Background

Every year, approximately 70,000 young adults (ages 18–39 years) are diagnosed with cancer. These young adults immediately face multiple challenges that will affect their health-related quality of life, including the potential for their disease and/or treatment to affect their fertility. Yet research has demonstrated that few of these individuals are provided with information about the fertility-impairing potential of their treatments,^1,2 and this can be a common source of regret among young adult survivors of cancer.³

Providing information at the time of diagnosis about risk of infertility and the availability of existing interventions to maintain biological reproductive potential is critical to assist patients in making decisions about fertility preservation.⁴ The American Society for Reproductive Medicine (ASRM),⁵ ASCO,^6,7 and NCCN⁸ have each established national guidelines to enhance provider adherence and facilitate patient and provider discussions about fertility preservation options. Guidelines recommend oncologists should inform all patients about the potential effects of their disease and/or treatment on their fertility. Patients who ask for information about fertility preservation should receive it, and be referred to a reproductive specialist when appropriate.

Previous studies suggest that oncologists face many communication challenges when discussing fertility preservation with patients. Challenges are related to physician attributes (eg, knowledge barriers), patient attributes (eg, cultural or religious prohibitions for assisted reproduction), and health care factors (eg, time demands).⁹ Findings from these studies suggest that oncologists do not often refer their patients to fertility preservation specialists, with less than 50% of providers following recommended ASRM/ASCO/NCCN guidelines.⁹ Oncologist characteristics have also been related to referrals, with male and older oncologists less likely to provide referrals.^9,10 Educational materials exist yet are rarely used (<25% of US oncologists were either aware of or distributed educational materials to their patients).¹⁰ Furthermore, research suggests that some providers are uncomfortable engaging in these discussions, which may be due to the belief that fertility preservation adds more stress to the situation or is inappropriate for patients with a poor prognosis.¹¹ Other providers may lack sufficient knowledge to address questions about fertility preservation options.⁹

This study examines provider adherence to guidelines for discussing fertility preservation as documented in the medical record within a single institution for both men and women across multiple years. Although the ASCO guidelines, written to foster greater patient and provider communication about potential treatment-related infertility, were published in 2006, very little has been published on clinician- or patient-level data from the medical record in subsequent years.¹² To address this scientific gap, we sought to examine compliance with treatment guidelines at an NCI-designated comprehensive cancer center, Robert H. Lurie Comprehensive Cancer Center (RHLCCC) of Northwestern University. Based on the published self-report data of providers' attitudes and practices regarding fertility preservation discussions, it is hypothesized that (1) RHLCCC's medical oncologists will be more adherent to fertility preservation guidelines (>50%), and (2) a higher rate of fertility preservation discussions will be initiated by RHLCCC's female oncologists or when patients are diagnosed at a younger age.

Methods

Results

Data retrieved from the EDW included 1,018 new encounters with young adults seen during the 3-year period from 2010 to 2012. Individual chart review to confirm eligibility resulted in 527 cases being excluded. Cases were excluded for the following reasons: 205 patients had a cancer diagnosis and/or cancer treatments before 2010, 106 patients were being seen for a second opinion only, 123 patients had initiated treatment at an outside institution and were transferring their care to the RHLCCC, 26 patients had multiple primaries, 33 patients had experienced a recurrence, 30 patients had a history of infertility or fertility treatment before their cancer diagnosis, and 4 patients did not have an eligible cancer diagnosis. In addition, because neuro-oncology and surgical oncology encounters did not generate the automated fertility preservation questions, encounters for 36 patients seen in neuro-oncology and 1 seen in surgical oncology were also excluded, leaving 454 analyzable encounters.

