Introduction

An important revolution in medical practice and research started since the World Health Organization defined health as a “state of complete physical, mental and social well-being and not merely the absence of disease or infirmity” [1]: The biomedical approach was superseded by a bio-psycho-social one [2], the notion of a “patient-centered” approach began [3], and, with the advent of the concepts of quality of life (QoL) and health-related quality of life (Hr-QoL), relative assessment tools were developed. Thus, a widening of perspective took place: Attention shifted from the mere focus on the body and its organic affection processes to the patients and to the repercussions on patients' whole being.

Along with the need of holistically assessing patients' needs, the demand for reporting both the “psycho” and “social” components of the disease experience is constantly growing. This demand has, in turn, produced the remarkable number of different assessment tools (mostly questionnaires, inventories, and scales) available today in literature [4, 5].

In Table 1, the principal QoL assessment tools in oncology that dedicate a subscale to social aspects are listed as an example. Each of the 14 identified instruments [622] seems to conceptualize the social component of experiencing cancer in a different way, referring to the social or functioning role (e.g., EORTC Quality of Life Core 30 [9]), to social support (e.g., Psychological Screen for Cancer [18]), to partner relationships (e.g., Brief Cancer Related Worry Inventory [6]), to relatives (e.g., Functional Assessment of Cancer Therapy—General Measure [10]), and to social dysfunction (e.g., General Health Questionnaire [11]), and dedicate a variable number of items to investigating this domain (from one-twelfth to half of the entire set).

Table 1 Psychological tools commonly used in oncology containing dimensions with social content items

In this way, considering only the social dimensions included in the QoL instruments used most in oncology, a broad and complex vision can be obtained of what can generate, moderate, and describe the social impact of being ill in oncological patients. At the same time, these instruments allow us to appreciate the role of the social dimension in the global quality of life for oncological patients, as well as their association with more physical, functional, and psycho-emotional aspects.

In light of the contents of the social subscales within the cited QoL instruments used in oncology, this work defines modifications of the patients' interpersonal and social relationships, as well as changes in social needs, social resources, and social difficulties related to the oncological experience as the “social impact” of cancer. Considering that a multidimensional QoL instrument cannot be thoroughly exhaustive of all of this variety and complexity, it might be necessary to utilize a more in-depth and complex instrument.

This paper aims to review questionnaires that were specifically designed or adapted to investigate the social impact of cancer in oncology. On one hand, this work will further an understanding of how this domain may be theorized and, subsequently, which practical and daily aspects of personal experience are employed to build it. On the other hand, it will present an overview of instruments (described both in terms of psychometrics and of content) so as to facilitate an informed choice about what type of knowledge objectives to pursue, be it in a clinical or research setting. The focus is on the oncological patient, not on caregivers or relatives, because we planned to address the social impact of cancer from the perspective of those who experience it in person.

Method

Data sources and search

A computer-based literature search was performed to identify articles about tools used in assessing the social component of patients' cancer experience published between January 2000 and June 2011. MedLine and PsycINFO databases were chosen for this search, as they both contain research publications across a wide range of health professions in the field of oncology, and they are a leading source for journals related to clinical and psycho-social cancer care. The search terms that we used refer to three separate conceptual areas that were intermixed during research: terms regarding the “social” domain (i.e., communication, couple, dyadic relationship, employment, family, interpersonal problem, professional relationship, public image, relationship, self-perceived burden, social activity, social adjustment, social concern, social contentment, social difficulty, social discomfort, social environment, social exclusion, social experience, social factor, social function, social functioning, social impact, social interaction, social morbidity, social network, social problem, social support, social well-being, stigmatization, support), terms regarding the oncological disease (i.e., cancer, neoplasm, oncology), and terms regarding the creation/validation/adaptation of measurement (i.e., inventory, measure, measurement, psychometric property, psychometrics, questionnaire, reliability, scale, self-assessment, self-rating scale, tool, validity). While the terms that are associated with the last two conceptual areas are more intuitive and shared, in order to identify those terms that are associated with the social component of the disease experience, we turned to the contents of the items included in the social subscales included in the QoL instruments listed in Table 1.

