This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Martin, S. Articles by Bleyer, A. Search for Related Content PubMed PubMed Citation Articles by Martin, S. Articles by Bleyer, A.

Epidemiology and Population Studies

Delays in Cancer Diagnosis in Underinsured Young Adults and Older Adolescents

^aBaylor College of Medicine, Houston, Texas, USA; ^bUniversity of Rochester, Rochester, New York, USA; ^cUniversity of Texas School of Public Health, Houston, Texas, USA; ^dUniversity of Texas M.D. Anderson Cancer Center, Houston, Texas, USA; ^eChildren's Oncology Group, Arcadia, California, USA

Key Words. Cancer • Diagnosis • Health insurance • Health services research • Adolescents

Correspondence: Sean Martin, M.D., University of Colorado Health Sciences Center, 7370 East Florida Avenue #1009, Denver, Colorado 80231, USA. Telephone: 775-771-0201; Fax: 541-385-6341; e-mail: Sean.Martin{at}uchsc.edu

Received February 8, 2007; accepted for publication April 23, 2007.

	Learning Objectives

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
After completing this course the reader will be able to:

1. Describe the relative lack of progress in survival improvement in young adults and older adolescents with cancer.
   
2. Explain the variability in the time to diagnosis of cancer among young adults and older adolescents.
   
3. Identify the potential role of health insurance in determining the time to cancer diagnosis in young adults and older adolescents.
   
4. Discuss the impact that universal health insurance may have in shortening the time to cancer diagnosis and treatment in young adults and older adolescents.
   

	ABSTRACT

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
Background. In the U.S., adolescents and young adults diagnosed with cancer have had less survival improvement than older or younger patients, a deficit that may be a result of delays in diagnosis in an age group with the lowest rates of health insurance.

Methods. The relationship between health insurance status and the time from the onset of first cancer-specific symptom or sign to definitive diagnosis (lagtime) was retrospectively compared with other sociodemographic factors in newly diagnosed cancer patients aged 15–29 years who were evaluated between June 2001 and June 2003. Data on 270 patients with the six most common cancer types in this cohort (leukemia, Hodgkin's and non-Hodgkin's lymphoma, sarcoma, brain tumors, thyroid cancer) were retrospectively collected in 2004.

Results. Lagtimes were evaluable in 235 (88%) patients. In multivariate analysis, the type of cancer and health insurance were significantly associated with lagtime, whereas race/ethnicity, age, gender, marital status, and surrogate measures of socioeconomic status were not. The mean lagtime in patients with public or no health insurance was 13.1 weeks longer than in patients with private health insurance, and longer in four of six evaluable histology-specific types of cancer. In cancers evaluable for stage at diagnosis, advanced stage was associated with longer lagtimes.

Conclusion. In the U.S., older adolescents and young adults with cancer are likely to have a delay in diagnosis because of inadequate health insurance and consequently present with a more advanced stage of disease.

Disclosure of potential conflicts of interest is found at the end of this article.

	INTRODUCTION

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
Despite the advances in cancer diagnosis and treatment, older adolescents and young adults with cancer in the U.S., especially those in the 20–39 years age group, have not experienced the same survival prolongation or mortality reduction as their older and younger counterparts [1, 2]. Reasons for the failure to improve survival in this cohort may be related to delays in diagnosis, which are known to occur more frequently in cancer patients in this age group than in others [3–11]. Also, the 18–24 years age group is both the most underinsured in the U.S. [12–16] as well as the least likely to access health care [17], and the 25–34 years age group is the next most uninsured age group [12]. The hypothesis of this study, therefore, was that a lack of health insurance is associated with delay in diagnosis in this age group. Accordingly, we determined the interval from symptom onset to diagnosis (lagtime) in newly diagnosed cancer patients who were 15–29 years of age when diagnosed, the same age group for which the epidemiology and outcomes of cancer were recently described [18]. The objective was to assess the relationship between the diagnosis lagtime and health insurance status in cancer patients who, historically, have had the least improvement in cancer survival as well as the least health insurance.

