This analysis of 1122 child and adolescent cancer survivors sought to determine the prevalence of eligible survivors attending the survivorship clinic at a major academic medical center in the Deep South and to determine factors associated with receipt of survivorship care. We dichotomized the sample as those that had received at least one survivorship visit vs. those who had no visits to determine patterns and barriers in establishing survivorship care at our center. Approximately half of the eligible cohort (52%) attended the survivorship clinic. There were significant associations between clinic attendance by diagnosis era, age at diagnosis, primary payer, cancer type, and distance to the clinic. There was no difference in attendance by rurality or race. Those diagnosed after 2009 were three times more likely to attend the survivorship clinic. This reflects the significant effort at our institution to re-engage patients lost to long-term follow-up care in the most recent area; a strategy that is working, though there remain sub-groups with lower clinic attendance.

In our cohort, each additional year of age at cancer diagnosis was associated with decreased odds of survivorship clinic attendance. Further, those diagnosed as adolescents (15–19 years at diagnosis) were least likely to attend the clinic, when compared with survivors who were diagnosed at a younger age. The unique needs of adolescent and young adult (AYA) cancer survivors are well documented in the literature [9,10,11]. In addition to medical long-term follow-up, AYA cancer survivors have unique psychosocial needs [10]. For this reason, AYAs in particular may benefit from the resources (medical and beyond) available in a long-term follow-up clinic, and it is concerning that survivors in our cohort who were diagnosed during adolescence were least likely to attend the survivorship clinic. The reasons for lack of attendance in this older age group are likely multifactorial and may include barriers related to insurance/employment opportunities or financial challenges [12, 13]. Older childhood cancer survivors may also lack awareness about their ongoing health needs or have negative memories and beliefs about their prior cancer diagnosis and treatment [14,15,16].

Those with private insurance were most likely to attend the clinic, while those who were uninsured or primarily insured with Medicaid at cancer diagnosis were less likely to attend. While lacking health insurance entirely is a clear financial barrier to care and has been previously documented among childhood cancer survivors [17], the role of Medicaid coverage is more complex. The Childhood Cancer Survivor Study reported higher rates of survivorship care utilization among those with active Medicaid (in adulthood) as compared to those with private insurance [17]. In contrast, our study applied insurance status at cancer diagnosis. It is not uncommon for childhood cancer survivors to experience a change in insurance status from cancer diagnosis to survivorship, including loss of coverage. This is particularly true for survivors after reaching age of majority due to changes in eligibility for parental plans or Medicaid. Prior studies have shown that any insurance status change, particularly loss of coverage, is associated with lower utilization of cancer related follow-up care [18]. A possible explanation for lower survivorship attendance among those without private insurance in our cohort could be insurance status change or loss, highlighting the importance of screening for barriers to care including insurance enrollment across the spectrum of care (at diagnosis, during changes in treatment, the transition point into survivorship care, and when reaching the age of majority).

The published data regarding cancer type and survivorship care is inconclusive, but several studies have suggested higher attendance among those with leukemia/lymphoma and lower attendance among those with CNS tumors [19]. Our study adds to the body of literature supporting these associations [20,21,22]. Attendance in our cohort was highest among those with leukemia or lymphoma while those with brain/CNS tumors were least likely to attend. In our cohort, those with solid tumors or other rare cancers were also less likely to attend. Differences in receipt of survivorship care by cancer type may reflect varying system level practices and care patterns across different institutions and care contexts. Nonetheless, the association between brain/CNS tumors and low uptake of survivorship care in our practice setting warrants further exploration given the significant risk of late effects in this group.

There are conflicting data regarding rurality as a risk factor for adverse outcomes in childhood cancer. While a large national sample found no association, a smaller study in Washington State using census track level rather than county measures found rurality to be a risk factor for adverse outcomes [23, 24]. In our study, childhood cancer survivors from rural/non-metro areas at cancer diagnosis were as likely to attend the survivorship clinic as their urban counterparts. This mirrors our prior study that showed no survival differences by rurality among children with cancer in Alabama [25].

Our prior study in Alabama showed a significant association with distance to care and inferior survival among children with cancer in the state. Alabama is unique in the distribution of pediatric cancer care, with most children in the state traveling to our major academic medical center, and >20% traveling more than 100 miles to care [25]. We previously showed those living farther from care had an increased risk of death at 1 year from cancer diagnosis, and the increased risk persisted into survivorship (5 years and 10 years from diagnosis) [25]. In the present study, distance (per 25-mile increase) is associated with reduced attendance at the survivorship clinic, consistent with two previously published cohorts [22, 26]. Considering the present and past Alabama studies together illustrates that those living far from survivorship care in the state are less likely to receive long-term follow-up care in the TLC Clinic, and more likely to experience late mortality.

Lastly, race/ethnicity was not associated with survivorship clinic attendance in our cohort. Non-Hispanic Black patients were well represented in the sample and were as likely to attend the TLC Clinic as non-Hispanic White patients. This finding contrasts with prior studies that report lower odds of survivorship care among Hispanic, non-Hispanic Black, and other races/ethnicities as compared to non-Hispanic White childhood cancer survivors [14, 27,28,29]. The present study describes initial establishment of survivorship care, and we should continue to evaluate potential differences by race/ethnicity in longitudinal survivorship follow-up at our institution.

Our study has some limitations. A projected end of therapy date by cancer type was applied for those lacking a precisely documented end of therapy date (N=138). However, we tailored the estimated treatment durations to the various cancer types and combined this information with disease status and vital status to determine eligibility for survivorship care. We excluded patients who did not receive systemic therapy (e.g., surgery alone). There were also some patients for whom we did not have complete treatment records available. Additionally, the variables “primary payer” and “distance to care” were collected at the time of cancer diagnosis and may have changed by the time the child was eligible for survivorship care. This study did not assess survivors pursuing survivorship care in a primary care or other long-term follow-up setting rather than the TLC Clinic and did not evaluate longitudinal follow-up in the TLC Clinic over time. Lastly, we lacked individual level social determinants of health variables (transportation, socioeconomic status) that may impact access to care in this population.

These limitations notwithstanding, the study has several strengths including the large and diverse cohort of child and adolescent cancer survivors. The cohort reflects a large proportion of patients with residence in rural areas (25%) and far from the clinic (23% were >100 miles) allowing for analysis of geographic variables. The practice environment is also a strength as the survivorship care for our academic medical center is centralized at the TLC Clinic across ages (no upper age limit). Because attrition is a consideration regarding engagement in survivorship-focused care, the TLC Clinic targets recruitment beginning 2 years following completion of therapy, which aligns with recommendations from the Children’s Oncology Group [6]. Lastly, using the ASCR dataset as the source dataset is a strength, as the data requirements for the state registry are rigorously standardized and maintained.

In conclusion, though the reasons for not attending the survivorship clinic at our major academic medical center are diverse, they present an opportunity to intervene towards the goal of having all eligible patients establish long-term follow-up care. Moving forward, it will be important to assess longitudinal follow-up in addition to establishing survivorship care. We will study local referral practices by cancer type to understand and target potential health system improvements. We identified health care access as a key barrier towards survivorship clinic attendance, including insurance status and distance to the clinic. Insurance status is a potential intervention target that may be addressed by screening/counseling at the time of transition into survivorship and at age of majority. Barriers related to distance present the opportunity to partner with primary care physicians across Alabama and/or to leverage telehealth services to decentralize survivorship care for appropriate subgroups. In all efforts, special attention to AYA survivors is warranted, as they were least likely in our cohort to receive survivorship care.