Background
Adolescents and young adults (AYAs; defined as age 15–39 years)1 diagnosed with cancer have historically been understudied, despite the fact that it is a growing population.2–5 As a result, many improvements in survival outcomes observed for pediatric and older adult patients have not been observed in the AYA population.2,3,6 Within the older adult population, significant mortality disparities have been demonstrated for patients identifying as racial/ethnic minorities or those with lower socioeconomic status (SES),7–11 but less is known about any similar mortality disparities in the AYA population.
The AYA population faces unique challenges when diagnosed with cancer, including lack of access to care,2,3,12 lack of financial security,2,3 lack of psychosocial support,2 competing life demands,2,3 lack of patient education,2,3 inadequate insurance coverage,2,3,13–15 limited survivorship-specific resources and care,3,6,16 reproductive- and fertility-related health issues,2,3,17 and lower rates of clinical trial enrollment.2,13,18–20 Limited studies have identified that these challenges can disproportionally impact AYA patients who identify as Hispanic, non-Hispanic (NH)–Asian/Pacific Islander, and NH-Black,2,3,6 as well as socioeconomically deprived groups.2,3,6 Furthermore, few studies have identified lower survival rates for non-White4,14,21–24 AYA patients and AYA patients with lower SES.4,21,23–26
However, more research is needed to understand trends in mortality among AYAs with cancer more precisely. It is well-known that both uninsured and low SES populations have worse access to health care, and many patients of lower SES are underinsured or uninsured.27–29 Many of the aforementioned studies examined disparities in survivorship by SES and race/ethnicity in populations receiving care from a variety of health systems. This may limit interpretability of results, given that wide-ranging access to and delivery of health care among included patients may impact mortality outcomes. To address this gap, we evaluated the associations of SES and race/ethnicity to mortality outcomes among insured AYA patients treated within a large integrated health care system with similar access to care within a diverse population of Southern California.
Methods
Research Setting and Subjects
Our study population included patients within the Kaiser Permanente Southern California (KPSC) database aged 15 to 39 years and diagnosed with the 14 most common cancer types by incidence between January 1, 2010, and December 31, 2018. KPSC is a large, vertically integrated health care system that affords accessible, high-quality multidisciplinary care delivery; ease of communication among care providers; use of a single electronic health record (EHR); internalized imaging and pharmacy; and accessible patient data. The KPSC region includes residents living in 6 counties: Imperial, Los Angeles, Orange, Riverside, San Bernardino, and San Diego. We determined the 15 most common cancers in our study population, excluding thyroid cancer: breast, testicular, leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, melanoma, colon, cervical, uterine, renal, brain, soft tissue, rectosigmoid, ovarian, and gastric. These cancers represented 85% of total AYA cancers in our data set. Thyroid cancer was excluded due to its well-known high survival rate, which in our population was 99.20%, and thus less relevant to the purposes of our analysis. The remaining cancer types, including anal, biliary, bladder, bone, larynx, liver, lung, pancreas, penile, prostate, salivary, tongue, and vulvar, were excluded due to low sample sizes. The end date of the study follow-up period was September 14, 2020. Included patients had completed their first course of treatment with KPSC. Patients were followed for a maximum of 10 years after their original date of diagnosis. This study received approval from the Institutional Review Boards of Kaiser Permanente and the State of California, both of which waived any requirement for verbal or written consent due to the use of deidentified data.
Data Collection
AYAs with cancer were identified through KPSC’s SEER-affiliated cancer registry, to which all patient-level EHR data are linked. Race/ethnicity, sex, age and stage at diagnosis, cancer type, insurance status, and other demographic data were derived from the EHR. Race/ethnicity is documented via self-report when a patient is enrolled into a KPSC insurance plan, which is then reflected in the EHR. The major outcome of interest was all-cause mortality, which was ascertained through our internal cancer registry for all patients meeting the inclusion criteria.
We used the Neighborhood Deprivation Index (NDI), created by Messer et al,30 as a marker of SES at the US Census tract level. The NDI is a comprehensive composite measure of the multifactorial nature of SES. It includes several metrics, such as unemployment level and number of households on public assistance, below the poverty line, or with an income of <$30,000 per year, and then combines them into a single variable. A greater NDI score indicates a higher level of neighborhood deprivation, whereas a lesser score indicates a lower level of deprivation. The NDI was used following Messer’s protocols, and we chose to stratify patients into 4 levels of deprivation based on 2018 Kaiser Permanente SCAL data, with quartile 1 (Q1) being the least deprived and quartile 4 (Q4) being the most deprived, for easier interpretability. Patient-level NDI was estimated using census tract based on home address.
