Evaluation of Factors Associated With Unmet Needs in Adult Cancer Survivors in Canada

This cross-sectional survey study analyzes findings from a nationwide survey of adult cancer survivors in Canada, including the physical, emotional, and practical concerns and associated unmet needs 1 year or more after the end of treatment.


Introduction
The Canadian health care system is facing a critical challenge: how do we provide comprehensive follow-up care that meets the individual needs of the growing number of cancer survivors in an oftensiloed system? Currently, more than 1.6 million Canadian individuals are living after cancer treatment. 1 Survival is improving for most cancers, primarily owing to early diagnosis, better staging, multimodal therapies, and improved surgical techniques. Evidence indicates the presence of physical, psychosocial, practical, and emotional well-being challenges (eg, fear of recurrence, living with uncertainty, adjusting to a new normal, and complicated comorbidities) as patients transition from treatment to the posttreatment phase of their cancer trajectory. [2][3][4][5][6][7][8][9][10] Without appropriate support systems in place, these cancer survivors face a substantial burden because of the inadequacy of health care services in meeting the broad spectrum of needs that affect their health and quality of life. 2,[11][12][13][14][15][16][17][18] However, existing literature is limited, preventing a comprehensive understanding of the extent of unmet physical, emotional, and practical needs of cancer survivors at a national level. Moreover, existing survey tools do not capture the full extent of unmet needs, lacking comprehensiveness, internal consistency, content and structural validity, and a conceptual framework in the development process. 19 To address this gap in knowledge, the Canadian Partnership Against Cancer, in collaboration with the 10 provincial cancer agencies, developed and disseminated the national Experiences of Cancer Patients in Transitions Study. The Canadian Partnership Against Cancer is an independent organization funded by the Canadian federal government with the long-term objectives of reducing the incidence of cancer, reducing the likelihood of dying from cancer, and improving the quality of life of those affected by cancer.
In the present study, we analyzed the Experiences of Cancer Patients in Transitions Study (referred to hereafter as survey) to identify the physical, emotional, and practical challenges encountered by adult cancer survivors between 1 and 3 years after completing curative-intent treatment. Knowledge of the gaps in care can guide the development and implementation of programs and services to address the broad scope of needs of cancer survivors.

Survey Development and Dissemination
Ethics and privacy approvals for the survey were obtained by the cancer agencies in the 10 Canadian provinces from their respective ethics boards before data collection. Ethics approval for this analysis was obtained from the McMaster Research Ethics Board. Written informed consent was obtained from participants in Ontario and Quebec only. In the remaining provinces, completion of the questionnaire indicated informed consent. The present study followed the American Association for Public Opinion Research (AAPOR) reporting guidelines.
The survey was developed through an iterative process based on a conceptual framework guided by literature reviews, consultations with stakeholders (ie, cancer survivors, clinicians, and system leaders), and extrapolation of existing surveys. The survey was designed to (1) assess the emotional, physical, and practical needs of cancer survivors between 1 and 3 years after treatment; (2) identify the specific needs of most survivors; (3) identify the most vulnerable survivors, along with factors associated with unmet needs; and (4) ascertain the factors and/or resources associated with needs being met. A thorough description of the survey methods, development process, pilot testing, and dissemination is published elsewhere. 20 Twenty different concerns (listed in Table 1; eAppendix in the Supplement) were clustered into broader physical, emotional, and practical domains. The survey inquired about the severity of concern (response options: big, moderate, small, or not a concern), whether help was sought (yes or no), and how easy it was to get help for the concern (very easy, easy, hard, very hard, or didn't get any help). Excerpts from the survey pertaining to physical, emotional, and practical domains are included in eAppendix in the Supplement; the survey is available on the Canadian Partnership Against Cancer System Performance website.
The survey was disseminated in 2016 to eligible participants through the mail but provided recipients with the option of completing it online. The survey package included a cover letter, with a description of the study, confidentiality agreement, URL link and personal identification number to the encrypted online survey, and informed consent documents in Ontario and Quebec (in which written consent was required, whereas completion of the questionnaire indicated consent in the other provinces); the paper survey with a preprinted barcode and associated personal identification number (for those who preferred to complete the paper version); and a preaddressed, prepaid return envelope. Designated staff were available to assist respondents during the data collection phase of the study.
The study population was selected through probability sampling. A response rate of 30% was assumed, and thus the adult sample population was calculated with a margin of sampling error of ±5% at the 95% CI by disease site and province. 20 The number of eligible survivors for each province was estimated from the prevalence of specific cancer type at the national level and the incidence at the provincial level. The smaller provinces were less likely to achieve adequate sample size, owing to CI precision. Therefore, all eligible survivors were surveyed in smaller provinces. In larger provinces, the number of eligible participants was larger than that required according to the sample size calculation; therefore, random samples for disease sites were drawn. Furthermore, an external vendor was hired to assist with survey distribution, design, and preliminary analysis. Therefore, confidential data on survivor characteristics could not be shared with the vendor, which was required to assign weighting to the survey to make the sample representative of all Canadian cancer survivors.
Thus, this sample should not be generalized to represent all survivors in Canada.