The search was restricted to articles written in English or Italian on human adult subjects.

After identifying and classifying the relevant publications, a database search by tool name was performed. Finally, a manual search of the reference lists of the selected papers was performed to find any further relevant literature.

Data extraction

The papers that we identified were categorized according to which instrument they refer to. For each instrument, we then identified the type of oncological population it was validated on, the potential other contexts it was used in, what dimensions were studied, validation data (construct, concurrent, criterion, predictive, external, cross-cultural), reliability (internal consistency and temporal stability), and feasibility (compilation time, availability of the instrument, or other associated materials). Furthermore, for each instrument, we showed the social referrals it points to (the partner, caregiver, or close other; family; friends; the community, including peers; professional helpers). We also showed the topics, organized in three broad categories: social forms (regarding relationship quality, attachment, communication, and the patient's social functioning), social resources (like for instance social support), and social difficulties/issues. Lastly, we identified which instruments are utilized for one or more cancer types or whether they are specific to one of the two genders.

Each of the two authors has independently classified the instruments across the five social referral categories, the three topic categories, and the two specificity (type of cancer, gender) categories. The respective results were then compared (K agreement test, p < 0.01), and if necessary, an agreement was reached through discussion.

Results

The initial search (including manually retrieved articles) yielded 469 articles. After excluding articles on the basis of the aforementioned criteria and duplications, we considered 27 articles [2349] dealing with 14 different tools: Cancer Communication Assessment Tool for Patients and Families (CCAT-PF), Cancer Experience and Efficacy Scale (CEES), Chinese Social Adjustment Scale (Ch-SAS), Couples' Illness Communication Scale (CICS), Cancer Perceived Agents of Social Support (CPASS), Cancer Related Communication Problems (CRCP), Experiences in Close Relationships Scale (modified and brief, ECR-M16), Experiences in Close Relationships Scale (modified and brief, ECR-M36), Prostate Cancer Peer Support Inventory (PCPSI), Social Difficulties Inventory (SDI), Social Difficulties Inventory 16 (SDI-16), Self-perceived Burden Scale (SPBS), Social Relational Quality Scale (SRQS), and Social Support Network Inventory (SSNI). Although ECR-M16 and ECR-M36 are abridged versions of the same tool (I.E. ECR), we considered them separately because psychometric data have been provided separately for each instrument. For the same reason, we considered SDI and SDI-16 as two distinct tools, despite SDI-16 being the abridged version of SDI (SDI21).

Of the 27 selected papers, 17 (63 %) refer to psychometric properties of the tools in oncological settings, and 10 (37 %, references in brackets in column 2 of Table 2) refer to the past or future of the tool.

Table 2 Tools assessing social impact of patients' cancer experience: psychometric properties

Psychometric properties in oncology of the identified tools

Table 2 lists the 14 identified tools and summarizes their psychometric properties.

Ten tools (71.4 %) were specifically designed for cancer patients, whereas four (28.6 %) were adapted for this kind of users from general population (i.e., ECR-M16, ECR-M36, SSNI) or other clinical domains (i.e., SPSB).

Ten tools (71.4 %) were tested with a sample of patients with homogenous cancer types, whereas for four tools, heterogeneous samples of cancer diagnosis were enrolled (i.e., CICS, SDI, SDI-16, SPSB).

Nine (64.3 %) tools were tested in Western or English-speaking countries, three in China (CEES, Ch-SAS, SRQS), one in Brazil (SSNI), and one in Israel (CPASS), but only for SDI data on cross-cultural validation for three South Asian languages were provided. However, the paper on SSNI presents an adaptation of the tool for Brazilian cancer patients.

