Abstract
Purpose:
The reproductive concerns of emerging adult survivors of childhood cancer are not well described, and valid measurement tools tailored to this population are lacking. The purpose of this analysis was to evaluate a modified version of the Reproductive Concerns Scale (mRCS) among male and female survivors of childhood cancer.
Methods:
This is a secondary analysis of cross-sectional survey data collected from patients enrolled on an infertility-educational intervention study. Participants completed the mRCS at baseline. Cancer treatment data were abstracted from participant medical records. Principal component analyses were conducted to evaluate the factor structure of the mRCS for males, females, and the entire sample. Internal consistency was evaluated using Cronbach's alpha. Open-ended responses were analyzed and used to assess the validity of relevant quantitative items on the mRCS.
Results:
The sample consisted of N = 98 participants who were an average of 19.1 (±1.1) years of age, 45.9% were male, and 61.2% were non-Hispanic white. Factor analyses revealed three domains: Fertility Concerns (Cronbach's alpha = 0.77), Health Concerns (α = 0.74), and Information Seeking (α = 0.57). Sex-specific factor analyses identified differences in scale items for males. The open-ended responses aligned well with participant scores on the Fertility Concerns subscale.
Conclusion:
The mRCS consists of three subscales relevant to emerging adult survivors of childhood cancer. Further analysis by sex suggests that separate scales for males and females are warranted. Future research is warranted to determine the clinical utility of using the mRCS as a screening tool to identify and address reproductive concerns among emerging adult survivors.
Introduction
With current survival rates for childhood cancer approaching 85%, the majority of children and adolescents treated for cancer will become long-term survivors. 1 Children and adolescents are often exposed to gonadotoxic therapies as part of their cancer treatment and are at risk for infertility. 2 Overall, infertility rates among childhood cancer survivors range from 42% to 66% for males2,3 and 11% to 25% for females.2,4 Future fertility is a priority for patients and families from the time of cancer diagnosis into long-term survivorship,5–9 and potential infertility is a source of distress for both male and female survivors of childhood cancer.10–12 Reproductive and fertility-related concerns are multidimensional and can encompass worry, uncertainty, frustration, and anger, as well as desire and hope for biologic children and parenthood.13,14 Among cancer survivors, the prevalence of fertility-related concerns ranges from 20% to 60% and is associated with distress, anxiety, depression, and trauma.10,15–17 The unfulfilled desire for a child and/or nulliparous status in survivorship are associated with higher levels of distress and poorer mental health.16,18,19 Survivors of childhood cancer who are in their emerging adulthood years (i.e., 18–22 years old) are unlikely to have had children before, or since, their cancer diagnosis, leaving them uncertain about their fertility status and vulnerable to fertility-related distress.
Emerging adulthood, a developmental period between adolescence and established adulthood, represents a period of identity formation beyond that of adolescence. 20 Emerging adults are diverse in terms of educational, vocational, and relationship status, and this group is heterogeneous with regards to interest in family building. 20 For survivors of childhood cancer, emerging adulthood often represents a time of transition from the pediatric to adult health care systems and unmet informational needs regarding potential infertility. 21
The Reproductive Concerns Scale (RCS) was developed to assess a range of fertility-related concerns among adult female cancer survivors. 22 Items for the RCS were drafted and refined based on findings from the literature, cognitive interviews with female survivors, and feedback from health care provider content experts. In an initial study of 231 adult female survivors of lymphoma or gynecological cancer (mean age of 45 years [±5.3]) who were 5 to 10 years postcompletion of cancer treatment, the 14-item RCS demonstrated high internal reliability (Cronbach's alpha = 0.91). 22 The RCS has subsequently been used in a variety of adult female cancer survivor studies.15,16,18,23 The RCS has been adapted to include nine quantitative items and a single open-ended item for use among adolescent female cancer survivors. 24 Readability and relevance of the adapted items were assessed in a pilot sample of 14 adolescent females (aged 12–18 years), and results supported the items as relevant and comprehensible. 25 Overall, the RCS is an established measure for assessing reproductive concerns among adult female survivors, and relevant adaptations have been made to capture age-appropriate fertility-related concerns among adolescent female survivors.
Reproductive concerns among emerging adult survivors of childhood cancer (EA survivors) are not well described, and validated measurement tools for this growing population are lacking. The purpose of this article was to conduct an exploratory factor analysis of a modified version of the RCS (mRCS) in a population of male and female EA survivors, 18 to 22 years of age. We hypothesized that the mRCS would demonstrate subscales for fertility worry, blame, and health-related concerns, as suggested by the content of the items and prior qualitative studies supporting these concepts among male and female EA survivors. 26 Open-ended responses were also explored as a validation approach of the mRCS quantitative items.
