily living needs. Although 37% of the sample resided in regional or remote areas, our results did not support the findings from previous studies of an association between rural location and unmet needs. On account of the range of study factors examined in this study, a number of associations were established for the first time. Low levels of social support and maladaptive coping styles were associated with multiple domains of unmet need. Notably, survivors who were identified as a case on anxious preoccupation coping had more than twice the odds of reporting unmet needs across all five domains. While causation cannot be inferred, the new associations identified in this study are particularly valuable because social support and coping style are potentially amenable to intervention. In particular, attention could be directed towards exploring the contribution that targeted coping interventions focusing on anxiety and helplessness, could make towards the prevention of or reduction in survivors’ unmet needs across a number of domains.
Strengths and limitations
While previous needs assessments have also included a diversity of recent cancer survivors [
], the population-based sampling method used in this study is a major strength as it increases the generalisability of the results. In Australia, the notification of cancer to the cancer registry is a statutory requirement under the state and territory Public Health Acts. Indices of registry data quality demonstrate that the level of case ascertainment is high and the data collected are accurate [
]. However, the overall response rate was 41% (1360/3315 eligible individuals) and may raise concerns about response bias. While this response rate seems low, it is higher than that achieved by other studies which also used cancer registries to recruit diverse samples of recent survivors [
]. Survivors who were not proficient in English were excluded due to the prohibitive cost of translating the questionnaire into other languages and may have resulted in an underestimate of the prevalence of unmet needs given that language barriers have been associated with poorer access to health care services. Our outcome measure, the SCNS-SF34, is a well-validated tool for assessing multiple dimensions of supportive care need and was developed with diverse samples of individuals diagnosed with cancer in terms of cancer type and time since diagnosis [
]. However, it is possible that the SCNS-SF34 may not fully capture the unique needs of cancer survivors in the late treatment to early survivorship phase of care, and therefore this study may underestimate the prevalence of unmet need reported by survivors at six months post-diagnosis. Since this study commenced, two cancer survivor-specific needs assessment tools [
] have been developed and should be considered for use in future studies.
About one-third of cancer survivors in the transition from late treatment to early survivorship had moderate to high levels of unmet need, particularly in the psychological and daily living domains. Our findings directly inform health care professionals and organisations involved in the provision of survivorship care about the actions, resources and services most needed by subgroups of survivors. Our findings also suggest that coping support interventions may have the potential to contribute to the prevention or reduction of survivors’ unmet needs across all domains. However, it is important not to overlook the finding that 63% of survivors in this study reported no or low level unmet needs at six months post-diagnosis and for whom current care appears to adequately meet their needs. On the basis that a valuable new perspective about how to prevent or reduce cancer survivors’ unmet needs could be gained from those with no unmet needs, future research should seek to identify and better understand this subgroup of survivors.
The research on which this paper is based was conducted as part of the
Cancer Survival Study led by Allison Boyes and Afaf Girgis. Funding for this study was provided by the National Health & Medical Research Council (ID 252418), Cancer Council NSW, Hunter Medical Research Institute, Honda Foundation and University of Newcastle. The Victorian Cancer Registry (Cancer Council Victoria) and NSW Central Cancer Registry (NSW Department of Health and Cancer Institute NSW) assisted with case recruitment. Our sincere thanks to the cancer survivors who provided the survey data, and Christophe Lecathelinais for statistical assistance.
The authors declare that they have no competing interests.
AB participated in study conception, design and acquisition of funding; was responsible for implementing the study protocol; performed some of the statistical analysis; interpreted the data and drafted the manuscript. AG participated in study conception, design and acquisition of funding. CD participated in the study design and acquisition of funding, and coordinated the statistical analysis. AZ helped to implement the study protocol and performed some of the statistical analysis. All authors participated in revising the manuscript, and read and approved the final version.