The majority of tools (N = 9; 64.3 %) are multidimensional, involving a number of factors between 2 and 6 and a number of items between 12 and 36; for the five one-dimensional tools (CCAT-PF, CICS, SDI, SPSB, SSNI), the item range is 4–21.

Data on internal consistency were provided for all tools: Consistency ranges from acceptable to good for most of them, except CCAT-PF and the “Negative peer support” subscale of PCPSI. Conversely, temporal stability was reported only for five tools including CCAT-PF, which does not reach the conventional standard of acceptability for temporal stability.

The construct validity was assessed for 11 tools (78.6 %) by means of confirmatory factor analysis (two tools), Rasch analysis (one tool), exploratory factor analysis (seven tools), and in concordance with a previous version of the tool (one tool). Concurrent validity data were provided for all tools except PCPSI, SDI, and SDI-16 using mainly QoL, emotional distress, and social and/or familial functioning scales. However, for SDI, criterion validity was assessed by comparing self-report SDI scores with patient evaluations by social workers, and for both SDI and SDI-16, indications for the minimal clinical meaningful differences were provided. Predictive validity was provided only for CICS, reporting it on the emotional representation of illness after 3 months. Finally, external validity data were provided for CICS, CPASS, and CRCP (17.6 % of all tools), while cross-validation information was provided for SDI only.

For ten tools (71.4 %), an item list and answer format were provided in the appendix or in tables. However, since SDI-16, SPSBI, and SSNI are adaptations of previously published tools, their items can be drawn from the original papers; for SRQS only, the item content is available in the paper. Despite the fact that these are short questionnaires, no explicit indications were provided for their completion time except for SPSB.

Qualitative features of the identified tools

Table 3 summarizes some of the qualitative features as they emerged from the comparisons from the two independent judges.

Table 3 Tools assessing social impact of patients' cancer experience: qualitative features

Regarding the main social referrals, we found that the partners, caregivers, or closer others were the most represented category (11 instruments), while the professional helpers were the least represented (two instruments).

Regarding the social topic—meaning the social aspect explored by the instrument in non-mutually exclusive categories—the descriptors, the resources, and the difficulties/issues are represented quite equally (five, five, and four instruments, respectively). This evaluation was not possible for the SQRS because single items were not published.

Lastly, even though only four of the tools were tested on heterogeneous cancer-type samples, our content analysis reveals that only the PCPSI (prostate tumor) cannot be extended to persons with a different cancer type than the study sample. This evaluation was not possible for the SRQS.

Similarly, regardless of the samples in the original studies, 13 of the 14 tools seem to be appropriate for both female and male oncological patients—here the PCPSI is the exception.

Discussion

Attention to the socio-psychological aspects of disease and its treatments—a direct consequence of the WHO definition of health as global well-being [1]—and of the widening of focus of the medical intervention from the organic disease to the patient is constantly advocated by the principal international agencies dealing with health issues (i.e., National Cancer Control Network, National Breast Cancer Centre). At the same time, QoL measures constitute an endpoint in numerous randomized clinical studies [5, 50].

The present work intends to investigate the meaning of the “social impact of cancer” by examining the instruments created for the assessment of the social impact of cancer to better understand how this domain was operationalized in the different assessment tools and what experiences, situations, and concepts were employed by different researchers to flesh out this domain.

We first considered that, in oncology, many psychological assessment tools dedicate some items to the “social” component, even though they refer to different aspects. A systematic review of the literature allowed us to identify 14 instruments specifically designed for the assessment of the social aspects of the oncological experience (or adapted to oncological populations) for a total of 27 articles. The first observation that we made was that, despite the high number of instruments, they seem to be specially designed for that study and they were seldom employed in subsequent studies. Consequently, only for a few tools, the validation process went further than verifying internal consistency as well as construct and concurrent validity. In other words, the quantity of the identified instruments seems to indicate the need for a specific tool for evaluating these components of the oncologic experience, but in the literature, there does not seem to be a tool of choice, considering that a whole validation process to assess all of the principal psychometric properties has not been undertaken yet for any of these instruments (except for SDI and SDI-16). Nevertheless, some cultural aspects (such as the tendency to collectivism rather than individualism) may make a cross-cultural or cross-national validation difficult, just like the total or partial visibility of a mutilation, or of a disfiguring scar may impair the external validity with respect to different oncological or other pathologies or to different stages of the disease.