Statistical Analysis
We calculated mortality rates per 1,000 person-years for the study population as a whole and by race/ethnicity and NDI quartiles separately. We used multivariable Cox regression to estimate hazard ratios (HRs) for all cause-mortality by SES and race/ethnicity. The proportional hazard assumption was validated in our analysis through inspection of the log-log survival curves. When calculating HRs for each NDI quartile, the least deprived group (Q1) was the reference group. When calculating HRs for each racial/ethnic group, NH-White patients was the reference group. The Cox model included race/ethnicity and NDI subgroup, as well as sex (female, male), age (15–39 years) and stage (0, I, II, III, IV, unstaged/unknown) at diagnosis, insurance type (commercial, Medicaid, Medicare or dual, unknown), and cancer type (see Table 1). Lastly, to address concerns related to disparities in initial stage of presentation for different patient populations within our study population, we performed Mann-Whitney U tests to determine any variability within our cohort.
Incidence and Mortality Rate per 1,000 Person-Years for All Cancer Types
Results
Study Population
A total of 6,379 patients met our inclusion criteria. Tables 2 and 3 show demographic data for all patients stratified by race/ethnicity and NDI quartiles. Of the overall cohort, 59% of patients were female and the median age at diagnosis was 33 years. The cohort was racially/ethnically diverse: 45% Hispanic, 36% NH-White, 10% NH-Asian/Pacific Islander, 7% NH-Black, and 3% other/unknown race/ethnicity. Based on the aforementioned NDI quartiles, 21% of patients were in Q1 (least deprived), 30% were in Q2, 29% were in Q3, and 21% were in Q4 (most deprived). According to AJCC stages, 37% of patients were stage I at diagnosis, 16% were stage II, 10% were stage III, 8% were stage IV, 11% were not staged/unstageable, and the stage was unknown for 18%. Insurance types represented by the study population included 89% commercial coverage, 6% Medicaid, 4% unknown, and 1% Medicare or dual coverage. A total of 753 deaths occurred during the study period. Overall, the most common cancer types in this cohort were breast, testicular, and melanoma.
Patient Demographics by Race/Ethnicity
Patient Demographics by NDI Quartile
Table 3 shows both the incidence and mortality rate per 1,000 person-years of the study population and for each individual cancer type included in the study. The most fatal cancer types included gastric, brain, and colon, whereas the least fatal cancer types included testicular, Hodgkin lymphoma, and melanoma.
Mann-Whitney U tests of our cohort showed that NH-White patients had a significantly higher chance to have a less severe stage at presentation compared with Hispanic, NH-Black, and NH-Asian/Pacific Islander patients.
All-Cause Mortality Rates by Race/Ethnicity and SES
All-cause mortality rates per 1,000 patient-years differed for each racial/ethnic group. Patients of other/unknown race (29.1%) and NH-Black (27.2%), NH-Asian/Pacific Islander (26.2%), and Hispanic (24.5%) patients all experienced higher mortality rates compared with NH-White (16.4%) patients, implying some survivorship disparity among patients of different races/ethnicities.
Patients belonging to the most deprived quartile (Q4) had the highest all-cause mortality rate per 1,000 patient-years, whereas patients belonging to the second least deprived quartile had the lowest mortality rates per 1,000 person-years, followed closely by the least deprived group (Q1: 19.9%; Q2: 19.8%; Q3: 24.1%; Q4: 27.4%).
All-Cause Mortality by Race/Ethnicity and SES in Adjusted Model
Table 4 shows both the crude and adjusted association between all-cause mortality and race/ethnicity in our cohort. Similar to the crude mortality rates, non-White patients experienced higher risks of mortality in crude association. Covariates extracted for adjusted HRs included sex, age, cancer type, insurance type, stage at diagnosis, and SES. Compared with NH-White patients, Hispanic patients (adjusted HR, 1.31; 95% CI, 1.09–1.59; P=.004) demonstrated a statistically significant higher risk for all-cause mortality, even when controlling for all covariates. Similarly, NH-Black patients (adjusted HR, 1.34; 95% CI, 1.00–1.83; P=.05) experienced a higher risk of mortality compared with NH-White patients of marginal significance.
