Fax: (801) 585-0900
Limitations in health care access and utilization among long-term survivors of adolescent and young adult cancer
Version of Record online: 24 SEP 2012
Copyright © 2012 American Cancer Society
Volume 118, Issue 23, pages 5964–5972, 1 December 2012
How to Cite
Kirchhoff, A. C., Lyles, C. R., Fluchel, M., Wright, J. and Leisenring, W. (2012), Limitations in health care access and utilization among long-term survivors of adolescent and young adult cancer. Cancer, 118: 5964–5972. doi: 10.1002/cncr.27537
- Issue online: 19 NOV 2012
- Version of Record online: 24 SEP 2012
- Manuscript Accepted: 21 FEB 2012
- Manuscript Revised: 2 FEB 2012
- Manuscript Received: 12 DEC 2011
- survivors of adolescent and young adult cancer;
- health care access;
- health insurance;
- cost of medical care
Health care outcomes for long-term survivors of adolescent and young adult (AYA) cancer were compared with young adults without a cancer history, using the 2009 Behavioral Risk Factor Surveillance System data.
Eligible participants were 20 to 39 years of age. There were N = 979 who self-reported a cancer diagnosis between the ages of 15 to 34 years and were at least 5 years from diagnosis (excluding nonmelanoma skin cancer). The remaining 67,216 participants with no cancer history were used as controls. Using multivariable regressions, relative risks and 95% confidence intervals were generated to examine the relationship of survivor status on indicators of poor health care (uninsured, no personal health care provider, no routine care, and avoiding seeing a doctor due to cost). Adjusted proportions were calculated by demographic groups. Results are weighted by Behavioral Risk Factor Surveillance System survey design.
Although the proportion uninsured did not differ (21% of survivors vs 23% of controls), AYA survivors reported forgoing care due to cost at higher levels than controls (relative risk = 1.67, 95% CI = 1.44-1.94). Cost barriers were particularly high for survivors aged 20 to 29 years (44% vs 16% of controls; P < .001) and female survivors (35% vs 18% of controls; P < .001). Survivors reporting poorer health had more cost barriers. Moreover, uninsured survivors tended to report lower use of health care than did controls.
AYA cancer survivors may forgo health care due to cost barriers, potentially inhibiting the early detection of late effects. Expanding health insurance coverage for young cancer survivors may be insufficient without adequate strategies to reduce their medical cost burdens. Cancer 2012. © 2012 American Cancer Society.
More than 69,000 adolescent and young adults (AYA) aged 15 to 39 years are diagnosed with cancer each year in the United States, and the majority will transition to long-term survivorship.1 AYAs are often diagnosed with different cancers than those traditionally seen in adult oncology clinics, with germ cell tumors, sarcomas, leukemias, lymphomas, melanomas, and thyroid cancers comprising the majority of cancers in this age cohort.2, 3 Although 80% of AYA patients with cancer survive at least 5 years after diagnosis, many are at risk for developing chronic health conditions from the chemotherapy, radiation, and surgery received during treatment.2, 4 Medical care in the years following cancer therapy is particularly important to screen survivors for late effects, such as secondary cancers, infertility, and cardiac conditions.
Although adolescent cancer patients have often been included in studies of childhood cancer survivors, young adult cancer survivors aged 20 to 39 years are underrepresented in the survivorship literature.5 Specific to health care access, this age group is often included in aggregate analyses of survivors under age 65 years,6, 7 but few studies have examined the unique health care needs of AYAs. Studies of younger survivor populations, such as pediatric and adolescent cancer survivors, have found that survivors under age 35 are more likely to be on public insurance and in general, are more likely to be uninsured than their siblings.8 This is of particular concern because uninsured survivors report lower use of medical care and cancer-specific follow-up care.9, 10 In the general population, insurance coverage can be particularly unstable for young adults due to school and job transitions. Younger adults in their 20s, who are often less financially stable and more likely to be employed in jobs that do not offer health insurance benefits, are uninsured more often compared with adults in their 30s.11, 12
For AYA cancer survivors, a cancer diagnosis during adolescence and young adulthood could have substantial effects on their health care by impacting their education and labor force outcomes, further limiting their access to important resources such as employer-sponsored health insurance coverage.13 Although AYA cancer survivors report that adequate health insurance coverage is one of their most important health care needs,14 we are aware of no population-level studies for this age group that have assessed insurance coverage and other markers of health care access, such as having a primary care provider and the ability to afford health care. Because the majority of AYA cancer survivors will transition back into the primary care setting following their cancer treatment, assessing limitations in health care access for this population will have important implications for improving their quality of care.