	METHODS

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
A retrospective analysis was performed on 503 patients aged 15–29 years who were evaluated for a previously untreated cancer at the University of Texas M.D. Anderson Cancer Center (MDACC) in Houston, Texas, between June 1, 2001 and May 31, 2003 [19, 20]. Following institutional review board approval, study participants were identified via the MDACC Tumor Registry. After review of the available patients and their types of cancer, the six most common malignancies in the age group were selected for analysis: leukemia, non-Hodgkin's lymphoma, Hodgkin's lymphoma, sarcomas, brain tumors, and thyroid cancer. Patients with the next most frequent cancer, carcinoma of the uterine cervix, were inevaluable for lagtime because most had no symptoms and were diagnosed by Pap smear screening. International patients were also excluded because of the varying nature of their health care systems and travel times to the U.S. Patient data were collected in 2004 and included the primary symptom leading to a cancer diagnosis, lagtime (defined below), cancer type and stage, and patient age, gender, race/ethnicity, marital status, zip code of residence, and health insurance status when referred to MDACC.

Lagtime
Diagnosis lagtime was defined as the interval between the onset of evidence of an existing malignancy (symptom or sign) and the date on which a specific cancer type was verified by pathological confirmation. Cancer-specific symptoms reported as beginning a week, month, or year ago were given lagtime values of 7 days, 30 days, and 365 days, respectively. Symptoms reported to have started in a given month or season were assumed to have begun at the midpoint of the month or season.

Health Insurance Status
Health insurance status was determined from billing data and grouped into three tiers of coverage: public insurance, private insurance, and self-pay. Public insurance included state/national programs (Medicaid), county programs, and no insurance (indigent patients). Indigent patients were determined by MDACC to be unable to pay for a cancer-related visit costing more than $500 and to have an income at or below 185% of the federal poverty guidelines based on the patient's previous income tax return. Private insurance included managed care plans, health maintenance organizations, and other group insurance plans. Although military insurance is public, patients with Champus were grouped in the private insurance category because of the inherent lack of barriers to care for military personnel, particularly in the young adult age group. Self-pay patients included those who paid with personal or private funds. Self-pay patients were defined as those without health insurance who were responsible for 50%–100% of their cost of care as determined by income. These patients were classified as having the highest tier of health care coverage because of shorter waiting times given the absence of paperwork and verification from third-party payors.

Socioeconomic Status
The patient's zip code of residence and U.S. census 2000 data determined the median household income, population density, and percentage of people living in urban areas. Population density was defined by the percentage of urban households: low, <33%; intermediate, 33%–66%; high, >66%.

Other Demographic Data
Race/ethnicity, marital status, and religion were derived from a questionnaire completed by the patient during his/her first hospital visit.

Stage of Disease
Disease stage was determined in accordance with American Joint Committee on Cancer and Ann Arbor Staging System criteria. Extensive regional or distant metastases at diagnosis of sarcoma and thyroid cancer were regarded as advanced disease. For Hodgkin's lymphoma, available data included presence or absence of symptoms at diagnosis, with the latter occurring when biopsy of a painless mass led to the diagnosis.

Statistical Analyses
As is conventional for lagtime measures, all analyses were performed on the logarithms of lagtimes (geometric means and variances) and converted back to original units of measurement (days), while maintaining the same coefficient of variation as on the log scale, to allow for more meaningful interpretation. For comparison of two group, t-tests and Mann-Whitney tests were used, with one-tailed p-values accepted for insurance comparisons based on the expectation that lagtimes would be longer in underinsured groups. The Kruskal-Wallis test was used to assess differences among three or more groups. A standard multiple linear regression model was applied to the logarithm of the lagtimes, with a stepwise variable selection procedure used to eliminate factors from the model until only those factors that were statistically significant at the 0.05 significance level remained. With a sample of 235 patients and an F-test at a significance level of 0.05, the model had 88% power to detect an R² of 0.03 attributable to public insurance versus other insurance while adjusting for the additional independent variables. The three additional independent variables included in the model define type of disease and yielded a model R² of 0.28. Because the study was not powered to correlate disease stage with lagtime, statistical analyses were not performed on disease stage or symptom data.