Patient and oncologist characteristics for the 454 included encounters are presented in Table 1. The most common cancers affecting young adults in this patient population were breast cancer, Hodgkin lymphoma, non-Hodgkin's lymphoma, testicular cancer, and leukemia. Cancer types represented by this sample were generally consistent with SEER data.¹⁶ Based on oncologists' responses to the EMR questions about potential treatment-related infertility, more than half of the patients were informed of the effect treatment may have on their fertility (275 of 431 [63.8%; 95% CI, 59.1%–68.4%]). When examined by year, the rate was highest in 2012 (2010 = 62.2%; 2011 = 58.6%; 2012 = 70.6%); however, this finding was not significant (P=.136). Of those who were informed, 56% were interested in fertility preservation and 77% of those interested were referred to a specialist (Table 2).

Of the 114 encounters (26.4% of all encounters in the final sample) for which oncologists documented a “not applicable” response for informing their patients about the potential impact of treatment on their fertility, the probable reasons are listed in Table 2. Patients for whom chemotherapy was not indicated were coded as “probably adherent,” as were those who had already received fertility preservation recommendations, were currently pregnant, or were not interested in fertility preservation. Patients with HIV or metastatic disease and those with missing or unclear justification of the “not applicable” response were coded as “probably not adherent.” Patients who declined treatment and those for whom treatment was not indicated or the plan was undecided were

Table 1.

Patient (N=454) and Provider (N=36) Characteristics From Eligible Encounters

View Table

When adherence was examined at the bivariate level (Table 3) and after accounting for clustering within oncologist, several characteristics were associated with a greater likelihood of being informed of the potential effect treatment would have on fertility. For patients, being of a younger age at the first medical oncology visit (P=.019), being white (P=.042), being female (P=.005), and having an earlier stage of cancer (P=.022) were associated with being informed by their oncologist of the impact treatment might have on fertility. Female oncologists were equally likely as their male counterparts to communicate with patients about the impact treatment might have on fertility (P=.418).

Finally, in a mixed effects model including all variables and adjusting for clustering within oncologists, only patient sex remained significant (Table 4). Female patients were more likely to be informed by their physician of the potential impact of treatment on fertility compared with male patients (odds ratio, 3.6; CI, 1.3, 9.6; P=.012). Significant bivariate relationships between adherence and patient age at first medical oncology visit or between adherence and being a white patient were no longer significant in the mixed effects model but did reveal trends (P=.069 and P=.052, respectively).

Discussion

This study describes the degree of compliance with national clinical guidelines for discussing treatment-related infertility with newly diagnosed young adults with cancer seen at an NCI-designated Comprehensive Cancer Center over multiple years. Some of the most common barriers to fertility preservation discussions are time demands and lack of information.

Table 2.

Frequency Responses to Fertility Preservation Information Questions

View Table

We found support for our hypothesis that compliance with national fertility preservation guidelines at RHLCCC would exceed 50%. That said, initial compliance rates from our sample were suboptimal at 63.8% (ie, oncologists indicated “yes” they had informed patients of the potential impact of treatment on their fertility). When the probable “not applicable” encounters were recoded, compliance rates were substantially higher (83.2%). This represents a wide discrepancy in compliance rates and probably suggests “best-case” and “worst-case” scenarios. Although oncologists rarely provided justification for using “not applicable” within the stop gap question, our chart abstraction approach allowed our research team to identify probable reasons for these responses. This revealed a number of cases in which oncologists had indeed informed patients about the potential impact of treatment on their fertility and initiated a discussion about options to preserve their fertility. It

Table 3.

Bivariate Associations Between Clinical or Demographic Variables and Adherence^a

View Table

Table 4.

Mixed Effects Model Associations Among Clinical or Demographic Variables and Adherence^a

View Table

Our hypothesis that female oncologists would have a higher rate of compliance was not supported. After accounting for the influence of patient and provider characteristics, our mixed effects model demonstrated that female oncologists were as likely as male oncologists to inform patients about the potential of treatment-related infertility. In contrast, there was support for our hypothesis regarding patient age, because this variable demonstrated a trend in our mixed effects model. Younger patients were slightly more likely to have been informed of the effect of treatment on their fertility than older patients (P=.069). Although this finding was independent of sex, oncologists may have perceived that older women were more likely to have greater parity or be content with their family size.

