Brief Report: Children's Response to Serious Illness: Perceptions of Benefit and Burden in a Pediatric Cancer Population

Abstract

Objective To examine a revised measure of benefit finding for children, in relation to perceptions of illness-related burden, and other measures of child psychological functioning. Methods A sample of 78 children with cancer completed the newly revised Benefit/Burden Scale for Children (BBSC) and measures of optimism/pessimism, positive/negative affect, anxiety and defensiveness. Results Factor analysis of the BBSC revealed a clear two-factor solution, with benefit finding and illness-related burden representing orthogonal factors. Both scales were internally consistent and demonstrated different patterns of correlation with the other measures assessed in the study. Conclusion The BBSC is internally reliable and preliminary data supports the validity of separate benefit and burden constructs. Children report positive and negative aspects of their illness simultaneously, and perceptions of benefit and burden function as independent constructs. The BBSC is a useful measure for pediatric cancer patients that could be applied to children experiencing other significant life events.

Introduction

Over the last decade, research has begun to focus on the role of benefit finding in adaptation to highly stressful experiences (Helgeson, Reynolds, & Tomich, 2006). Although the onset of a serious illness or another potentially traumatic event can precipitate a number of debilitating psychological symptoms, these experiences can also serve as catalysts for personal growth (Tedeschi & Calhoun, 2004). Perhaps due to the frequent life-threatening implications, cancer has received considerable attention as an experience that may be traumatic for patients but also foster positive life changes (Kangas, Henry, & Bryant, 2002). Cancer patients often report finding benefit or paradoxically uncovering a “silver lining” in the personal or social consequences of their illness. For example, they may experience enhanced empathy, reordered life priorities, a closer connection to other people, or increased emotional strength resulting from the various challenges associated with cancer (Antoni et al., 2001; Carver & Antoni, 2004; Cordova, Cunningham, Carlson, & Andrykowski, 2001; Tomich & Hegelson, 2004). Importantly, the manner in which patients perceive positive or negative implications of their illness may influence current and future psychological adjustment. Despite inconsistent findings in the literature (Helgeson et al., 2006), research has indeed shown that positive changes stemming from cancer can predict improved functioning up to eight years following the illness (Carver & Antoni, 2004).

Research on the role of benefit finding in adaptation to cancer has focused largely on adult patients, with only a few pediatric studies (Barakat, Alderfer, & Kazak, 2006; Phipps, Long, & Ogden, 2007). In addition, research has generally not explored the possibility of perceiving positive consequences but simultaneously feeling psychologically burdened as a result of the illness. That is, some patients may recognize illness-related growth but also experience ongoing negative consequences from having cancer, which for children may include feeling insecure in their appearance, concerns about worrying their parents, or no longer having as much time to interact with friends. Consistent with the prevailing view that positive and negative emotion are not opposite ends of a single continuum but represent relatively distinct domains (e.g., Salovey, Rothman, Detweiler, & Steward, 2000; Watson & Clark, 1997; Watson, Clark, & Tellegen, 1988), benefit finding and perceived burden may each have a unique influence on adaptation to cancer or other potentially traumatic experiences. For instance, finding benefit in one's illness or other traumatic stressor may have a buffering effect, preventing or reducing distress, whereas perception of event-related burden may increase vulnerability to distress.

In a prior investigation, Phipps, Long, and Ogden (2007) developed the Benefit Finding Scale for Children (BFSC). In a childhood cancer sample, the BFSC was shown to be a unidimensional measure with excellent internal consistency. There was also evidence of construct validity in that BFCS scores correlated positively with optimism and self-esteem and negatively with anxiety (Phipps et al., 2007). Although the initial report of the use of the BFSC suggested its usefulness as a measure of benefit finding in children, the instrument contained only positively worded items, an approach that may encourage a socially desirable response set that could artificially inflate estimates of benefit. It has been suggested that including both positive and negative effects of the potentially traumatic event within the same instrument may provide a more valid measure of benefit finding (Tomich & Hegelson, 2004). To address this, the BFSC was revised to include additional items reflecting negative aspects, or perceived burden associated with the cancer experience. We present the preliminary psychometric properties of the revised measure and examine associations of benefit finding and perceived burden with other relevant outcomes. We hypothesized that perception of benefit and burden would represent relatively distinct constructs (as opposed to opposite ends of a single continuum), but would be moderately inversely correlated with each other.

