INTRODUCTION
Cancer is a disease whose illness trajectory varies according to tumor site and grade (Aslett et al., Reference Aslett, Morrison and Zinovieff2009). The symptoms, treatments, and level of care needed by the cancer patient parallel this, and their disease course may range from a sudden and rapid decline to a prolonged course filled with sudden deteriorations in condition. Informal caregivers are expected to take on many roles: medical and financial decision maker, patient advocate, and care provider (National Cancer Institute, 2015). Moreover, in contrast to other diseases, cancer caregivers' tasks involve patients who enter the survivorship phase, including the psychological burden related to the possibility of disease recurrence. At the same time, the normal daily grind of their lives as well as that of the other family members they are already caring for continue, which exponentially increases their workload. The relationship between the patient and caregiver also plays a pivotal role, with a poor preexisting relationship as well as disagreements and conflicts within the family complicating care and decision making (Siminoff et al., Reference Siminoff, Rose and Zhang2006; Fried et al., Reference Fried, Bradley and Towle2003).
As the patient's illness becomes recurrent or terminal, caregivers are expected to provide increased care due to the decline in functional status and greater symptomatology. Caregivers have been found to report low quality of life and other health scores, even when compared to cancer patients (Weitzner et al., Reference Weitzner, Jacobsen and Wagner1999a , Reference Weitzner, McMillan and Jacobsen b ; Sales, Reference Sales1992; Cameron et al., Reference Cameron, Franche and Cheung2002; Glajchen, Reference Glajchen, Bruera and Portenoy2009; Kim & Given, Reference Kim and Given2008). Physical, psychosocial, and financial complications occur, including insomnia (Hearson & Clement, Reference Hearson and Clement2007), fatigue (Teel & Press, Reference Teel and Press1999), cardiovascular disease (von Kanel et al., Reference von Kanel, Mausbach and Patterson2008), increased mortality (Christakis & Allison, Reference Christakis and Allison2006), increased rates of anxiety and depression (Given et al., Reference Given, Wyatt and Given2004), and intensified feelings of fear and hopelessness (Dumont et al., Reference Dumont, Turgeon and Allard2006). In addition, due to the time-consuming aspect of caring for patients with advanced cancer, social isolation is likely to occur. Adequate social support, financial security, and work stability can offset this social impact (National Cancer Institute, 2015). Financial distress occurs due to accumulating copayments, insurance deductibles, travel costs, home care service expenses, and lost wages (National Cancer Institute, 2015; Glajchen, Reference Glajchen, Bruera and Portenoy2009; Yabroff et al., Reference Yabroff, Davis and Lamont2007). Early referral to palliative care has been shown to help improve the physical and psychological well-being of both patients and caregivers as well as enhance family satisfaction (Casarett et al., Reference Casarett, Pickard and Bailey2008; Gelfman et al., Reference Gelfman, Meier and Morrison2008). Spirituality has also been found to act as a safeguard against caregiver distress by helping caregivers gain a new perspective from the cancer experience and alleviate hopelessness, resulting in improved well-being (Ward et al., Reference Ward, Berry and Misiewicz1996; Kim et al., Reference Kim, Wellisch and Spillers2007).
Applebaum and Breitbart (Reference Applebaum and Breitbart2012) recently reviewed available interventions for informal caregivers, ranging from psychoeducation to complementary and alternative medicines. The majority of these interventions have shown benefits for informal caregivers. However, caregiver distress is still not monitored regularly, resulting in a need to address the availability of clinically applicable and effective caregiver tools that screen and detect any distress early on, which promotes more timely implementation of these interventions.
The objectives of our systematic review were: to determine the availability and adequacy of caregiver tools that were developed for cancer caregivers; to identify variations in the available cancer caregiver tools regarding different domains of distress, including such physical and psychosocial factors as caregiver needs, burden, and quality of life; and to identify the cancer caregiver tools that have been created for and utilized in the adult and pediatric populations.
METHODS
We conducted a literature search of all articles pertaining to tools for caregivers of cancer patients. During the initial search, restrictions were not placed in terms of language or publication type.
We employed the following databases to obtain relevant studies for this review (the original search was conducted up through October of 2013):
-
• MEDLINE (1946–present)
-
• EMBASE (1947–present)
-
• CINAHL (1937–present)
-
• Scopus (1996–present)
-
• PubMed ePubs (until October 2013)
We retrieved all articles through the strategy listed in Supplementary Appendix A for the MEDLINE search. The same strategy was adopted for the other databases. The methods and results section of each article were reviewed to determine if any reliability and validity tests had been conducted. In addition, we checked for references that might lead to further research on that tool to identify if reliability and validity were measured. See Tables 1–5 for details on the included studies.
*Self-report instruments with ≤20 items, positive reliability, and validity.
*Self-report instruments with ≤20 items, and positive reliability and validity.
**Healthcare provider a dministered instrument with ≤20 items, and positive reliability and validity.
*Self-report instruments with ≤20 items and positive reliability and validity.
*Self-report instruments with ≤20 items, and positive reliability and validity.