Study Population
The original study population consisted of adolescent and adult patients with cancer who completed treatment within the past 1 to 3 years (n = 40 790

Outcomes
The outcomes of this study were as follows: (1)

Statistical Analysis
The binary variables for physical, emotional, and practical domains and subsequent unmet needs were derived by aggregating the responses to individual concerns under these 3 domains. These variables were used to identify the extent of concerns and associated unmet needs. For instance, if the participants reported any concern (of low, moderate, or high severity) under emotional domain, the emotional concern variable was coded as yes. Subsequently, the unmet need for the emotional concern variable was coded as yes if the respondents did not seek any help, sought help for a particular concern but reported that it was hard or very hard to get help, or did not get any help despite seeking it (eAppendix in the Supplement includes relevant excerpts from the survey).
Univariate analysis for binary and categorical variables was performed using the χ 2 test to identify the important factors for inclusion in the regression models. Multivariate logistic regression analysis was performed to determine the adjusted associations between reporting of unmet needs in emotional, physical, and practical domains and independent variables. Using a forward stepwise approach, we sequentially included in the models the patient demographic-, treatment-, clinician-, and site-specific P values. A 2-sided P < .05 was considered to be statistically significant.
All analyses were performed with Stata, version 9.3 (StataCorp LLC). Data synthesis and quality assessment were conducted in 2017. Data analysis was completed by independent researchers in July 2019.

Results
In total, 10 717 adult respondents were included in this analysis. Among these respondents, 5660  of the type of emotional concern. In contrast, less than 25% of the respondents reported any type of practical concern. The extent of unmet practical needs ranged from getting to and from   (Figure). Help for all of their concerns was sought by 1105 participants (27%) with 3 concerns in all 3 domains, by 1273 (31%) of those with 2 concerns in 2 domains, and by 783 (45%) of those with concern in 1 domain (Figure).

Factors Associated With Unmet Needs
The results of multivariate analyses are summarized in needs compared with those who received surgical treatment alone.   Survivors of breast cancer had the largest sample size (n = 3607) and reported one of the highest proportions of unmet needs across all 3 domains in unadjusted analysis (Table 2). Therefore, breast cancer was used as a reference group in the adjusted analysis to identify the differences in unmet needs by cancer type. When compared with breast cancer survivors, melanoma cancer survivors had a significantly higher likelihood of reporting unmet emotional needs (OR, 1.75; 95% CI, 1.17-2.61; P = .01), whereas survivors of prostate (OR, 0.60; 95% CI, 0.43-0.84; P < .001) and

JAMA Network Open | Health Policy
hematological (OR, 0.70; 95% CI, 0.50-0.99; P = .04) cancers were significantly less likely to report unmet needs for physical concerns. No differences in reporting of unmet practical needs were found between cancer types.  30 Despite the widespread acknowledgment of the importance of integrated survivorship care, the optimal survivorship care model that clarifies practitioner roles and responsibilities remains poorly understood, and implementation of efforts remains fragmented. 31 We suggest that such a model be tailored to address barriers, including funding constraints, staffing shortages, and lack of shared information systems between different health care sectors. 31 The actionable recommendations from these findings are as follows: (1) better patient education, to teach patients to recognize health care concerns that are not a normal part of cancer survivorship; (2) early involvement of primary care physicians in cancer care, to ensure the smooth transition from cancer treatment to survivorship care overseen by a primary care physician; (3) improved coordination and communication between primary care physicians and oncologists, to provide holistic support to survivors; (4) access to alternative care practitioners (eg, nurses, occupational therapists, and social workers) in geographic areas with a shortage of primary care physicians; and (5) development of tools for risk stratification of patients during cancer treatment and preemptive allocation of resources to address an anticipated high level of needs based on predisposing factors. More research is required into how communication, coordination, and integration between cancer care teams and primary care physicians can successfully occur to ensure that survivor needs are met in the right way, at the right time, and by the most appropriate professional.

Discussion
limit the generalizability to all cancer survivors in Canada. However, this survey was a large national study of participants selected from provincial cancer registries and hence was more generalizable than others in the literature. Third, the sample may include a large proportion of older adults with other comorbidities that may cause symptoms such as pain, mobility limitations, fatigue, and emotional distress. Thus, distinguishing the symptoms associated with cancer from those associated with other comorbidities may not have been easy for survey respondents and may have led to overreporting in some cases. However, the presence of comorbidities was not a significant factor in unmet needs in the adjusted analysis. Nonetheless, living with comorbidities is a reality for survivors and may add to the complexities of recovery and coping with the aftermath of cancer.

Conclusions
As the number of cancer survivors continues to grow, identifying and addressing their needs is imperative to ensure the provision of high-quality survivorship care. Our findings suggest that the high prevalence of unmet needs in this population underscores the need for system-level changes, including earlier identification of patients at high risk for challenges during survivorship, models of shared care that include cancer specialists and primary care teams, and increased knowledge about the importance of providing adequate survivorship care that will enhance the quality of life for cancer survivors in Canada.