Disclosing a history of childhood cancer to romantic partners

To describe young adult childhood cancer survivors' disclosure of their cancer history (i.e., disclosure behavior, difficulty, and timing), perceived partner responses, and associations with relationship status satisfaction.


| BACKGROUND
Cancer during childhood can have major long-term effects on survivors' subsequent lives, including their psychosocial functioning and romantic/psychosexual development during puberty and emerging adulthood. [1][2][3][4][5][6][7][8][9] Initiating or maintaining romantic relationships may be disrupted by physical impairments, worries about infertility, body image concerns, sexual dysfunction, stigmatization, impaired social skills and prolonged time away from peers during treatment. [1][2][3][10][11][12][13] Qualitative studies showed that young adult survivors of childhood cancer sometimes felt unfit as dating partners, unattractive, or less sexually adventurous than healthy peers. 12,[14][15][16][17] Even after survivors successfully initiate relationships, they often worry about breakups. 18 In addition to potential effects of cancer and treatment on dating, survivors' behaviors of (not) disclosing their cancer history (i.e., communicating facts, thoughts, and feelings about the cancer experience), may also directly affect their romantic relationships. In general, disclosing personal life events, emotions, or experiences when dating can enhance closeness and intimacy with a (potential) partner. [19][20][21][22] Yet, such disclosure may be particularly difficult when one has to disclose a cancer history, as indicated in qualitative studies in young adult survivors. 1,2,13,20 They were unsure about when and which information to disclose, 13 while also worrying about partners' responses, being rejected, and being perceived as different or vulnerable. 2,13,20,21,23 Another qualitative study reported that survivors felt dismissed, disappointed, or frustrated by responses after disclosing their cancer history. 2 At the same time, survivors might feel obliged to share such information with potential partners. 13 Disclosing or withholding a cancer history may also influence survivors' satisfaction with their relationship status (i.e., how happy they are being single or partnered), which could also affect their general well-being and satisfaction with life. 24 Survivors who hesitate to disclose may avoid dating 13,25 and remain single unintentionally.
Survivors who are partnered may strain their existing relationships by hesitating or delaying disclosure of their cancer history, 13 potentially jeopardizing partners' trust, their own satisfaction, or quality of life.
Although the aforementioned qualitative studies have yielded some insights into childhood cancer survivors' experiences and concerns surrounding romantic relationships and disclosure, large-scale systematic data are missing. Such insights are needed to offer evidence-based psychosocial counseling and patient education materials, to support survivors in their psychosexual/psychosocial development later in life. 26 Therefore, this study will extend the previous literature by describing childhood cancer survivors' disclosure history (i.e., disclosure behavior, difficulty, timing), perceived partner responses, and relationship status satisfaction. Associations between these variables and differences by background factors will be tested. Moreover, these quantitative responses will be substantiated by qualitatively analyzing open-ended comments of survivors.

| Procedures
An embedded mixed methods design was adopted, which entailed the simultaneous collection of qualitative data to enrich quantitative data. 27 Presented data are part of a larger collaboration (E-Surv), containing two study parts: VIVE (focus: medical late effects) and InRel (focus: intimate relationships of childhood cancer survivors). 28 Data were collected by mailing information packages and reminders between April-August 2018 through the German Childhood Cancer Registry (GCCR). Survivors were randomly assigned to complete an online or paper questionnaire, and provided informed consent accordingly. Survivors were assigned an anonymized ID and only the GCCR had access to the key code linking IDs and survivors. Thus, survey responses and identifiable information were stored separately and personal data were protected at all times. The Medical Ethical Committee and data protection officer of the University Medical Center Bonn approved the study (#138/17).
Eligible participants had been diagnosed with any type of cancer before age 18, were long-term survivors (≥5 years post-diagnosis), emerging adults (age 20-25 years), and lived in Germany at the time of study. Based on these criteria, the GCCR randomly selected N = 2000 eligible survivors (note: participants were age 21-26 years at participation due to logistical delays).
A total of n = 622 survivors responded (31.1%), of whom n = 89 refused to participate and n = 7 withdrew participation (see InRel). 28 Responders were somewhat younger (23.3 vs. 24.0 years; p < 0.001) and more often female (39.6% female vs. 24.2% male; p < 0.001) than the initial pool of 2000 eligible participants, but they did not differ by type of diagnosis (p = 0.463) nor age at diagnosis (p = 0.369). Responders (n = 526) were able to skip any question throughout the survey and, for the presented analyses, we retained any survivor with complete data on disclosure, resulting in a final sample of N = 509.
One quarter of these survivors (n = 126/509) shared additional open-ended comments about disclosure for qualitative content analyses.