An article was included if the instrument cited was designed specifically for caregivers of cancer patients to ensure that the instrument was originally designed with cancer caregiver distress in mind. If an instrument was created for a different population, including cancer patients but not caregivers, it was excluded. Since we intended to capture the whole caregiver experience, there were no restrictions set on the area that a tool assessed. Since there may be differences in the patient–caregiver relationship in the pediatric and adult cancer populations, we included instruments that measured caregivers of both adult and pediatric cancer patients and divided the instruments by population within the various areas of the caregiving experience.
The authors carried out a study selection to determine that all the articles met our inclusion criteria. Any disagreements about a particular study were resolved by discussion and consensus. Articles that were not studies on caregivers of cancer patients were excluded. Articles about cancer caregivers but with no standardized tools reported were also excluded. Articles that were only published as conference abstracts, research protocols, or theses, with no corresponding published manuscript, were also excluded. Finally, we also excluded articles for which no English translation was available.
The data were then extracted into the following areas of focus: purpose of the tool; dimensions measured; population originally developed for; number of items; and psychometric data. In order to allow these tools to be implemented in a clinical setting and avoid excessive burden to caregivers and clinicians in completing them, our team chose a cutoff of 20 items. Table 6 summarizes the distribution of tools and their number of items.
Definition of Burden
The term “burden” has been employed to define the impact on clinical and psychosocial factors within families. In Given's study on the Caregiver Reaction Assessment Scale, “caregiver burden” was defined as the physical, psychological, social, emotional, and financial problems experienced by family members (including both objective and subjective factors) (Given et al., Reference Given, Given and Stommel1992; Gwyther & George, Reference Gwyther and George1986; Montgomery et al., Reference Montgomery, Gonyea and Hooyman1985).
Definition of Need
“Caregiver needs” have been defined as needs related to state of health and state of healthcare that are not being met (Hudson et al., Reference Hudson, Trauer and Graham2010; Hileman & Lackey, Reference Hileman and Lackey1990).
Definition of Quality of Life
The Centers for Disease Control and Prevention (2015) defines “quality of life” (QoL) as a subjective evaluation of both the positive and negative aspects of life. QoL instruments typically measure psychological, social, and physical well-being but may also include one's spiritual state.
Definition of Satisfaction with Care
“Satisfaction” has been defined as the fulfillment of a need or want (Merriam-Webster, 2015). Satisfaction with care has been determined to be a significant factor, as the patient and caregiver will go through multiple aspects of care through the cancer trajectory (diagnosis, treatments, recurrence, survivorship, end of life), as well as the various healthcare specialties to which the patient will be referred.
Definition of Other Aspects
“Other aspects” were defined as any other aspect in caregiving that did not fit exclusively into any of the previous categories, including emotions between caregivers and patients, caregiver comorbidity, and anticipatory grief.
RESULTS
The database search identified 5,541 articles. These were narrowed down to 135 articles (2.4%) based on our inclusion criteria. All 135 were extensively discussed by the authors, and a consensus was reached. From these we extracted 59 instruments that measured different dimensions of cancer caregiving distress. Instruments that had similar names to other tools were labeled according to their country/region of origin. The abbreviated names of the instruments were also listed. The PRISMA diagram is shown in Supplementary Appendix B.
The instruments were divided into the following categories: burden (n = 26, 44%); needs (n = 14, 23.7%); quality of life ([QoL] n = 9, 15.3%); satisfaction with healthcare delivery (n = 5, 8.5%); and various other issues (n = 5; 8.5%). Among the 59 instruments, 45 (74.6%) measured caregiver distress in adult cancer patients and 15 (25.4%) in caregivers of pediatric patients. Forty-one of the (69.5%) instruments were found to be reliable and valid. The list of instruments and associated information are presented in Tables 1–5.
The median number of items was 29 (4–125). Twenty instruments (33.9%) had ≤20 items. There were 9 instruments that had ≥50 items (15.3%). The breakdown of number of items is shown in Table 6. Thirteen instruments (22%) had ≤20 items and were valid and reliable, 12 (92.3%) of which were self-report instruments. These instruments are denoted in bold and shown first in Tables 1–5.
Of the caregiver burden instruments, 9 of the 26 (34.6%) had ≤20 items, with 5 (55.6%) being reliable, valid, and self-report. Eighteen (69.2%) of the tools measured caregiver burden in adult patients. Of the caregiver needs instruments, 3 of 14 (21.4%) were reliable and valid, and had ≤20 items, with 1 of these 3 (33.3%) requiring administration by healthcare providers. Twelve of the 14 (85.7%) needs instruments measured the caregiver needs of adult patients. Of the instruments measuring satisfaction with care, 2 of 5 (40%) were self-report, reliable, and valid and had ≤20 items. All instruments for caregiver satisfaction were for caregivers of adult patients. Of the caregiver QoL instruments, 3 of 9 (33.3%) were self-report, reliable, and valid and had ≤20 items. Seven of the 9 (77.8%) measured caregiver QoL in adult patients.
DISCUSSION
Our systematic review of the literature identified 135 of 5,541 articles that met our inclusion criteria, with 59 instruments found to be developed initially for caregivers of cancer patients. Some of these instruments were mentioned in multiple articles, thereby explaining the larger number of articles than instruments.