Association Between All-Cause Mortality and Race/Ethnicity for All Patients
Table 5 shows both the crude and adjusted association between all-cause mortality and NDI quartile in our patient population. All covariates—sex, age, cancer type, stage at diagnosis, insurance type, and race/ethnicity—were included when evaluating the adjusted HRs. Crude mortality risk was progressively higher for each socioeconomic quartile compared with the least deprived group, with increasingly higher risk with higher levels of deprivation. However, when using our aforementioned model, these differences in crude mortality risk were eliminated.
Association Between All-Cause Mortality and SES for All Patients
Discussion
Our study suggests that racial/ethnic disparities in all-cause mortality exist for insured AYAs with cancer identifying as Hispanic or NH-Black, but no differences between patients of different socioeconomic background were observed. In unadjusted models, NH-White patients had both lower mortality rates and lower risk of all-cause mortality compared with Hispanic, NH-Asian/Pacific Islander, and NH-Black patients. In our adjusted model, Hispanic and NH-Black patients had a significantly higher risk of mortality compared with NH-White patients, whereas the remaining racial/ethnic groups showed no differences in all-cause mortality risk compared with NH-White patients. Several factors could contribute to these observed trends, including reduced levels of clinical trial enrollment,18–20 health literacy,31,32 and challenges and/or disparities regarding cost32,33 and access to care,34–36 as well as other comorbid medical conditions12,37,38 or lifestyle factors37 excluded from our analysis. Additionally, disparities may exist between patients of different racial/ethnic identities regarding their stage at presentation, which could impact mortality rates or quality of life of patients based on treatment intensity and risk for developing late effects. For our cohort, we performed Mann-Whitney U tests and found that NH-White patients had a significantly higher chance of presenting with a less severe cancer stage compared with Hispanic, NH-Black, and NH-Asian/Pacific Islander patients, which could impact differences in treatments received and other outcomes. Importantly, this analysis does not control for cancer type or other factors, thus several possible confounding variables exist. We also included cancer stage as a covariate in our adjusted model to try to control for these differences and their influence on mortality rates. We recognize that some of these factors may be driven by historic and ongoing structural and institutional racism,39 and thus the drivers of racial/ethnic survival disparities are most likely inherently multifactorial in nature.
When analyzing differences in mortality rates among various socioeconomic strata, crude results suggest that patients with cancer from lower SES, measured through the use of the NDI, have higher rates of all-cause mortality compared with patients of higher SES. However, this association was eliminated upon controlling for several potential contributing factors in our model, including elements such as patient demographics, insurance type, cancer stage and type, and age at diagnosis. AYA patients with higher levels of deprivation may have limited access to health care,1,2,15 and thus may present at later clinical stages, resulting in a higher risk of dying from their cancer. To evaluate for differences in our study population, we performed Mann-Whitney U tests and found that patients from the most deprived NDI quartile had a significantly higher chance to present at a more severe stage compared with patients from the 2 least deprived quartiles. Multiple variables may influence these findings, because this analysis does not control for cancer type or other covariates, but these differences could influence treatment, quality of life, or survival outcomes. By including stage at presentation as a covariate in our model, we aimed to control for some of this effect when analyzing survival outcomes. Our study was conducted within a vertically integrated health care system that affords all members relatively equal access to care, and so perhaps this model helped mitigate socioeconomic drivers of mortality disparities that have been observed in other studies. These findings may suggest that access to care and/or insurance status may play significant roles in socioeconomic mortality disparities observed in other studies, but this is difficult to interpret because our study does not offer a corresponding comparison group to elucidate this possibility.
Our study consists of several strengths. First, we examined mortality disparities within an insured population with relatively equal access to care, which allowed us to evaluate the effect of race/ethnicity and SES without concern for confounding effect of insurance coverage. Second, use of the NDI tool afforded a more holistic interpretation of patient deprivation as opposed to relying solely on household income as a marker of SES. Additionally, our cohort represented a racially/ethnically and socioeconomically diverse patient population. In comparison with a cohort of 497,452 AYAs with cancer analyzed between the years 1973 and 2015,40 our study population showed similar distribution for age and female predominance (59%). Racially/ethnically, our population was more diverse. The referenced general population was reported as 79.9% NH-White, 10.3% NH-Black, and 8.2% NH-Asian/Pacific Islander and American Indian, compared with our study population distributed as 35.8% NH-White, 9.7% NH-Asian/Pacific Islander, and 6.6% NH-Black, and 2.7% other/unknown; 45% of our patients identified as Hispanic, which was not reported in the comparison analysis referenced. Furthermore, previous analysis of this data set has shown similarities for both the distribution of cancer types and overall 5-year survival rate for our AY.