Method

Participants

Children and their parents were recruited from outpatient clinics at a major children's oncology center. Children were eligible to participate if they were between the ages of 8 and 18 years, spoke English as their primary language, had a primary diagnosis of malignancy, were between 3 months and 5 years from time of diagnosis, and did not have any significant cognitive or sensory impairments that would preclude their completion of study measures. Of 107 parent–child dyads approached, 80 (75%) agreed to participate in the study. Two children failed to complete all study measures, leaving a sample of 78 evaluable participants.

The mean age in the sample was 12.9 years (SD = 3.2), and 48.7% were male. The sample was 75.7% Caucasian; 19.2% African American, and 5.2% of other ethnicities. There was broad distribution across socioeconomic strata, with 13% in SES group I (i.e., highest SES); II = 36%; III = 32%; and IV and V (lowest SES) = 19% (Barratt, 2006). Nearly two thirds of the sample (61.6%) had completed treatment, and the remainder was still on active therapy. Diagnostic breakdown included leukemia (28.2%), lymphomas (20.5%), non-CNS solid tumors (34.6%), and brain tumors (16.7%).

Measures

The Benefit and Burden Scale for Children (BBSC). This 20-item self-report instrument was developed for use in this study. The benefit finding items were taken from the Benefit Finding Scale for Children (BFSC, Phipps et al., 2007). In order to avoid a socially desirable response set, the revised measure included negatively worded items interspersed with the benefit items. Rather than simply inserting negative items as “filler”, an attempt was made to create a meaningful set of items reflecting relevant negative aspects of the cancer experience. A group of pediatric psychologists and trainees in clinical psychology developed a list of items to assess a range of areas that may be influenced negatively by illness, including affect (e.g. “Has made me more moody and irritable”), relationships with peers (e.g., “I have less time to be with my friends than I used to”), and family relationships (e.g., “I worry about being a burden to my family”), and sought consensus on those most relevant and appropriate. Each item on the resulting BBSC depicted a potential benefit or burden of illness that could be answered on a 5-point Likert scale ranging from “not at all true for me” to “very true for me”. The piloted version contained 10 a priori benefit items and 10 a priori burden items.

The Youth Life Orientation Test (YLOT; Ey et al., 2005). The YLOT is a measure of dispositional optimism developed as a child analogue of the widely used Life Orientation Test (Scheier & Carver, 1985). The 16-item measure (with 2 filler items) contains seven optimism and seven pessimism items. In the current sample, levels of internal consistency (α) were .722 and .693 for optimism and pessimism, respectively.

Positive and Negative Affect Schedule for Children (PANAS-C; Crook, Beaver, & Bell, 1998). The original PANAS was developed for adults by Watson, Clark and Tellegen (1988). The PANAS contained two 10-item scales, one measuring positive affect that other measuring negative affect. As described by Crook et al. (1998), the version of the PANAS-C used in this study was comprised of 20 items generating two 10-items subscales for positive and negative affect. In the current sample, positive affect items (α = .908) and negative affect items (α = .891) each showed strong internal consistency.

The State-Trait Anxiety Inventory for Children (STAIC; Speilberger, 1973). This is a widely used measure of anxiety symptoms in children. Only the trait scale was administered in the current study. It consists of 20 anxiety-related statements that are responded to on a 3-point scale (hardly ever, sometimes, often). Internal consistency (α) was .840 in this sample.

Children's Social Desirability Scale (CSD, Crandall, Crandall, & Katkovsky, 1965; Phipps & Srivastava, 1997). The CSD, a measure of social desirability, consists of items representing behaviors and attitudes that are socially desirable but improbable (e.g., “I always do as I am told”; “I never tell a lie”). Items are answered in a yes/no format. We utilized a revised 25-item version that has been used previously in populations of children with cancer (Phipps & Srivastava, 1997). The internal reliability (α) of this version was .799 in the current study.

Procedure

The study was approved the hospital's IRB, and conducted in accordance with institutional and APA guidelines. Informed consent was obtained from all parents and adult participants and assent from minor child participants. After obtaining consent, children completed self-report measures in a research laboratory in the hospital's Behavioral Medicine clinic. Research assistants were nearby to help participants who had difficulty with any of the measures, but most patients were able to complete the measures on their own.