As a general comment, we found it difficult to document whether a particular tool was developed for research or clinical purposes. Instead, we included the number of items in each instrument to indirectly illustrate if the tool was too complex or time-consuming to complete during a regular healthcare visit. Among all instruments, about 20% were reliable, valid, had ≤20 items, and were self-report questionnaires. We believe this is an encouraging number of instruments that are psychometrically sound and, more importantly, that can be applied effectively in a clinical setting. Among these instruments, McMillan's Caregiver Quality of Life Index appeared to be the simplest questionnaire, with four items measuring a caregiver's physical, emotional, social, and financial well-being (McMillan & Mahon, Reference McMillan and Mahon1994). Although interesting for research purposes in understanding the complexities of caregiver distress, lengthy instruments can place a heavy burden on already-stressed caregivers and, likewise, may increase the burden on clinicians in trying to apply these instruments in a busy clinical setting. Self-report instruments may also help decrease the burden on busy healthcare providers, as caregivers are able to complete them without taking time away from providers.
Our review demonstrated a considerable number of instruments measuring caregiver burden (n = 26) and caregiver needs (n = 14). While Deeken et al. (Reference Deeken, Taylor and Mangan2003), in their review of self-report instruments for informal caregivers, appropriately identified the disproportion of quality-of-life tools between cancer patients and caregivers, our search was actually able to identify a higher number of caregiver QoL tools compared to previous literature reviews (Hudson et al., Reference Hudson, Trauer and Graham2010; Deeken et al., Reference Deeken, Taylor and Mangan2003). On the contrary, there is a dearth of instruments measuring family satisfaction with care, grief, and bereavement.
We found that there were 44 of 59 (74.6%) adult patient-based cancer caregiver instruments compared to 15 (25.4%) pediatric patient-based cancer caregiver instruments, which is not surprising when compared to overall cancer statistics (Murphy et al., Reference Murphy, Xu and Kochanek2013; Howlader et al., Reference Howlader, Noone and Krapcho2013; NIH, 2015). Although a statistical analysis suggests that cancer is overwhelmingly a disease of adults, the diagnosis of cancer in a child puts enormous burden on caregivers, so that assessment tools are of great importance in that population (Sargent et al., Reference Sargent, Sahler and Roghmann1995; Woodgate, Reference Woodgate2006; Patterson et al., Reference Patterson, Holm and Gurney2004). Furthermore, there are very specific issues in pediatric caregiving, including the effects of advanced disease on children and adolescents, and those need to be addressed in future research.
One of the limitations of the study is that we did not address tools that were not designed initially for cancer caregivers but can and are being used currently for that population. We believe that the level of complexity of caring for cancer patients may need specifically designed tools to capture their level of distress. There were several valuable tools that were excluded from the final tally. These tools were initially developed for other diseases or for patients, and were then eventually found to be applicable for caregivers as well. Some of these tools included the Zarit Burden Interview, developed for caregivers of dementia patients (Zarit et al., Reference Zarit, Reever and Bach-Peterson1980); the Distress Thermometer, initially utilized with prostate cancer patients (Roth et al., Reference Roth, Kornblith and Batel-Copel1998); and the Family Relationships Index, initially used in cancer patients and was part of the Family Environment Scale (Edwards & Clarke, Reference Edwards and Clarke2005). Although not meeting the inclusion criteria for our review, these tools should not be neglected and are particularly valuable in detecting distress for both caregivers and patients with different disease processes.
Another limitation is the choice of 20 items as a cutoff for the instruments, which we based on our clinical experience in using other regular clinical assessment tools, including the Edmonton Symptom Assessment System (ESAS) and the Memorial Delirium Assessment Scale (MDAS). It might be possible that certain lengthier instruments would also be feasible and time-efficient in the clinical setting.
As compared to the reviews conducted by Deeken et al. (Reference Deeken, Taylor and Mangan2003), Hudson et al. (Reference Hudson, Trauer and Graham2010), and Applebaum and Breitbart (Reference Applebaum and Breitbart2012), our research supports their findings and also adds more instruments to the list that can be utilized for caregivers of adults and pediatric cancer patients. We also confirmed that there is a significant number of tools that have ≤20 items and that can be applied clinically. The challenge for the future is how to implement them regularly, improving caregiver care and measuring outcomes in the adaptation of these tools. Since there is no gold standard or real comparison between the different tools, clinical teams will have to determine which of these tools are more suitable to their practice. Ideally, in the future, receiving caregiver feedback will help identify the most effective.
Our review of tools has demonstrated that very few instruments have been incorporated into regular clinical practice. In addition, there are no studies documenting how the use of these tools helps modify clinical and/or health service utilization outcomes. There is also a lack of data on the comparative clinical performance of these instruments. These are all areas where research is needed.
We have found that there are several self-report cancer caregiver distress instruments that are concise, reliable, and valid. Clinical utilization and outcomes measures are certainly important areas for future research.
SUPPLEMENTARY MATERIALS AND METHODS
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1478951516000079.