	RESULTS

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
Initial evaluation yielded 268 eligible participants with the six most common cancers in the age group, of which 235 (88%) had evaluable lagtimes (Table 1). Of the 235 patients evaluated, 121 (51%) were male and 114 (49%) were female. The mean age at diagnosis was 22.2 years. Seventeen (7%) of the patients had no insurance, 33 patients (14%) had public insurance, 34 patients (15%) were self-pay, and 151 (64%) had or used private insurance. The latter included five patients with military insurance as explained in the Methods section. The mean (± standard deviation [SD]) ages were comparable in the self-pay, private, and public insurance groups: 21.2 (±4.0), 22.4 (±4.3), and 22.4 (±4.3) years, respectively. Of the 75 patients aged 15–19 years, 12 (16%) were self-pay patients, 49 (65%) had private insurance, and 14 (19%) had public insurance. Among patients aged 20–24 years, 14 (18%) were self-pay, 44 (57%) were privately insured, and 19 (25%) were publicly insured. In patients aged 25–29 years, 8 (10%) were self-pay whereas 58 (70%) and 17 (20%) had private insurance and public insurance, respectively.

Multivariate analysis showed that only the type of cancer and status of health insurance prior to diagnosis were significantly associated (p < .01) with the lagtime to diagnosis (Table 2). Non-Hodgkin's lymphoma (p < .0001), Hodgkin's disease (p = .0007), and leukemia (p < .0001) were each predictive of shorter lagtimes, while public insurance (p = .0021) was predictive of longer lagtimes. No differences were noted for gender, race/ethnicity, marital status, median household income of zip code of residence, population density of residence, or age (within the 15–29 year range).

Excluding leukemia and brain tumors, for which staging is not applicable, four histology-specific subgroups could be evaluated for stage at diagnosis: diffuse large cell non-Hodgkin's lymphoma, Hodgkin's lymphoma, osteosarcoma, and follicular thyroid carcinoma including mixed papillary-follicular (Table 4). All four had longer lagtimes among advanced-stage patients than among patients with localized disease, with greater delays of 17 (p = .095), 19 (p < .01), 259 (p < .01), and 195 days (p < .01), respectively. In Hodgkin's lymphoma, the absence or presence of symptoms at diagnosis could be evaluated because 22 of 46 (non–self-pay) patients had biopsies of asymptomatic masses. Among patients with stage I or II Hodgkin's lymphoma, the lagtime was 35 days longer in symptomatic patients than in asymptomatic patients (p = .16). The corresponding difference in lagtime between symptomatic and asymptomatic patients with stage III or IV Hodgkin's lymphoma was 17 days (p = .45). The study was not designed (powered) to assess correlations between stage or symptoms and lagtime.

	Discussion

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

As a single-institution study, our results may not be generalizable. The lagtimes we observed, however, are consistent with those in the largest study conducted to date, reported by Pollock and associates in the Pediatric Oncology Group [4]: 16 weeks in both their study and ours for bone sarcomas, 14 weeks in both studies for brain tumors, 17 weeks in their study versus 9 weeks in ours for Hodgkin's lymphoma, 7 versus 3 weeks for leukemia, and 9 versus 5 weeks for non-Hodgkin's lymphoma. In addition to a lack of generalizability, several cancers in our study had small numbers of patients (e.g., 19 each for acute myeloid leukemia and non-Hodgkin's lymphoma). Nonetheless, the observed relationship between lagtime and insurance status was consistent across all but one of the cancer types and histologies evaluated. Because this study was not designed a priori to evaluate associations between lagtimes and either stage or symptoms at diagnosis, one cannot conclude from the data that longer lagtimes result in more advanced disease or, ultimately, worse outcomes.

A number of studies have reported a direct correlation between lagtime and age in the pediatric age range, with adolescents experiencing longer lagtimes than children [3, 4, 8–10]. Greater lagtimes have been observed in white American adolescents [4]. In Israel, significant correlations between lagtime and religion, father's ethnicity (Arabic children had shorter lagtimes than children of Jewish parents), and location of residence have been observed [10]. Non-Hispanic whites had the longest interval to diagnosis in our study, although neither race/ethnicity nor marital status was significantly correlated with lagtime. The median household income within the residence zip code was not as strongly correlated with lagtime as was insurance status, which may be a result of the use of a zip code proxy [22] rather than the actual individual household income. Education and employment status, two factors also likely to influence lagtime, were not evaluated in our study.