Although no specific hypotheses were advanced, there was a significant finding for patient sex and a trend for race. Female patients were more than 3 times as likely as their male counterparts to be informed about a potential loss of fertility. Given the varied options for fertility preservation for women and the potential impact any fertility preservation procedures might have on treatment initiation, it is encouraging that these discussions be prioritized for female patients during the initial medical oncology encounter. Although the relative frequency of communication about potential infertility is less among young men, all patients with testicular cancer in this sample were informed of the potential impact of treatment on their fertility.

Similarly, the baseline rate of adherence to discussions of potential treatment-related infertility was high among several of the most common cancers affecting young adults. For example, patients with breast cancer, leukemia, or Hodgkin or non-Hodgkin's lymphoma were also more likely to be informed about a potential loss of fertility. Given the increased awareness of potential infertility among patients with breast or testicular cancer and those with lymphoma,^1,17–20 this is not surprising and may reflect clinic-specific practices among medical oncologists. Oncology providers at RHLCCC specialize in specific cancer diseases, and those specializing in the aforementioned cancer types see more young adult patients and theoretically are more comfortable discussing and referring patients for fertility preservation consults.

In addition, white patients were twice as likely as non-white patients to be informed of the potential effect of treatment on their fertility. It is not clear why this finding emerged, but it is possible that race may be assumed to be associated with socioeconomic status, and oncologists may be disinclined to refer patients of lower socioeconomic status for fertility preservation discussions.²¹ Fertility preservation options are often costly and cost-prohibitive for patients with limited financial resources.²²

This study has limitations. First, as already noted, the data abstracted from the EMR were an underestimate of true compliance rates, and the “not applicable” responses were apparently used to indicate a range of responses that suggested both adherence and nonadherence. This has obvious implications for conclusions about the “upper bound” of adherence rates by oncologists in our sample. A related concern is that data were abstracted from the EMR for only the patients' first medical oncology visit, because this was when the medical oncologists would have responded to the stop gap question about fertility preservation options. For a subset of patients, their disease stage and treatment plan were unknown at that time, making it difficult for our study team to determine if those “not applicable” responses were suggestive of adherence or nonadherence. Second, not all oncologists use these questions to document their discussions with young adults about potential treatment-related infertility. For example, despite the fact that central nervous system tumors are one of the more prevalent cancer types among young adults, neuro-oncology does not currently have these questions programmed into the EMR for their new patient encounters. Third, the study was conducted at a comprehensive cancer center with dedicated resources to support decision-making about fertility preservation, and therefore results may not generalize to other institutions. However, most institutions do not even collect these data or know what their baseline compliance rate is or should be.² This underscores the unique and important aspect of this research despite these limitations.

Conclusions

Given the potential for a cancer diagnosis and subsequent treatments to affect fertility, it is important for newly diagnosed young adults with cancer to receive appropriate information from their oncologists early enough in their care to maximize their options to preserve their fertility, if they choose to do so. Comprehensive cancer centers and Fertile Hope Centers of Excellence often have institutional resources to help foster these discussions, but additional work is needed to ensure these discussions occur and patients clearly understand their options. This work could be advanced through a few different strategies. For example, qualitative research approaches would be useful to better understand patient and provider barriers and facilitators of fertility preservation decision-making.^23,24 In turn, this could guide more effective use of the EMR to increase guideline adherence and improve health services delivery.

Variability in institutional resources and policies and the evolving information about fertility preservation options pose challenges to implementing best practices.^4,25 Approximately 85% of young adult patients are not treated in large academic institutions or cancer centers but in a variety of community-based settings and small oncology group practices,²⁶ and these practices often have few resources to support patients who are confronted with treatment that might impair their fertility. As such, it is possible the rate of adherence to fertility preservation guidelines at nonacademic institutions or smaller cancer centers is much lower than reported in the current study. The development and testing of fertility preservation decision-making tools through randomized controlled trials may be a beneficial future direction. Improving provider adherence to fertility preservation guidelines is a meaningful first step to provision of family-building information and fertility preservation options.