A cancer population compared with national SEER-level data.41 Another strength was our ability to analyze the association of all-cause mortality between race/ethnicity and SES independently of one another. Lastly, we used state cancer registry mortality data for our analysis, thus hopefully reducing survivorship or selection biases.
Several limitations exist within our study. First, our study may be underpowered due to low sample sizes within our racial/ethnic and NDI quartile subgroups, thus potentially limiting our ability to demonstrate significant mortality differences between some groups. Second, although our KPSC health care system strives to incorporate patient’s self-identified race/ethnicity into the EHR through self-reporting processes on enrollment, the listed race/ethnicity in our dataset is subject to potential bias or assumed race/ethnicity. Third, our use of all-cause mortality may not be reflective of cancer-driven mortality outcomes for all patients; however, a recent study demonstrated that 86.2% of deaths in a similar AYA cancer population from the same KPSC health care system were cancer-related, thus mitigating this concern.42 Fourth, although we included cancer type as a covariate, we recognize that different cancer types may have differences in prognoses, average stage at clinical presentation, and disease progression, and therefore analyzing them in a combined fashion may mask results that may be different for certain cancer types. Future research should examine individual cancer types to better elucidate disease-specific factors. Fifth, although the NDI is a useful and validated tool for evaluating the influence of SES on health outcomes, the tool operates via census tracts at a neighborhood level, thereby negating individual-level scope of analysis. Additionally, the AYA population is transient in nature due to shifting employment, education, or personal opportunities. Therefore, a secondary limitation of using the NDI tool is that any SES determination will be predicted based on their address at diagnosis and will not account for any migration during the course of treatment. Sixth, although we focused on survival disparities specifically, many other types of disparities may exist along the cancer care continuum for underserved patients that should be further examined. Additionally, we excluded thyroid cancer due to the known high survival rate, but it is important to recognize that disparities in mortality and other aspects of care for patients with thyroid cancer of various racial/ethnic or socioeconomic backgrounds may exist. Lastly, we were limited to a maximum follow-up period of 10 years following initial diagnosis, and so longer-term survival disparities could not be evaluated.
Several additional opportunities for future research exist for further evaluation of disparities in patient outcomes for patients belonging to various SES strata and racial/ethnic groups within the AYA population and beyond. Our study adds to the growing literature that racial/ethnic disparities do exist within the AYA cancer population. Furthermore, SES influences may drive different mortality outcomes as well. Future research and interventional efforts must explore, identify, and combat these inequities on a systems level. Interventions specific to the AYA population should ensure equitable access to supportive, interdisciplinary services throughout the course of treatment; improved and proportional clinical trial enrollment; and improved insurance coverage and reduced costs. Furthermore, identification of the various social determinants of health, including factors such as access to affordable and accessible cancer care, housing and food insecurity, and difficulty with transportation to cancer care facilities, will be essential for health care systems to ensure optimal patient care and outcomes. Upon recognition of social needs, health care systems must develop workflows that connect patients with broader community-based organizations that can support their needs. In January 2023, the Children’s Oncology Group released a blueprint regarding identifying and reducing disparities in AYA cancer care, with a focus on access to care, improved survival, unmet financial needs, symptom management, and sexual health.43 Thus, further research regarding optimal implementation strategies for identifying and acting on social drivers of health is imperative to ensure equitable care for AYAs with cancer.
Conclusions
Our study demonstrates that racial/ethnic differences in all-cause mortality exist for insured AYAs with cancer in that Hispanic and NH-Black patients have a higher risk of death compared with NH-White patients when controlling for several covariates. No mortality differences existed among patients of different socioeconomic strata as measured by the NDI when controlling for certain confounding variables. Furthermore, we leveraged our vertically integrated health care system to better elucidate drivers of socioeconomic disparities in all-cause mortality in the AYA population by evaluating patients with relatively equal access to care. Future research must further evaluate long-term racial/ethnic and socioeconomic differences in survival as well as best practices and interventions to reduce disparities in clinical trial enrollment, access to care, or financial burden for the AYA cancer population.
References
- 1.↑
US National Cancer Institute. Adolescents and young adults with cancer. Accessed January 5, 2024. Available at: https://www.cancer.gov/types/aya
- 2.↑
Overholser L, Kilbourn K, Liu A. Survivorship issues in AYA oncology. Med Clin North Am 2017;101:1075–1084.