Statistical Approach

The 20 items of the BBSC were first factor-analyzed using a principal components analysis with varimax rotation, an approach recommended for an empirical summary of the data set (Tabachnick & Fidell, 2007). Descriptive statistics were used to characterize the resulting subscales. T-tests and analysis of variance were used to examine differences in benefit and burden across demographic and illness-related variables, and Pearson product moment correlations were used to examine the relationship between the BBSC scales and the continuous psychological variables.

Results

As presented in Table I, factor analysis revealed a clear two-factor solution, representing the two a priori subscales for perceptions of benefit (eigenvalue = 4.91) and burden (eigenvalue = 3.69) related to the cancer experience. Cumulatively, the two factors accounted for 43% of the total variance in BBSC scores. The two 10-item subscales were also not correlated with one another [r (76) = −.10, p = .41), and each subscale demonstrated excellent internal reliability (α = .852 for benefit items and α = .804 for burden items). Descriptively, the mean and standard deviation of the benefit scale (M = 37.0; SD = 8.6) was nearly identical to that reported for BFSC (M = 37.4; SD = 7.8; Phipps et al., 2007). As was the case with the BFSC, the benefit scale of the BBSC was reasonably normally distributed with a mild negative skew. The burden scale showed a stronger positive skew.

Table l.

Factor Loadings for the BBSC

Item	Loadings on Factor 1 (perceived benefit)	Loadings on Factor 2 (perceived burden)
I am not able to enjoy myself the way I used to	−.08	.68
Has helped me become a stronger person	.60	−.19
Has made me less happy with my life	−.08	.73
Has helped me learn who my real friends are	.61	.03
I worry about being a burden to my family	.15	.40
Has helped me know how much I am loved	.52	−.09
Has helped me make some new best friends	.42	.16
Has made me feel embarrassed about being seen in public	.03	.40
I have less time to do fun things, or just play around	−.09	.49
Has helped me learn to deal better with my problems	.59	−.35
Has taught me to be more patient	.51	.08
I have less time to be with my friends than I used to	−.15	.45
Has taught me to be more loving to others	.77	.07
Has made me more moody and irritable	.17	.47
Has brought my family closer together	.55	−.01
Has made me less hopeful about my life	−.2	.59
Has taught me what is really important in life	.86	−.06
I worry that I will bring other people down or upset them	.08	.68
Has taught me to be happy and enjoy the good things when they happen	.65	−.18
I have less confidence in myself	−.14	.64
Mean (SD)	37.0 (8.6)	21.0 (7.7)
Median (range)	39 (17–50)	20 (10–45)
Coefficient α	.850	.804

Note. Loadings on the primary factor in bold.

Analyses did not show that children's benefit finding scores differed as a function of gender, race/ethnicity, socioeconomic status, diagnostic category, or treatment status, i.e., on versus off therapy (all p values >.2). Likewise there was not a significant correlation of benefit scores with time elapsed since diagnosis. However, there was some suggestion that older children in the sample had a greater likelihood of finding benefit than younger children, r(76) = .24, p = .04. Perceptions of illness-related burden did not differ by any of the demographic or medical factors assessed in the study (all p values >.2).

Correlation analyses between benefit finding, burden and the other psychological functioning variables yielded unique patterns of results (see Table II), further supporting the independence of the two domains. Perceptions of illness-related benefits were positively associated with optimism and demonstrated a negative correlation with pessimism. In contrast, perceived burden was associated with less optimism and higher degrees of pessimism among the children. Benefit finding was also associated with greater positive affect but did not yield significant relations with negative affect or anxiety. However, children who endorsed higher levels of burden experienced significantly less positive affect, more negative affect, and worse anxiety symptoms. Neither benefit nor burden scales were significantly related to the measure of social desirability.

Table ll.

Correlations Between Perceived Benefit and Burden and Other Variables

	Benefit	Burden
Benefit	–	−.10
YLOT optimism	0.41***	−.46***
YLOT pessimism	−.27*	.46***
PANAS-C positive affect	.62***	−.33**
PANAS-C negative affect	−.05	.61***
STAIC	−.04	.51***
CSD	.18	−.15

*p < .05; **p < .01; ***p < .001.