In the malignancies for which we had data on stage at diagnosis, stage was directly correlated with lagtime, albeit not at a statistically significant level (our study was not powered to evaluate this relationship). In Europe, population-based studies of breast cancer and osteosarcoma patients demonstrated a direct correlation between lagtime and stage at diagnosis [23, 24]. The implication is that the delays in diagnosis lead to more advanced cancer at the time of diagnosis, a potentially worse outcome, and a requisite increase in therapy, whether more multimodal or of greater intensity or duration. Using osteosarcoma as an example, the patient who presents with pulmonary metastases and is thought to be operable not only usually undergoes lung resection(s) in an attempt to remove the metastases, but more aggressive chemotherapy is usually given, and despite the additional therapy generally has less than half the survival rate of patients who are diagnosed with localized disease. In Hodgkin's lymphoma, the survival of patients with stage III disease may not be different from those with stage I or II disease as it is in osteosarcoma, but the more advanced patients receive far more treatment, usually combining chemotherapy and radiation therapy, for a longer duration, and with greater likelihood of adverse late effects such as infertility and another malignant neoplasm. Stage IV Hodgkin's patients have a worse survival, even with bone marrow transplantation and high-dose chemotherapy and total body or total nodal irradiation.

In certain cancers, the symptoms may be too nonspecific to lead to an early diagnosis, such as in brain tumors [25, 26]. In our study, brain tumors were one of only two types of cancer in which a significant correlation between lagtime and insurance status was not apparent. In osteosarcoma, separate studies have reported an inverse relationship and no correlation between lagtime and tumor spread [27, 28]. In Europe, a median lagtime for osteosarcoma was reported to be 69 days [27], comparable with that observed in this study among patients with private insurance, but much shorter than that noted in patients with public insurance (Table 3). Of interest, because universal health insurance is available throughout Europe, the fact that the European lagtime matches the U.S. lagtime for (privately) insured persons not only validates the data reported herein, but also underscores the disadvantage of noninsured or poorly insured (publicly) persons in the U.S. who are at risk for substantially longer times to diagnosis of a fatal disease that is potentially curable in its earlier phase.

Thus, what happens after diagnosis of cancer with respect to further delays in staging and treatment initiation may also depend on insurance status. A lack of insurance coverage is associated with less expenditures on medical evaluation and treatment among cancer patients, suggesting that cancer patients lacking health insurance are more likely to receive less care for financial reasons [29]. In a study of breast cancer, minority women did not have a longer interval to therapy initiation [30], but in a larger, more recent analysis, the degree to which standardized guidelines for cancer treatment were followed was dependent on the quality of health insurance [31]. The study reported here was not designed to evaluate relationships between lagtime and survival or other treatment outcomes after diagnosis.

The delay in cancer diagnosis among young adults between 15 and 30 years of age may result in a more advanced stage of cancer and a worse prognosis and outcome. With only a 3–4 week difference in diagnostic lagtime observed between patients presenting with localized lymphoma and those presenting with metastatic lymphoma, and a 3–4 week difference between the mean lagtime of those patients with Hodgkin's lymphoma who were symptomatic at presentation and that of those who were not, significant clinical progression of the more rapidly growing cancers may take place during relatively short delays in diagnosis. That osteosarcoma and thyroid carcinoma also tended to have more advanced cancer in association with longer lagtimes implies that solid tumor patients who have the longest lagtimes are similarly subject to delays resulting from less health insurance. The insurance effect may thereby apply to many cancer types, regardless of rapidity of growth and clinical detection.

The prognosis of a given type of cancer may decline significantly as a function of age, as exemplified by acute lymphoblastic leukemia [32]. The age dependence may be a result of changes in the biology of the tumor such that it is intrinsically more therapy resistant at older ages. Another reason for the worse prognosis as a function of age may be health insurance, which in the U.S. also worsens with age such that as young Americans "age out" of insurance coverage they become at greater risk for diagnosis delays and a worse outcome (and/or greater treatment needs). In our study, the multivariate analysis indicated that type of cancer and health insurance status were more important than race or ethnicity, gender, household income per se, marital status, place of residence, or age within the 15- to 29-year range. To what extent health insurance may explain the previously observed inverse correlation between age and prognosis in individual types of cancer remains to be determined.

In summary, the age group in the U.S. with the highest rate of uninsured and underinsured persons appears to be at risk for delays in cancer diagnosis, a disease in which delays have been associated with more advanced presentations, a requirement for more therapy, and a likelihood of a worse outcome. Other diseases are likely to be similarly affected, and a solution is needed to improve health insurance coverage of young people in the U.S., particularly those intrinsically most vulnerable to delays in diagnosis: young adults and older adolescents.

	DISCLOSURES

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
The authors indicate no potential conflicts of interest.