The aim of the current study was to determine whether survivors of AYA cancer have poorer health care access and utilization than the general population. Using a population-based sample from the 2009 Behavioral Risk Factor Surveillance System (BRFSS) data, we examined whether survivors of AYA cancer are at a higher risk of being uninsured, having no medical provider, having no routine care in the past year, and forgoing care due to cost in the past year, compared to similarly aged adults without a cancer history. We then assessed whether certain subgroups of AYA survivors, including younger survivors, those lacking health insurance, and those with health limitations, may be at particular risk for health care disparities.
MATERIALS AND METHODS
We used the 2009 BRFSS data, an annual, nationally representative survey implemented in the United States, the District of Columbia, Puerto Rico, Guam, and the US Virgin Islands. BRFSS uses random-digit telephone dialing methods to sample noninstitutionalized adults aged 18 years or older. There were 432,607 adults who participated in the 2009 BRFSS, which was the first year that cancer survivorship questions were asked of the full sample. Participants were asked “Have you ever been told by a doctor, nurse, or other health professional that you had cancer?” The test–retest reliability for this question is adequate at k = 0.91.15 The BRFSS survey design, sample characteristics, and questionnaires are available at www.cdc.gov/brfss.
We limited our sample eligibility to participants who were aged 20 to 39 years at interview. Of the 74,433 respondents meeting this age range, 5161 did not know whether they had a cancer history, refused to answer the question, or were missing. This left N = 67,216 eligible participants as controls and N = 2056 who self-reported having a history of cancer. Of those with a cancer history, 1616 reported being diagnosed with cancer between the ages of 15 and 34 years. Because we were interested in long-term survivors of AYA cancer (that is, the majority had completed their cancer treatment), we excluded 516 who were less than 5 years after diagnosis. We also excluded 97 who reported a history of nonmelanoma skin cancer, because these cancers are typically noninvasive and are treated with minimal outpatient procedures. An additional 24 patients were excluded who did not know their cancer diagnosis or refused to answer the question. With these exclusions, we had N = 979 AYA cancer survivors available for analysis.
In addition, we performed secondary analyses using a restricted cancer sample that excluded participants who reported a history of cervical, thyroid, or testicular cancer, as well as cancers grouped as “other.” Because treatment information is not available in BRFSS, this secondary analysis was run to exclude cancers that often present as in situ and those which often only require surgical resection without chemotherapy or radiation. We had N = 396 cancer survivors in the restricted sample.
Participants indicating a cancer history were asked the age at diagnosis for their first primary cancer and whether they had multiple cancers. Survivors were asked to report the type of cancer for their most recent diagnosis. For diagnosis type, respondents were provided with a list of cancers only if they needed prompting. Using participants' current age and age at first diagnosis, we calculated the years since diagnosis.
Health Care Access and Utilization Outcomes
All BRFSS participants responded to several questions about their health care access and use. Health insurance coverage was asked as “Do you have any kind of health care coverage, including health insurance, prepaid plans such as health maintenance organizations, or government plans such as Medicare?”; we classified those who said “yes” as insured. Participants who responded as having 1 or more provider to the question “Do you have one person you think of as your personal doctor or health care provider?” were indicated as having a personal medical provider.
Use of routine medical services was asked as “About how long has it been since you last visited a doctor for a routine checkup?” This question specified a routine checkup as a general physical exam, not an exam for a specific injury, illness, or condition. Participants indicating a visit within the past year were designated as having routine care. Inability to access care due to cost was determined by the question “Was there a time in the past 12 months when you needed to see a doctor but could not because of cost?” All participants indicating “yes” were designated as having cost barriers to medical care.