- 3.↑
Burkart M, Sanford S, Dinner S, et al. Future health of AYA survivors. Pediatr Blood Cancer 2019;66:e27516.
- 4.↑
Murphy CC, Lupo PJ, Roth ME, et al. Disparities in cancer survival among adolescents and young adults: a population-based study of 88,000 patients. J Natl Cancer Inst 2021;113:1074–1083.
- 5.↑
Smith MA, Seibel NL, Altekruse SF, et al. Outcomes for children and adolescents with cancer: challenges for the twenty-first century. J Clin Oncol 2010;28:2625–2634.
- 6.↑
Keegan TH, Lichtensztajn DY, Kato I, et al. Unmet adolescent and young adult cancer survivors information and service needs: a population-based cancer registry study. J Cancer Surviv 2012;6:239–250.
- 7.↑
Cooper RM, Chung J, Hogan T, et al. Patterns of overall mortality by race/ethnicity and socioeconomic status in insured cancer patients in Southern California. Cancer Causes Control 2021;32:609–616.
- 8.↑
Smedley BD, Stith AY, Nelson AR, et al. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. National Academies Press; 2003.
- 9.↑
Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/ethnicity and socioeconomic status. CA Cancer J Clin 2004;54:78–93.
- 10.↑
Velasco-Mondragon E, Jimenez A, Palladino-Davis AG, et al. Hispanic health in the USA: a scoping review of the literature. Public Health Rev 2016;37:31.
- 11.↑
Singh GK, Jemal A. Socioeconomic and racial/ethnic disparities in cancer mortality, incidence, and survival in the United States, 1950–2014: over six decades of changing patterns and widening inequalities. J Environ Public Health 2017;2017:2819372.
- 12.↑
Chao C, Bhatia S, Xu L, et al. Chronic comorbidities among survivors of adolescent and young adult cancer. J Clin Oncol 2020;38:3161–3174.
- 13.↑
Parsons HM, Harlan LC, Seibel NL, et al. Clinical trial participation and time to treatment among adolescents and young adults with cancer: does age at diagnosis or insurance make a difference? J Clin Oncol 2011;20;29:4045–4053.
- 14.↑
Moke DJ, Tsai K, Hamilton AS, et al. Emerging cancer survival trends, disparities, and priorities in adolescents and young adults: a California Cancer Registry-based study. JNCI Cancer Spectr 2019;3:pkz031.
- 15.↑
Alvarez EM, Keegan THM, Johnston EE, et al. Adolescent and young adult (AYA) oncology patients: disparities in access to specialized cancer centers. J Clin Oncol 2016;123(Suppl):Abstract e18020.
- 16.↑
Smits-Seemann RR, Kaul S, Zamora ER, et al. Barriers to follow-up care among survivors of adolescent and young adult cancer. J Cancer Surviv 2017;11:126–132.
- 17.↑
Coker Appiah L, Fei YF, Olsen M, et al. Disparities in female pediatric, adolescent and young adult oncofertility: a needs assessment. Cancers (Basel) 2021;13:5419.
- 18.↑
Siembida EJ, Loomans-Kropp HA, Trivedi N, et al. Systematic review of barriers and facilitators to clinical trial enrollment among adolescents and young adults with cancer: identifying opportunities for intervention. Cancer 2020;126:949–957.
- 19.↑
Muffly L, Yin J, Jacobson S, et al. Disparities in trial enrollment and outcomes of Hispanic adolescent and young adult acute lymphoblastic leukemia. Blood Adv 2022;6:4085–4092.
- 20.↑
Mittal N, Saha A, Avutu V, et al. Shared barriers and facilitators to enrollment of adolescents and young adults on cancer clinical trials. Sci Rep 2022;12:3875.
- 21.↑
DeRouen MC, Parsons HM, Kent EE, et al. Sociodemographic disparities in survival for AYAs with cancer differ by health insurance status. Cancer Causes Control 2017;28:841–851.
- 22.↑
Tehranifar P, Neugut AI, Phelan JC, et al. Medical advances and racial/ethnic disparities in cancer survival. Cancer Epidemiol Biomarkers Prev 2009;18:2701–2708.
- 23.↑
Keegan THM, DeRouen MC, Parsons HM, et al. Impact of treatment and insurance on socioeconomic disparities in survival after adolescent and young adult Hodgkin lymphoma: a population-based study. Cancer Epidemiol Biomarkers Prev 2016;25:264–273.