	ACKNOWLEDGMENTS

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 
We are indebted to the members of the Children's Oncology Group Adolescent and Young Adult Committee and the Aflac Insurance Co. for their support of and contributions to this investigation.

This work was supported by grants from the National Cancer Institute (NCI U10 CA98543) and Aflac Insurance Co.

Presented in part at the Annual Meeting of the American Society of Clinical Oncology, Orlando, Florida, May 15, 2005, and at the 10th Annual Meeting of the International Society of Pharmacoeconomics and Outcomes Research, Washington, DC, May 17, 2005.

S.M. is now affiliated with University of Colorado Health Sciences Center, Denver, CO; A.B. is now affiliated with St. Charles Medical Center, Bend, OR, and CureSearch/National Childhood Cancer Foundation, California.

	REFERENCES

Top Learning Objectives Abstract Introduction Methods Results Discussion Disclosures Acknowledgments References

 

1. Hampton T. Cancer treatment's trade-off: Years of added life can have long-term costs. JAMA 2005;294:167–168.[Free Full Text]
2. National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER) Program. SEER*Stat Database: Incidence - SEER 13 Regs Public-Use, Nov 2004 Sub for Expanded Races (1992–2002), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Branch, released April 2005, based on the November 2004 submission. Available at http://www.seer.cancer.gov. Accessed January 25, 2007.
3. Saha V, Love S, Eden T et al. Determinants of symptom interval in childhood cancer. Arch Dis Child 1993;68:771–774.[Abstract/Free Full Text]
4. Pollock BH, Krischer JP, Vietti TJ. Interval between symptom onset and diagnosis of pediatric solid tumors. J Pediatr 1991;119:725–732.[CrossRef][Medline]
5. Goddard AG, Kingston JE, Hungerford JL. Delay in diagnosis of retinoblastoma: Risk factors and treatment outcome. Br J Ophthalmol 1999;83:1320–1323.[Abstract/Free Full Text]
6. Butros LJ, Abramson DH, Dunkel IJ. Delayed diagnosis of retinoblastoma: Analysis of degree, cause, and potential consequences. Pediatrics 2002;109:E45.[CrossRef][Medline]
7. Goyal S, Roscoe J, Ryder WDJ et al. Symptom interval in young people with bone cancer. Eur J Cancer 2004;40:2280–2286.[CrossRef][Medline]
8. Pratt CB, Smith JW, Woerner S et al. Factors leading to delay in the diagnosis and affecting survival of children with head and neck rhabdomyosarcoma. Pediatrics 1978;61:30–34.[Abstract/Free Full Text]
9. Flores LE, Williams DL, Bell BA et al. Delay in the diagnosis of pediatric brain tumors. Am J Dis Child 1986;140:684–686.[Abstract/Free Full Text]
10. Haimi M, Peretz Nahum M, Ben Arush MW. Delay in diagnosis of children with cancer: A retrospective study of 315 children. Pediatr Hematol Oncol 2004;21:37–48.[CrossRef][Medline]
11. Klein-Geltink J, Pogany L, Mery LS et al. Impact of age and diagnosis on waiting times between important treatment events among children 0 to 19 years cared for in pediatric units: The Canadian Childhood Cancer Surveillance and Control Program. J Pediatr Hematol Oncol 2006;28:433–439.[CrossRef][Medline]
12. Steinbrook R. Health care reform in Massachusetts—a work in progress. N Engl J Med 2006;354:2095–2098.[Free Full Text]
13. Collins SR, Schoen C, Tenney K et al. Right of Passage? Why Young Adults Become Uninsured and How New Policies Can Help. The Commonwealth Fund [monograph online]. Available at http://www.cmwf.org/usr_doc/collins_ritepassage2006_649_ib.pdf. Accessed June 13, 2007.
14. General Accounting Office. Health Insurance: Characteristics and Trends in the Uninsured Population. 2001, 1, 21, General Accounting Office, Washington, DC, Available at http://www.gao.gov/new.items/d01507t.pdf. Accessed January 25, 2007.
15. Fishman E. Aging out of coverage: Young adults with special health needs. Health Aff 2001;20;(6):254–266.[Abstract/Free Full Text]
16. White PH. Access to health care: Health insurance considerations for young adults with special health care needs/disabilities. Pediatrics 2002;110:1328–1335.[Abstract/Free Full Text]