Methods
Participants and procedures
This analysis used baseline data collected as part of an educational intervention targeting knowledge of infertility risk among young adult cancer survivors. 27 Childhood cancer survivors aged 18 to 22 years were recruited from a single institution pediatric survivor clinic between 2016 and 2017. Participants completed a baseline survey assessing sociodemographic factors and the mRCS. Participant clinical characteristics were abstracted from medical record data. Participants' infertility risk level (none, low, moderate, high) was calculated using previously published methods 27 by pediatric endocrinology providers based on the patient's gonadotoxic exposure(s) (i.e., cumulative doses of alkylating and/or heavy metal chemotherapy, sites and doses of radiation affecting the gonads, gonadectomy, retroperitoneal lymph node dissection, pelvic surgery, and/or hematopoietic cell transplant). The study was approved by the Institutional Review Boards at Emory University and Children's Healthcare of Atlanta (IRB No. 92097), and survivors provided informed consent before participating in the study. Study data were collected and managed using REDCap (Research Electronic Data Capture),28,29 a secure web-based software platform.
Modified Reproductive Concerns Scale
The mRCS used in this study includes 10 quantitative items asking participants to rate their agreement with statements regarding reproductive outcomes (e.g., “One day I would like to have children,” “I feel frustrated that I might not be able to have children in the future,” and so on) on a scale of 1 (not at all) to 5 (very much). In addition to the nine items used with adolescent survivors on the previously adapted RCS, 24 an item was added “I am worried about not having a child in the future” based on Quinn et al. published suggestions for revisions. The scale also includes a single fill-in-the-blank item, “If I cannot have children, I will be ____” where participants could complete the statement.
Data analysis
Descriptive statistics for the sample was calculated, including means and standard deviations for continuous data and frequencies and percentages for categorical data. The scale item “I feel like I have control over my ability to have a child in the future” was reverse coded before analysis to be consistent with endorsement of a concern. Inter-item and item-total correlations were explored using Pearson's r. Principal component analyses were conducted to evaluate the factor structure of the mRCS within the entire sample and among males and females separately. Items with low variability in responses were not included in the principal component analysis. Using orthogonal rotation (Varimax), items with a factor loading of <0.60 or that cross loaded on two factors >0.35 were removed. Factors with eigenvalues >1.0 were considered for inclusion as a final subscale. Internal consistency for each subscale was evaluated using Cronbach's alpha. All statistical analyses were completed using SPSS version 26.
Open-ended response data were analyzed by two members of the study team with training in qualitative research (B.C. and J.G.M.). First, the data were read as a whole, and content analysis resulted in the creation of thematic categories. Both researchers coded the data into themes, and any discrepancies in coding were resolved through discussion. Agreement between coders was evaluated using Cohen's kappa coefficient. The open-ended data were used to assess the validity of the quantitative responses on the mRCS Fertility Concerns subscale.
Results
A total of 102 EA survivors were approached to participate in the study, and 98 (96%) provided consent and completed the baseline survey. Participants were an average of 19.1 (±1.1) years of age, 45.9% were male, and 61.2% were non-Hispanic white (Table 1). On average, participants were diagnosed with cancer at 9.1 (±5.4) years of age and were 8.8 (±5.0) years from treatment completion. Only 5.1% (n = 5) of participants received no gonadotoxic treatment and were not at risk for treatment-related infertility. Infertility risk for the remaining survivors was categorized as low (n = 52, 53.1%), moderate (n = 13, 13.3%), and high (n = 28, 28.6%).
Participant Characteristics (N = 98)
Gonadotoxic exposure categories are not mutually exclusive.
Radiation fields that include the ovary, testis; cranial radiation ≥30 Gy.
Factor analysis
The mean total score for the mRCS was 28.90 (±6.22, range 14–42). The Cronbach's alpha for the total score was 0.67. Inter-item correlation was explored, and demonstrated variations in item correlation were observed (r = 0.02–0.68), suggesting that the items are measuring more than one concept and the identification of subscales was warranted. Before conducting factor analyses, one item (“If I cannot have a child, I would blame my doctor”) was removed due to extremely low variability in responses (i.e., 98.9% of the sample rated “not at all”) and a low item-total correlation (r = 0.025). The scree plot and eigenvalues demonstrated a three-factor structure (Table 2) accounting for 66% of variance in the responses. Two items (“If I cannot have a child, I would blame my illness” and “I feel like I have control over my ability to have a child in the future”) had factor loadings <0.60 and were not included in the subscales. Sufficient reliability was observed for the Fertility Concerns (α = 0.77) and Illness Concerns (α = 0.74) subscales, while the internal consistency of items in the Information Seeking subscale was lower (α = 0.57). Item responses and associated subscale mean and standard deviations are shown in Table 3.