- 24.↑
Chao C, Chiu V, Xu L, et al. Survival differences by race/ethnicity and neighborhood socioeconomic status in adolescents and young adults diagnosed with non-Hodgkin lymphoma. J Adolesc Young Adult Oncol 2015;4:76–83.
- 25.↑
Bownes LV, Stafman LL, Maizlin II, et al. Socioeconomic disparities affect survival in malignant ovarian germ cell tumors in AYA population. J Surg Res 2018;222:180–186.e3.
- 26.↑
Cuglievan B, Berkman A, Dibaj S, et al. Impact of lagtime, health insurance type, and income status at diagnosis on the long-term survival of adolescent and young adult cancer patients. J Adolesc Young Adult Oncol 2021;10:164–174.
- 27.↑
Call KT, McAlpine DD, Garcia CM, et al. Barriers to care in an ethnically diverse publicly insured population: is health care reform enough? Med Care 2014;52:720–727.
- 28.↑
Zhu J, Brawarsky P, Lipsitz S, et al. Massachusetts health reform and disparities in coverage, access and health status. J Gen Intern Med 2010;25:1356–1362.
- 29.↑
DeNavas-Walt C, Proctor BD, Smith JC. Income, poverty, and health insurance coverage in the United States: 2007. Accessed January 5, 2024. Available at: https://www.census.gov/library/publications/2008/demo/p60-235.html
- 30.↑
Messer LC, Laraia BA, Kaufman JS, et al. The development of a standardized neighborhood deprivation index. J Urban Health 2006;83:1041–1062.
- 31.↑
Richter D, Mehnert A, Forstmeyer D, et al. Health literacy in adolescent and young adult cancer patients and its association with health outcomes. J Adolesc Young Adult Oncol 2019;8:451–457.
- 32.↑
Prussien KV, Barakat LP, Darabos K, et al. Sociodemographics, health competence, and transition readiness among adolescent/young adult cancer survivors. J Pediatr Psychol 2022;47:1096–1106.
- 33.↑
Kaul S, Avila JC, Mehta HB, et al. Cost-related medication nonadherence among adolescent and young adult cancer survivors. Cancer 2017; 123:2726–2734.
- 34.↑
Kirchhoff AC, Lyles CR, Fluchel M, et al. Limitations in health care access and utilization among long-term survivors of adolescent and young adult cancer. Cancer 2012;118:5964–5972.
- 35.↑
Meernik C, Dorfman CS, Zullig LL, et al. Health care access barriers and self-reported health among adolescent and young adult cancer survivors. J Adolesc Young Adult Oncol 2024;13:112–122.
- 36.↑
Freyer DR, Brugieres L. Adolescent and young adult oncology: transition of care. Pediatr Blood Cancer 2008;50(5 Suppl):1116–1119.
- 37.↑
Smitherman AB, Anderson C, Lund JL, et al. Frailty and comorbidities among survivors of adolescent and young adult cancer: a cross-sectional examination of a hospital-based survivorship cohort. J Adolesc Young Adult Oncol 2018;7:374–383.
- 38.↑
Lazow MA, Salloum R, Pressey JG. Comorbidity burden in a cohort of adolescent and young adult patients diagnosed with central nervous system tumors and sarcomas. J Adolesc Young Adult Oncol 2023;12:215–223.
- 39.↑
Cheung CK, Tucker-Seeley R, Davies S, et al. A call to action: antiracist patient engagement in adolescent and young adult oncology research and advocacy. Future Oncol 2021;17:3743–3756.
- 40.↑
Scott AR, Stoltzfus KC, Tchelebi LT, et al. Trends in cancer incidence in US adolescents and young adults, 1973–2015. JAMA Netw Open 2020;3:e2027738.
- 41.↑
Chao C, Chiu V, Mueller LA, et al. Exploring the feasibility of establishing a retrospective cohort of survivors of adolescent and young adult cancer to study long-term health outcomes in an integrated managed care environment. J Adolesc Young Adult Oncol 2013;2:59–65.
- 42.↑
Armenian SH, Xu L, Cannavale KL, et al. Cause-specific mortality in survivors of adolescent and young adult cancer. Cancer 2020;126:2305–2316.
- 43.↑
Beauchemin MP, Roth ME, Parsons SK. Reducing adolescent and young adult cancer outcome disparities through optimized care delivery: a blueprint from the Children’s Oncology Group. J Adolesc Young Adult Oncol 2023;12:314–323.