We examined sex (male or female), current age (grouped as 20-29 and 30-39 years), and race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic/Latino, and other/mixed). Other demographic variables included marital status, highest educational attainment, household income, and current employment status, categorized as in Table 1.
|Characteristic||AYA Survivors N = 979||Controls N = 67,216||P|
|N||Weighted %||N||Weighted %|
|Highest completed educationb|
|≤High school education||339||37||21,521||34||.16|
|Divorced or separated||227||19||6,278||7|
|Out of work||105||11||6,246||12|
|Unable to work||112||9||2,019||3|
|Not in labor force||164||19||11,556||18|
|General health statusf|
|Frequent physical distressf|
|Frequent mental distressf|
We also used measures from the Centers for Disease Control and Prevention's health-related quality-of-life measure (CDC-HRQOL-4, known as the “Healthy Days Measures”) to assess self-rated general health and number of healthy days in the past month. The CDC-HRQOL-4 has been validated in several noninstitutionalized US populations, and performs well in populations with chronic conditions.16 We used the HRQOL-4 general health status measure, where respondents indicated their health as excellent, very good, good, fair, or poor (dichotomized as excellent/very good/good, or fair/poor), and summary measures of unhealthy physical and mental health days in the past month (Frequent Physical or Mental Distress = physical or mental health was not good for 15 or more days in the past 30 days).17
We generated demographic estimates for the cancer and control samples and examined cancer-specific variables by sex for the cancer survivors. We then assessed the relationship of survivor status with the 4 binary indicators of access to and utilization of health care (having no health insurance coverage, having no personal health care provider, having no routine check-up in the past year, and not seeing a doctor in the past year due to cost). Using multivariable generalized linear models with logit link and binomial family, we generated relative risks (RR) and 95% confidence intervals (95% CIs) to examine the association for these measures of interest.18 These models were run with the full cancer sample and then as secondary analyses using only the restricted cancer sample.
We also calculated adjusted proportions reporting the health care outcomes by age, race, and sex and, as relevant, insurance status, allowing for an interaction between survivor/control status and the measure of interest. Finally, in multivariable generalized linear models, we examined the association of health status with the health care measures among survivors, adjusting for time since diagnosis. Because BRFSS uses a complex sampling design, all results are reported weighted by BRFSS survey design and for nonresponse. We used the unconditional subpopulation approach available as part of the SVY command in Stata, release 12 (StataCorp, College Station, Tex) to correctly calculate the standard errors for our sample of interest.19 Statistical significance was α = 0.05. All results are based on the full sample unless otherwise reported.
AYA survivors, on average, were older than the control sample (33.1 years [standard deviation = 3.6 years] vs 30.0 years [standard deviation = 3.9 years]; P < .001). In Table 1, compared to controls, more AYA survivors were female (80% vs 50%; P < .001) and non-Hispanic white (75% vs 61%; P < .001). Survivors tended to be married more often and employed less often than the control sample. A total of 27% of survivors reported poor or fair health compared to 9% of the controls, and significantly greater proportions of survivors reported frequent physical or mental health distress compared with controls.
Among female survivors, the most common cancer was cervical (52.2%) followed by ovarian (9.2%) and melanoma (8.7%), whereas male survivors were more likely to report non-Hodgkin disease (21%), testicular (19.5%), and Hodgkin disease (13.2%) (Table 2). Although more women reported more than 1 cancer (9.8% vs 3.0% of men), men tended to be within 5 to 10 years of diagnosis at a higher proportion than females (68.4% vs 56.9%).
|Variable||Female AYA Survivors||Male AYA Survivors|
|N = 856||N = 123|
|N||Weighted %||N||Weighted %|
|Most recent diagnosis|
|Number of cancers|
|2 or more||89||9.8||6||3.0|
|Age at first diagnosis|
|Years since first diagnosis|
AYA Cancer Survivors Compared With Controls
Table 3 displays the proportions reporting barriers to health care and the multivariable regression estimates for the survivors compared with controls. The proportion reporting being uninsured was similar in unadjusted comparisons (21% of survivors vs 23% of those without cancer) and did not differ after adjustment in multivariable regressions. A total of 22% of the AYA survivors did not have a medical provider and 40% reported having no routine medical visits in the past year. Although somewhat lower than the controls (31% and 42%, respectively), these estimates were not statistically different and were not modified substantially in the adjusted regression analyses. However, in both the full and restricted sample regression models, survivors were more likely than the control group to report forgoing medical care due to costs in the past year (full sample RR = 1.67, 95% CI = 1.44-1.94; restricted sample RR = 1.55, 95% CI = 1.18-2.06).