| Background information
Relationship status and sex were self-reported. Age and clinical data (incl. age at and type of diagnosis) were supplied by the GCCR.

| Disclosure history and responses
Face-valid multiple choice questions (Supporting Information S1: Appendix A) were completed by survivors assessing (a) disclosure behaviors, that is, whether they typically disclose their cancer BOS-VAN DEN HOEK ET AL.
-905 experience to romantic partners (never; always; not to everyone; see all response options in Supporting Information S1: Appendix A), (b) disclosure difficulty was measured on a visual analogue scale, ranging from 0 to 10 (not at all-very much), and (c) disclosure timing, that is, when survivors typically disclose (never; around the first date; after a few dates; and an open-response option). Two items assessed (d) positive and (e) negative perceived responses from potential partners (never; once/several times/always; Supporting Information S1: Appendix A). Finally, survivors were able to provide (f) any additional comments about their experiences of disclosing cancer to (potential) partners.

| Relationship status satisfaction
The Relationship Status Satisfaction Scale (ReSta) 24 measures satisfaction with either being in a relationship/married or being single, depending on a person's relationship status. ReSta consists of five items answered on a 4-point Likert scale, which are summed to one total score with higher scores indicating greater satisfaction. ReSta was previously found to have sound psychometric properties. 24,29 Cronbach's alpha was 0.94 in this study.

| Analyses
Descriptive statistics of survivors' disclosure history (behaviors, difficulty, timing) and perceived partner responses were reported.
Associations of disclosure variables with each other and differences by background factors were tested, using χ 2 -, t-, or F-tests depending on the used factors. Open-ended responses were qualitatively analyzed by means of thematic content analysis. 30 Initial coding was independently done by two authors (DWBH, VL), and subsequently discussed among all authors until agreement was reached.
Descriptive statistics of ReSta were reported and tested for differences by background factors. Three subsequent models utilizing Analyses of Variance were conducted: The first model tested the effects of disclosure on relationship status satisfaction. Second, positive and negative partner responses and interaction terms with significantly related disclosure variables were added. In the third model, background factors, which were significantly (p < 0.05) related to ReSta at the univariate level, were added along with their interaction terms.
With N = 509 survivors, post hoc power analyses indicated ample power (>0.9) for the intended analyses to detect even small effects (Hedge's g = 0.2). Therefore, comparisons for continuous variables were accompanied by effect size calculations (g ≥ 0.2 considered small, ≥0.5 moderate, or ≥0.8 large effect), whereas percentages were used for categorical variables to better guide interpreting the clinical significance of findings.
Qualitative analyses indicated three main considerations for whether survivors disclosed their cancer history (see quotes in Supporting Information S1: Appendix B). First, survivors' attitude toward disclosure determined their subsequent behaviors: Some perceived cancer as taboo in society and feared misperceptions/ prejudices, making it difficult to disclose. Others regarded openness toward (potential) partners as vital. Second, survivors' disclosure behaviors were influenced by the extent to which cancer had become part of their identity/self-image. For example, even if negative, cancer had shaped them, changed their perspective on life, and sometimes made them proud. Others were hesitant to disclose due to worrying about being perceived as sick or weak. Thereby, the extent of late effects affecting survivors' daily life varied, which also affected their self-image and disclosure behaviors. Third, anticipated effects on romantic relationship influenced survivors' decisions about disclosure. Some experienced disclosing their cancer history as a test for the relationship or an opportunity to gain a deeper connection and understanding. Few survivors called their cancer history an asset while dating, because they felt they were more interesting for dates.
In contrast, others mentioned refraining from disclosure because it felt burdensome for the relationship or they were afraid of negative responses (see also Supporting Information S1: Appendix B).