17. Office of Technology Assessment. Adolescent Health: Summary and Policy Options, Volume I. 1991, 1, 196, United States Congress, Washington, DC, Available at http://govinfo.library.unt.edu/ota/Ota_2/DATA/1991/9102.PDF. Accessed January 30, 2007.
18. Bleyer WA, O'Leary M, Barr R et al, Cancer Epidemiology in Adolescents and Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival, 1975–2000. 2006, 1, 205, National Cancer Institute, Bethesda, MD, Available at http://www.seer.cancer.gov/publications/aya. Accessed January 30, 2007.
19. Bleyer A, Ulrich C, Martin S et al. Status of health insurance predicts time from symptom onset to cancer diagnosis in young adults. Proc Am Soc Clin Oncol 2005;23(suppl 16):547s.
20. Martin S, Ulrich C, Munsell M et al. Time to cancer diagnosis in young Americans depends on type of cancer and health insurance status. Value in Health 2005;8;(1):344a.
21. Gwyn K, Bondy ML, Cohen DS et al. Racial differences in diagnosis, treatment, and clinical delays in a population-based study of patients with newly diagnosed breast carcinoma. Cancer 2004;100:1595–1604.[CrossRef][Medline]
22. Geronimus A, Bound J, Wagner E. On the validity of using census geocode data to proxy individual socioeconomic characteristics. J Am Stat Assoc 1996;91:529–537.[CrossRef]
23. Arndt V, Sturmer T, Stegmaier C et al. Patient delay and stage of diagnosis among breast cancer patients in Germany – a population based study. Br J Cancer 2002;86:1034–1040.[CrossRef][Medline]
24. Bielack SS, Kempf-Bielack B, Delling G et al. Prognostic factors in high-grade osteosarcoma of the extremities or trunk: An analysis of 1,702 patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols. J Clin Oncol 2002;20:776–790.[Abstract/Free Full Text]
25. Edgeworth J, Bullock P, Bailey A et al. Why are brain tumours still being missed? Arch Dis Child 1996;74:148–151.[Abstract/Free Full Text]
26. Halperin EC, Watson DM, George SL. Duration of symptoms prior to diagnosis is related inversely to presenting disease stage in children with medulloblastoma. Cancer 2001;91:1444–1450.[CrossRef][Medline]
27. Bacci G, Ferrari S, Longhi A et al. High-grade osteosarcoma of the extremity: Differences between localized and metastatic tumors at presentation. J Pediatr Hematol Oncol 2002;24:27–30.[CrossRef][Medline]
28. Petrilli AS, de Camargo B, Filho VO et al. Results of the Brazilian Osteosarcoma Treatment Group Studies III and IV: Prognostic factors and impact on survival. J Clin Oncol 2006;24:1161–1168.[Abstract/Free Full Text]
29. Thorpe KE, Howard D. Health insurance and spending among cancer patients. Health Aff (Millwood) 2003, Suppl Web Exclusives):W3-189–W3-198.
30. Kerner JF, Yedidia M, Padgett D et al. Realizing the promise of breast cancer screening: Clinical follow-up after abnormal screening among black women. Prev Med 2003;37:92–101.[CrossRef][Medline]
31. Harlan LC, Greene AL, Clegg LX et al. Insurance status and the use of guideline therapy in the treatment of selected cancers. J Clin Oncol 2005;23:9079–9088.[Abstract/Free Full Text]
32. Mattano L, Nachman J, Ross J et al. Leukemias. Cancer Epidemiology in Older Adolescents and Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival, 1975–2000. 2006, 39, 51, National Cancer Institute, Bethesda MD, Available at http://www.seer.cancer.gov/publications. Accessed January 30, 2007.

This article has been cited by other articles:

				D. Pulte, A. Gondos, and H. Brenner Expected long-term survival of patients diagnosed with acute myeloblastic leukemia during 2006-2010 Ann. Onc., February 1, 2010; 21(2): 335 - 341. [Abstract] [Full Text] [PDF]

				D. Pulte, A. Gondos, and H. Brenner Improvement in survival in younger patients with acute lymphoblastic leukemia from the 1980s to the early 21st century Blood, February 12, 2009; 113(7): 1408 - 1411. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Martin, S. Articles by Bleyer, A. Search for Related Content PubMed PubMed Citation Articles by Martin, S. Articles by Bleyer, A.