|Uninsured||Not Having a Personal Medical Provider||Not Having Routine Medical Visit in Past Year||Forgoing Medical Care Due to Cost in Past Year|
|Unadjusted %||RR||95% CI||Unadjusted %||RR||95% CI||Unadjusted %||RR||95% CI||Unadjusted %||RR||95% CI|
|All AYA Cancer Survivors (N = 979) vs Controls (N = 67,216)|
|Restricted sampleb: AYA Cancer Survivors (N = 396) vs Controls (N = 67,216)|
Differences in Health Care Barriers by Insurance Coverage and Demographic Characteristics
Outcomes of having no provider, no routine care, and avoiding care due to cost are shown in Figure 1 by insurance status for the full sample. In adjusted proportions, more uninsured survivors compared to uninsured controls reported not having a medical provider (63% vs 56%, P = .01) or medical visit in the past year (73% vs 63%, P = .007). Moreover, 76% of uninsured survivors compared to 48% of the uninsured controls reported avoiding medical care in the past year due to costs (P < .001). Insured survivors also reported forgoing medical care more often than controls, but at much lower levels than the uninsured (21% of insured AYAs vs 11% of insured controls; P < .001).
We next compared adjusted proportions of insurance status within demographic groups to examine if barriers were more evident among specific survivor subgroups for the full sample (results not shown). Female survivors were uninsured more often (23%) than female controls (17%; P = .01), whereas the proportion uninsured did not differ significantly for males (22% vs 17%; P = .82). White survivors tended to be uninsured at higher levels than controls (22% vs 15%, P = .003). Black and other race survivors were uninsured at similar levels as white survivors (20% and 19%, respectively) and did not differ significantly from the controls, whereas Hispanic survivors had higher levels of being uninsured in both the survivor and control groups (31% and 36% respectively; nonsignificant).
There were no significant differences in not having a personal provider or routine visit by sex, age, or race, except that Hispanic survivors were less likely to have routine care than the Hispanic control group (66% vs 34%, P = .008; not shown). However, the AYA survivors differed substantially from controls by these demographic factors in forgoing care due to cost, even when results were adjusted for demographics and insurance status (Fig. 2). Specifically, 44% of AYAs aged 20 to 29 years and 26% of those aged 30 to 39 years reported avoiding care compared to 16% of controls in each of their respective age groups (P < .001 for each). A total of 22% of male survivors reported avoiding care due to costs, compared with 12% of controls, although this was not significantly different, whereas female AYAs did differ significantly as compared with controls (35% vs 18% of controls, P < .001). Thirty-three percent of white survivors versus 15% of the controls had cost barriers (P < .001). Although there were no significant differences for the other race/ethnicity categories, approximately 25% of black and other race survivors reported cost barriers, compared with <20% of the controls.
Health Status and Health Care Access Among AYA Survivors
The final set of models examined the relationship of health status (fair/poor health) and unhealthy days (≥15 days of poor physical or mental health per month) with the 4 health care access measures among survivors (Table 4). There were no significant associations with being uninsured or having no routine visits. In the fully adjusted model, survivors with 15 or more days of poor mental health were at greater risk for not having a health care provider (RR = 1.87, 95% CI = 1.09-3.24).
|N = 979||Uninsured||Not Having a Personal Medical Provider||Not Having Routine Medical Visit in Past Year||Foregoing Medical Care Due to Cost in Past Year|
|RR||95% CI||RR||95% CI||RR||95% CI||RR||95% CI|
|Excellent/very good/good (reference)||1||1||1||1|
|Days poor physical health|
|<15 days per month (reference)||1||1||1||1|
|≥15 days per month||0.92||0.56-1.51||0.69||0.42-1.14||1.10||0.84-1.45||1.49||1.17-1.91|
|Days poor mental health|
|<15 days per month (reference)||1||1||1||1|
|≥15 days per month||1.24||0.80-1.92||1.64||0.99-2.71||1.01||0.77-1.33||1.49||1.14-1.94|
|Excellent/very good/good (reference)||1||1||1||1|
|Days poor physical health|
|<15 days per month (reference)||1||1||1||1|
|≥15 days per month||0.79||0.45-1.37||0.58||0.31-1.05||0.99||0.73-1.36||1.16||0.83-1.62|
|Days poor mental health|
|<15 days per month (reference)||1||1||1||1|
|≥15 days per month||1.23||0.77-1.97||1.87||1.09-3.24||0.97||0.73-1.29||1.31||1.00-1.73|
AYA survivors reporting poor health were approximately 50% to 55% more likely to forgo care due to cost in the past year in models examining each measure independently. When the 3 factors were examined together in the full multivariable regression model, having 15 or more days of poor mental health per month remained marginally significant (RR = 1.31, 95% CI = 1.00-1.73), although the association was attenuated.