| Disclosure timing
Survivors' timing varied, but most disclosed after a couple of dates Timing was excluded from further analyses due to indicating different aspects (timing vs. reasoning) and unequal distributions.
Qualitative analyses largely substantiated the above results.
Survivors found it difficult to find the right time to disclose, but some chose specific moments: On the first/after a few dates, when the relationship became more serious, or upon becoming physically intimate. Other survivors could not intentionally choose when to disclose their cancer history because of its visibility (e.g., scars, spasms). In those cases, conversations were automatically steered toward disclosing their cancer history. Next to visibility, other facilitators of disclosure included: trust in a (potential) partner, getting older/passing of time and becoming more mature; and previous disclosure experiences that felt like practicing (Supporting Information S1: Appendix B).

| Satisfaction
Survivors were satisfied with their relationship status (M = 15.8, range: 5-20; Table 1  -909 with previous research. 26 Difficulty was strongly related to whether survivors disclosed at all (i.e., 80% with difficulties did not disclose).
Moreover, male, single, and CNS-tumor survivors were less likely to disclose. Yet, how people deal with (non-)disclosure is highly personal, 20 as further demonstrated in our qualitative analyses. Survivors' values and beliefs determined their disclosure behaviors, which included their (a) attitudes (e.g., "openness is essential" vs. "fearing taboos"), (b) identity (e.g., being proud or perceived as weak), and (c) anticipated effects on relationships ("good test" vs. "too burdensome"), which also corroborates previous qualitative studies. 2,12,13,20,21 More than half of survivors indicated they disclosed after several dates and few disclosed before/around the first date.
This aligns with healthy singles' expectations when dating a cancer survivor: 48%-76% would want to know after some dates, but hardly anyone before/at the first date. 31 Female survivors were more likely than males to disclose their cancer history, although they also experienced it as more difficult.
This corroborates previous qualitative research that women more frequently disclose, 20,21,32 suggesting they may perceive disclosure as essential to building intimacy. 33,34 Notably, survivors' relationship status differed by sex, with female survivors being somewhat more often partnered (56% vs. 44% of men), and partnered survivors disclosed and received positive responses more often. Hence, more research is needed to disentangle the potential role of sex in dating as a cancer survivor (e.g., female survivors may be more successful on the dating market than males). Lastly, and in line with previous findings, CNS-tumor survivors reported greater difficulties disclosing, while they were also more often single, 5,8,26 which may be due to impaired social or cognitive functioning later in life. 35,36 Most survivors experienced positive responses to their disclosure, which is rather surprising given that only one previous study indicated relatively positive perceived responses, 23  also necessitating the combination of subgroups/categories which may have different experiences or implications (e.g., disclosure "not to everyone"; see Supporting Information S1: Appendix B). Importantly, item responses were substantiated by survivors' open-ended answers, which also corroborated previous qualitative findings. Yet, our cross-sectional design did not allow for testing causal relations, and satisfaction scores are notorious for being skewed. Moreover, this study did not assess how and which information survivors discuss, which may also differ in different cultural contexts, and should be addressed in future research. For example, sharing clinical information about cancer was found easier than sharing emotional experiences, 2,20 and effects on potential partners should be tested. Lastly, our response rate was rather low (31%), but typical for this unique population. Although recruitment through the GCCR allowed for a population-based approach across Germany, some selection bias may have occurred (e.g., non-response due to discomfort with intimate topics).