We then examined the contents considered by the 14 identified instruments. When evaluating the identified factors together with the single component items, we noticed the remarkable heterogeneity of topics and of social referrals, which range from partners, to relatives, to friends, and to the community. None of the identified instruments seems to be comprehensive of all of the social targets presented in the sum of all of the other instruments. In other words, we cannot define one tool to be more complete than the others, even if we would simply consider the completeness of contents. However, the identified instruments, even though they were usually tested on homogenous samples of oncological patients, can generally be easily utilized on samples which are more heterogeneous in diagnosis and type.

In keeping with these considerations (but not forgetting the variety of criteria which should be considered when choosing an assessment tool, like the goal it is intended for, the amount of time available to perform the assessment, psychometric properties, and validation in one's own cultural context), it appears that the CICS, SDI, SDI-16, SPBS, and the SSNI are more appropriate than others, both because of the psychometrics they test up until now (and, in this context, the SSNI seems the weakest, even though we must remember that it is already in use in non-oncological settings), as well as because of their content. In particular, the CICS is an ultra-short (number of items = 4) tool that evaluates illness-related couple communication. Despite the fact that this tool was validated only on a sample of patients with ovarian cancer, external validity data (tested with partners and multiple sclerosis patients) are available, suggesting that the CICS could be used with oncological patients with other types of cancer. The SDI and SDI-16 are nimble lists of potential social difficulties that cancer patients could face during their illness. The multipoint validation process that these tools have been subjected to has made it possible to define the cut-off past which patients might benefit from discussion of social difficulties together with the minimal clinical meaningful differences; consequently, their use is particularly informative on both the research and clinical level. The SPBS (nine items) records the self-perceived burden, a construct that has as of yet been little studied, but which seems to connote the social relationships of severely ill patients, as was documented in previous validations of this instrument with hemodialysis and ALS patients. Lastly, the SSNI evaluates perceptions of social support—the main resource in times of trouble—regarding availability, reciprocity, and nature (emotional, practical, support in relation to illness). Considering that they are short and that the subjects and social referrals are complementary, these tests can also be used simultaneously to comprehensively record the social impact of the oncological experience.

In summary, it is possible that the great interest of the international psycho-oncology literature on this subject shows the importance of having a valid and reliable tool to assess the social impact of the oncological experience, but at the same time, further efforts are needed to investigate the psychometric properties of the instruments already in existence. Moreover, research on the social impact of cancer should probably consider employing one or more of these tools, as they are complementary to each other but do not overlap.

Considering the various social topics—forms, resources, difficulties, and worries related to being a “social person”—different potential social referrals taken into consideration by the identified tools could be the first step for the development of a more complete instrument.

Talking about the social impact of cancer also means considering the repercussions of the disease on the patient's partner, relatives, caregivers, friends, and community. In this work, we decided to limit the search of the instruments to those focused on the oncological patient, but other social entities certainly also need further investigation. A further limitation is connected to the fact that, in this paper, we only analyzed instruments for which validation data were available in the oncological context. Many other instruments that were not specially created could prove just as useful. Finally, since we restricted data searching to MedLine and PsycINFO, it is possible that extending the search to other databases (e.g., Embase, Cinahl) would yield additional pertinent papers. Notwithstanding these limitations, this paper is the first attempt that we are aware of to group several instruments that are being used in the oncological field with the goal of measuring the social impact of cancer—regardless of which facet of this broad concept is actually being investigated.