Factor Loadings, Eigenvalues, and Variance Explained by Subscales
Reproductive Concerns Scale Items and Subscale Scores
Subscale scores and Cronbach's alpha are bolded.
Regarding potential sex-based differences on the mRCS, the factor structure among females was similar to results of the entire sample (Supplementary Table S1). Among males, the three-factor structure remained with the items included in the Illness Concerns and Information Seeking subscales unchanged; however, there were differences in items included in the Fertility Concerns subscale (Supplementary Table S2). Specifically, the item “One day I would like to have children” did not load on to any subscale (factor loading <0.60), and the items “I feel like I have control of my ability to have a child in the future” and “If I cannot have a child I would blame my illness/cancer” loaded on to the Fertility Concerns subscale.
Content validation of Fertility Concerns subscale
Initial read through of the qualitative responses to the item “If I cannot have a child I will be___” identified four themes, (1) Open to Alternative Family Building, (2) Okay, (3) Upset, and (4) Unsure. Comparisons between coders demonstrated high inter-rater reliability (k = 0.91, p < 0.001). The open-ended responses allowed for an assessment of content validity that was conceptually aligned with the Fertility Concerns subscale. Table 4 displays the qualitative responses in relation to the mean scores on the Fertility Concerns subscale. Among participants with lower scores (mean score <3.0), a large proportion provided qualitative responses indicating lower concern about future fertility (e.g., “fine,” “okay”). Most participants with the highest scores on the Fertility Concerns subscale (mean score >4.0) describe disappointment, sadness, or anger regarding potential infertility, and many also plan to look into adoption or other family building options (Table 4).
Qualitative Responses by Fertility-Related Concerns Subscale Score
Qualitative responses available for 88 participants.
Discussion
This study evaluated the factor structure of the mRCS among a sample of EA survivors and three subscales emerged: Fertility Concerns, Illness Concerns, and Information Seeking. Validated assessment tools with solid psychometrics are critical for closing the knowledge gap regarding the multidimensional reproductive concerns among EA survivors. In the original development of the RCS, Wenzel et al., 22 used a summary score of all items to indicate higher levels of reproductive concerns. The mRCS includes subscales that can provide direction for education and intervention based on the survivors' concerns regarding fertility, illness, and/or desire for further information. Given the variation in inter-item correlation of the mRCS and distinct concepts assessed, examination of the individual subscale scores may be more informative in both the research and clinical settings than a single total score of concern. The population of EA survivors is developmentally distinct from survivors in established adulthood (e.g., 30 years and older) and is unlikely to currently have children or be attempting to conceive. Therefore, the discussion of potential infertility in this age range of 18 to 22 years is often future focused with education regarding their risk for treatment-related infertility and resources to assess their fertility status.
Other measures of reproductive concerns used in young adult cancer survivors (e.g., 18–39 years of age) capture different concepts impacting a survivor's distress and worry about fertility. The Fertility Problem Inventory, for example, was developed among adults in the general population and assesses fertility-related distress through the impact on social, sexual, and relationship concerns, as well as the need for parenthood and rejection of a child-free lifestyle.19,30–32 The Reproductive Concerns After Cancer scale was developed among young adult female cancer survivors (aged 18–39 years) and adapted for male survivors to assess concerns related to fertility potential, partner disclosure of fertility status, child's health, personal health, becoming pregnant, and acceptance of possible infertility.14,33,34 These measures include concepts that may not yet be relevant to EA survivors who have not tried to conceive, are not in serious romantic relationships, and are not developmentally primed to contemplate a child-free life or accept infertility. The subscales identified in the mRCS focus on fertility, illness concerns, and information seeking, which may be the priority to survivors during emerging adulthood, and addressing these concerns in very early adulthood may help reduce future fertility-related distress as survivors age.
The Fertility Concerns subscale demonstrated acceptable internal consistency. The open-ended responses aligned with scores from the Fertility Concerns subscale, suggesting that the scale is a valid measure of EA survivors' concerns. The frustration and worry captured by this subscale may be a function of misperception, uncertainty, or lack of awareness regarding assessment of fertility status and fertility preservation options. While the RCS does not have defined cut points, higher composite scores have been predictive of depressive symptoms. 16 The Fertility Concerns subscale shows promise as a clinical screening tool to identify EA survivors with concerns and provide further education regarding fertility assessment and options for preservation.