Young adult survivors of AYA cancers require consistent medical surveillance due to the long-term effects from cancer treatment. In our population-based assessment using national data from the 2009 BRFSS, we found that long-term AYA cancer survivors report substantial access- and cost-related barriers to utilizing health care, despite having comparable levels of health insurance coverage as similarly-aged adults without cancer. To our knowledge, this study is the first to examine general health care access and use among long-term survivors of AYA cancers, and suggests that improvements in posttreatment health care access must be prioritized for this population.
Access to care was poor for many AYA survivors. More than one-fifth of long-term AYA survivors did not have a medical provider, and 40% reported having no routine medical visits in the past year. Although these estimates did not differ significantly from the control sample in multivariable analyses, these findings are cause for concern in light of the health problems faced by many AYA cancer survivors that necessitate ongoing surveillance and/or care, and demonstrate a lower level of access to medical care than other survivors.6
Cost barriers to medical care were particularly common for survivors, because they were 55% to 67% more likely to report avoiding medical care in the past year due to cost than the controls. Uninsured AYA survivors faced the largest cost barriers to care, with 74% reporting forgoing care due to cost in the past year, substantially higher than the 54% reported in an earlier population-based assessment of uninsured cancer survivors aged 18 to 65 years.6 Although insured AYA survivors reported fewer cost barriers to care than did uninsured survivors, they were still twice as likely as the control group to forgo care due to cost in the past year. The economic costs of receiving health care for many with cancer go beyond out-of-pocket medical costs, and include other costs such as transportation, childcare, and lost wages.20 Even with insurance, the copayments and other out-of-pocket costs may be burdensome to AYA survivors, particularly for those of low income or in poor health who are high users of medical care, and this demonstrates the need to examine how young cancer survivors interpret medical cost and make decisions about when to access medical care.
Although adolescents and young adults are less likely to be insured and more likely than older adults to avoid medical care services,21 our findings suggest that subgroups of AYA survivors may be at additional risk for experiencing health care barriers beyond these generational differences in care. More than two-thirds of uninsured survivors had no personal provider and no routine medical care. Both younger survivors and female survivors reported forgoing medical care in the past year due to cost concerns at a higher level than controls, even in estimates that adjusted for health insurance coverage. Hispanic survivors had the lowest level of insurance among the survivors, and two-thirds of this racial/ethnic group reported having no routine medical care in the past year. Moreover, survivors who reported their health status as fair or poor, or who stated they had 15 or more days of poor mental or physical health per month, reported cost barriers to getting medical care.
AYA survivors in the United States will benefit from the Patient Protection and Affordable Care Act (ACA), because it allows young adults to stay on their parents' health plans until the age of 26, provides insurance plans for those with preexisting conditions, eliminates limits on coverage, and prohibits rescissions of coverage.22 By 2014, state-run insurance exchanges and the Medicaid expansion should provide insurance access to many survivors with chronic health limitations.22, 23 However, the ACA may prove insufficient for at-risk populations such as AYA survivors who require a high level of care. Although the state-run exchanges will provide coverage for individuals with preexisting conditions without access to employer plans,22 the cost sharing required by these plans may be burdensome for survivors with high out-of-pocket costs.24 Furthermore, the Supreme Court's ruling that individual states have the option not to expand their Medicaid programs, jeopardizes health insurance options for the lowest income AYA survivors. Although BRFSS respondents were not asked to report their type of health insurance coverage, it is likely that more AYA survivors are on public health insurance compared to those without a cancer history, similar to other populations of younger cancer survivors.8
Although the ACA is a substantial step toward health insurance parity in the United States, the barriers for AYA survivors go beyond expanding insurance access. AYA populations with the biggest barriers to care, such as uninsured, younger, and nonwhite survivors, need additional support as they transition out of the oncology setting, such as information on community health centers and financial assistance programs, to improve their access to care in the years following their cancer treatment. Also, community providers are often unaware of the health risks associated with AYA cancer. Although we did not have information on whether AYA survivors are receiving survivor-specific medical services (eg, screening for secondary cancers), the low level of general medical use found among the BRFSS AYA survivors suggests that many AYA survivors may require education about their long-term health care needs.25
Certain limitations should be considered in the interpretation of the findings. The response rate for BRFSS in 2009 was 53%,26 potentially affecting the generalizability of our findings to the overall AYA survivor population. Some reports have found that BRFSS may underrepresent hard-to-reach populations, such as minority populations and younger adults,27 suggesting our findings may underestimate the health care barriers in this population. Assessing cancer history via self-report may risk misclassification of cancers and noncancers, although previous studies of BRFSS have found cancer self-report reasonably reliable.15 In addition, we were unable to examine cancer-related medical visits; although some survivors did not report routine care, we may be under-assessing medical care use if some are still seeing their oncology provider. However, long-term cancer survivors should still be receiving primary care services to screen for other conditions, and therefore, this low level of such services is of concern. Also, the controls in our sample may be under-reporting cost barriers to care if they are healthier and do not perceive the need to obtain primary care services.