Discussion

Researchers have recently shifted from exclusively focusing on the negative aspects of highly stressful events to ways in which these experiences promote positive life changes. The majority of these efforts have not studied children or incorporated benefit finding measures that account for both positive and negative effects of the stressor. As a way of addressing these gaps in the literature, we revised the BFSC (Phipps et al., 2007) to assess perceptions of both benefit and burden relating to the potentially traumatic stressor, and with the aim of reducing possible response bias incurred when a measure includes only positively worded items. In a sample of child cancer patients, this instrument—the Benefit and Burden Scale for Children (BBSC)—loaded on two clear factors for the a priori dimensions of perceived benefit and burden. The two ten-item subscales each demonstrated excellent internal consistency, and outside of a small but significant positive correlation between benefit finding scores and child age, perceptions of benefit and burden did not differ as a function of demographic or medical factors. Children's reports of the positive and negative consequences of cancer were also uncorrelated, which suggests that these constructs may be distinct constructs and that many patients who perceived favorable changes were also experiencing harmful personal and social effects from their illness. Interestingly, the inclusion of burden items did not appear to impact children's report of benefit, as the mean and standard deviation of the benefit scale of the BBSC were nearly identical to those of the BFSC.

The apparent independence of benefit finding and children's perceptions of burden was further reflected in the correlations to personality and distress measures. Benefit finding related positively to optimism and positive affect, and had a negative association with pessimism. Consistent with other work (Antoni et al., 2001, Cordova et al., 2001, Phipps et al., 2007), benefit finding was unrelated to measures of psychological distress. However, in contrast to benefit, perceived burden was significantly associated with distress. Children who were feeling more burdened from cancer had significantly more negative affect and anxiety. Also, whereas perceived benefit correlated positively with other measures of positive psychological function (optimism, positive affect), perceived burden was associated with less of these qualities among the patients. Contrary to the benefit finding subscale, children who endorsed greater burden were also more pessimistic in their outlook on life.

These results highlight the concept that personal growth and other positive changes following highly stressful events do not reflect an absence of negative consequences or distress. Even in circumstances when children experience benefit from their illness or other trauma, they may feel an ongoing sense of burden and struggle to maintain healthy levels of functioning. It also appears that the level of perceived burden may be a better predictor of distress than the amount of benefit a child finds in his or her experience. Conversely, but consistent with broad trends in the literature (Helgeson et al., 2006), the strongest correlation was between benefit and positive affect, which suggests that benefit finding may reflect a positive outcome in its own right and not a mere lack of distress. Similar to the present view regarding the distinctness of positive and negative emotion (Salovey et al., 2000; Watson & Clark, 1997; Watson et al., 1988), the current findings indicate that benefit finding and burden capture different aspects of a child's response to a potentially traumatic experience. From a clinical standpoint, each of these constructs should be considered when examining children's level of psychological adjustment following stressful life events such as chronic illness.

There are limitations to this preliminary study that should be noted. First, we focused on children with cancer, which may limit generalization to other populations of children experiencing illness, trauma or significant life-events. Second, our results are based on cross-sectional data. Hence, we do not know whether benefit finding caused positive affect or whether burden led to more distress. These outcomes may have been antecedent conditions under which patients were more or less likely to perceive benefit or burden. The sample size was relatively small for factor analysis, and confirmatory factor analysis with larger samples is called for. Our population included a small number of minority participants, particularly Hispanic children. Future research should attempt to address potential differences in benefit and burden as a function of race/ethnicity. Future research might also consider including measures of coping response in attempt to disentangle the issue of whether benefit finding reflects a coping process or and independent outcome. In particular, it will be interesting to examine the relationship of benefit and burden to measures of emotion-focused coping and positive reappraisal, constructs that appear to share conceptual overlap. Given the newness and circumscribed nature of the burden scale of the BBSC, future investigators may wish to include other well established measures of negative outcomes in any clinical trials. Despite these limitations, the BBSC appears to be a useful measure that assesses a fuller picture of children's responses to highly stressful events.

Funding

National Institutes of Health (R01 CA82378); American Lebanese Syrian Associated Charities (ALSCAC).

Conflict of Interest: None declared.