The Illness Concerns subscale also demonstrated adequate internal consistency and is conceptually supported by prior research demonstrating that survivors worry about their health and the health of their offspring. 26 Survivors with health-related concerns may benefit from discussions regarding their risk for cancer recurrence and chronic health conditions and, for females, the ability to successfully carry a pregnancy. Survivors may be reassured that evidence suggests that there is not an increased risk for congenital anomalies among offspring of childhood and adolescent cancer survivors.35,36 In addition, survivors with a cancer predisposition syndrome would benefit from referral to a genetic counselor to discuss risk to future offspring. 37
The Information Seeking subscale demonstrated weaker internal consistency. These two items capture survivors' interest in learning about their risk for infertility and if they can discuss potential infertility with their parents. These concepts may not align for some survivors, as some may want more information about risk for infertility but have poor communication with their parents, and others may be able to talk with their parents about sexual health, but lack interest in learning about infertility risk at this age; these variations could contribute to the weaker internal consistency of this subscale. Health care providers can provide tailored education to EA survivors about their risk for infertility, explore options for alternative family building, and discuss survivors' health-related concerns about their own health and the potential health of their child. While survivors may not desire a child during the years of emerging adulthood, they may benefit from a tailored discussion regarding their options for parenthood and how to plan for these options in the future. EA survivors who express concerns about future fertility should be offered referral for a fertility status assessment with a reproductive specialist.
The RCS was developed and adapted among populations of female cancer survivors. In this analysis the factor structure of the mRCS was explored by sex and demonstrated that among the male subgroup, the perception of more control over having a child was associated with lower fertility concerns; this association was not found among females or in the combined factor analysis. Although the reasons for this difference could be multifold, including a variation in perceived risk for infertility between males and females or differences in utilization of fertility preservation before cancer treatment; further exploration into the need for sex-specific assessment of fertility concerns among EA survivors is needed. In addition, the clinical utility of the mRCS subscales among males may differ from females, requiring further refinement of this instrument to be relevant to male EA survivors.
Through the open-ended responses, a large proportion of survivors expressed openness to alternative family building options and confidence in becoming foster or adoptive parents, which may be a source of hope or optimism about future parenthood for cancer survivors. Interestingly, survivors' Fertility Concerns subscale scores did not demonstrate close alignment with the open-ended responses regarding alternative family building, as survivors identified adoption as an option if they were unable to have children across the scores. This suggests that alternative family building is conceptually distinct from fertility concerns, and a companion measure assessing interest in alternative family building may be beneficial.
There are several opportunities to incorporate findings from this study into clinical practice. Survivors report a strong desire for children in the future, underscoring the need for appropriate counseling and offering of fertility preservation at the time of cancer diagnosis and recurring education about infertility risk throughout survivorship. The mRCS could serve as a screening tool to identify survivors who are experiencing reproductive concerns and tailor conversations regarding infertility in the clinical setting. For example, higher Information Seeking scores can identify survivors interested in a fertility consultation, while higher Health Concerns scores could inform discussions regarding individual risk for cancer recurrence and risk for cancer in offspring. Survivors with high Fertility Concerns scores may benefit from a fertility status assessment through referral to a reproductive specialist to ensure the provision of comprehensive fertility services and potentially alleviate distress through focused education and psychological interventions.
Limitations
This secondary analysis includes survivors from a single institution who are followed in a pediatric cancer survivorship clinic, which limits the generalizability of these results. In addition, very few brain tumor survivors are represented in this study, as brain tumor survivors are followed in a separate clinic and were not actively recruited for the parent study. Findings may not be applicable to survivors outside of the age range of participants in this study (i.e., 18–22 years of age); however, the survivors in this age range represent a developmentally distinct group who are entering a stage ripe for considerations about family building. Convergent and discriminant validity were unable to be assessed due to the study design. Future research in this area should include a confirmatory factor analysis of the mRCS in the population of EA survivors and explore the utility of this scale to direct care in the clinical setting to address reproductive concerns.
Conclusions
Evidence-based assessment tools are critical to moving the science forward to assess and address reproductive concerns among EA survivors. The findings from this study present a critical evaluation of the mRCS in a novel population of EA survivors. Three subscales emerged within the mRCS representing the multidimensional aspects of reproductive concerns in this population. Further confirmatory analysis of this measure is warranted, as well as determining clinical utility of screening for concerns among EA survivors.
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