BRFSS does not include cancer treatment or other cancer-related information; thus, we were limited in exploring aspects such as disease stage at diagnosis or whether survivors were undergoing current cancer-related treatment. However, by focusing on survivors at 5 or more years beyond their first primary diagnosis, the vast majority of our sample should have completed active cancer treatment. In addition, the large number of cervical cancers may be due to female survivors reporting cervical cancer if they had atypical cells or in situ abnormalities, which may have required minimal therapy. However, the subanalyses wherein cervical, thyroid, testicular, and other cancers were removed allows for some control of differences in cancer or treatment severity, and these models were generally similar to our analyses of the full sample. Finally, because we examined several outcomes of health care access, we advise that these results should be interpreted cautiously in light of the multiple comparisons made.
Survivors of AYA cancer are a growing population in the United States with unique, long-term health care needs. Research to improve outcomes for this population is a national priority.28 Although adolescents and young adults in the United States, in general, have lower levels of insurance coverage and health care access, this population-based study shows that many AYA cancer survivors are not receiving primary medical care services and report cost barriers to care at levels beyond these general age differences. Prospective studies are necessary to determine how a cancer diagnosis during adolescence and young adulthood affects education, employment, and other resources, all of which play a vital role in influencing survivors' access to health care services. Given the need to monitor AYA cancer survivors for recurrence and late effects of treatment, increasing the access to health care for this population is of public health importance. Although AYA survivors should be educated about their health risks following cancer treatment, expanding health insurance coverage for young survivors may prove insufficient to increasing their health care access without adequate strategies to reduce their medical cost burdens.
Anne Kirchhoff is supported by the Huntsman Cancer Institute/Huntsman Cancer Foundation.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 1National Cancer Institute. A Snapshot of Adolescent and Young Adult Cancers. http://nci.nih.gov/aboutnci/servingpeople/snapshots/AYA.pdf. Updated October 2011.
- 2Canadian Cancer Society Steering Committee. Canadian Cancer Statistics 2009, Special Topic: Cancer in Adolescents and Young Adults. Toronto, Canada: Canadian Cancer Society; 2009.
- 4Cancer Epidemiology in Older Adolescents and Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival: 1975–2000. NIH Publication No. 06-5767. Bethesda, MD: National Cancer Institute; 2006., , , , editors.
- 11Rite of Passage? Why Young Adults Become Uninsured and How New Policies Can Help, 2009 Update. New York, NY: The Commonwealth Fund; 2009., , , , , .
- 12Employer-Sponsored Health Insurance: Trends in Cost and Access. AHRQ Publication No. 04-0085. Rockville, MD: Agency for Healthcare Research and Quality; 2004..
- 17Centers for Disease Control and Prevention. Measuring Healthy Days: Population Assessment of Health-Related Quality of Life. CDC: Atlanta, GA; 2000.
- 18Applied Logistic Regression.2nd ed. New York, NY: John Wiley & Sons; 2000., .
- 19Applied Survey Data Analysis. Boca Raton, FL: Chapman & Hall/CRC; 2010., , .
- 20Economic impact of cancer in the United States. In: Schottenfeld D, Fraumeni JFJr, editors. Cancer Epidemiology and Prevention.3rd edition. New York, NY: Oxford Univ Press; 2006: 202-214., .
- 24Realizing Health Reform's Potential: Will the Affordable Care Act Make Health Insurance Affordable? New York, NY: The Commonwealth Fund; 2011., .
- 26Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System 2009 Summary Data Quality Report. CDC: Atlanta, GA; 2011.
- 28National Cancer Institute. Closing the Gap: Research and Care Imperatives for Adolescents and Young Adults with Cancer. Report of the Adolescent and Young Adult Oncology Progress Review Group. Bethesda, MD: National Cancer Institute; 2006.