ADVANCED SEARCH
Keyword »
Topic »
Author »
Date »
to
JULY / AUGUST 2014 :: 75(4)
Cancer in North Carolina

Cancer is the leading cause of death in North Carolina. This issue of the NCMJ discusses cancer prevention, screening, treatment, and survivorship; disparities in incidence and mortality; and ethics of clinical trials. Highlighting the importance of comprehensive data for understanding cancer, original articles in this issue address how medical homes can reduce health care utilization among breast cancer patients and how distance to care affects receipt of radiation therapy.

ORIGINAL ARTICLE

Health Care Utilization From Chemotherapy-Related Adverse Events Among Low-Income Breast Cancer Patients: Effect of Enrollment in a Medical Home Program

Ravi K. Goyal, Stephanie B. Wheeler, Racquel E. Kohler, Kristen H. Lich, Ching-Ching Lin, Katherine Reeder-Hayes, Anne-Marie Meyer, Deborah K. Mayer

N C Med J. 2014;75(4):231-238.PDF | TABLE OF CONTENTS



Background Chemotherapy-related health care utilization by breast cancer patients can be expensive for payers and patients. This study evaluated the patient-centered medical home program Community Care of North Carolina (CCNC) in terms of its potential to reduce health care utilization associated with chemotherapy-related adverse events (AEs).

Methods Early-stage breast cancer cases diagnosed during the 5-year period 2003–2007 were identified in the North Carolina Central Cancer Registry; these cases were then linked to North Carolina Medicaid claims data. We measured health care utilization associated with chemotherapy-related AEs by setting (inpatient, outpatient, or emergency department) during a 15-month postdiagnosis follow-up period. Descriptive and multivariate analyses were performed to examine associations between CCNC enrollment and health care utilization associated with chemotherapy-related AEs.

Results A large proportion of breast cancer patients had at least 1 health care visit associated with a chemotherapy-related AE (n = 412 [72.3%]). The mean numbers of AE-related visits occurring in inpatient, outpatient, and emergency department settings were 0.30 (standard deviation [SD] = 0.83), 6.92 (SD = 10.94), and 0.46 (SD = 1.26), respectively. CCNC enrollment was associated with significantly fewer inpatient admissions (marginal effect, –0.1421; 95% confidence interval, –0.280 to –0.004).

Limitations In this observational study, we were unable to draw conclusions about the causality of these associations.

Conclusions Patients enrolled in CCNC had fewer inpatient health care visits associated with chemotherapy-related AEs. Future research should continue to explore the extent to which patient-centered medical homes can monitor and help manage the effects of cancer treatments.

Breast cancer, which is the most common type of cancer in women, made up more than one-third of all new cancers diagnosed among women in North Carolina in 2011 [1]. The 5-year incidence rate of female breast cancer per 100,000 population in North Carolina increased from 148.2 in 2001 to 157.4 in 2011; during the same period, the rate of mortality from breast cancer decreased from 26.5 to 22.8 per 100,000 persons [1]. The growing incidence of breast cancer in North Carolina, the increasing number of breast cancer survivors, and the aging of the US population mean that the demand for breast cancer care will likely increase, and we must consider how to ensure that the health care system is prepared to respond to this demand.

Although most of the health care utilization by cancer patients is associated with diagnostic and therapeutic management of the disease, management of chemotherapy-related adverse events (AEs) also requires substantial resources [2]. Several studies have shown that treatment-related AEs are associated with substantial costs resulting from higher rates of inpatient and emergency department (ED) admissions [3-5]. A few observational studies of breast cancer patients have shown that chemotherapy-related AEs have an enormous impact on health care expenditures and utilization [6, 7].

The effective management of chemotherapy-related AEs can be facilitated by community care management programs such as the patient-centered medical home (PCMH) model (also known as the primary care medical home model); integrating such a model into cancer care can enable more efficient utilization of available health care resources. PCMHs are designed to improve health care quality, to coordinate patient care across multiple providers, and to reduce the need for inappropriate or avoidable utilization of hospital EDs and inpatient beds [8]. In North Carolina, Community Care of North Carolina (CCNC) was established to provide care coordination, management, and prevention services for Medicaid beneficiaries [9]. Medicaid beneficiaries voluntarily enroll in the CCNC program and select a PCMH provider from the list of primary care providers serving their area; this provider is expected to coordinate the overall health care needs of that beneficiary, with support from a medical management team [10]. The CCNC program focuses on patients who are moving across care settings and provides transitional support services such as face-to-face counseling during inpatient admission, medication reconciliation after discharge, self-management training, and other types of patient education. The ultimate goal of this PCMH program is to promote better health outcomes in primary care settings, thereby reducing the utilization of inpatient and ED services.

To our knowledge, there are no studies assessing the impact of CCNC on health care utilization associated with cancer treatment–related AEs. Specifically, we lack data on the role that CCNC plays in the care of breast cancer patients experiencing chemotherapy-related AEs. Hence, the specific aims of this study were to describe the extent of health care utilization associated with chemotherapy-related AEs in the Medicaid breast cancer population in North Carolina and to assess the relationship between CCNC enrollment and utilization of such health care.

Methods
Data Sources
The North Carolina Integrated Cancer Information and Surveillance System (ICISS) provided access to cancer registry data and Medicaid claims data, diagnostic and procedural code lookup libraries, and analytic support for this study. [Editor’s note: For more information about ICISS, refer to the commentary by Meyer and colleagues on pages 265–269]. Specifically, North Carolina Central Cancer Registry (NCCCR) data were used to identify women diagnosed with early-stage breast cancer from January 1, 2003, through December 31, 2007. These data include clinical information relevant to the diagnosed cancer, such as primary site and tumor staging. The NCCCR data were then linked to Medicaid claims to identify women who were insured by Medicaid during this period. The Medicaid claims data provide details about health care utilization, including service dates, which allowed us to perform a longitudinal assessment of postdiagnosis treatments and health outcomes.

Cohort Selection
Our study sample consisted of women aged 18–64 years who had a diagnosis of early-stage breast cancer (stage 0, I, or II) based on the presence in the NCCCR data of an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code of 174.X, 238.3, or 239.3. Eligible women had to have been enrolled in Medicaid for at least 1 month before their index diagnosis, and they must not have had evidence of cancer prior to that time. We excluded cases diagnosed at autopsy or on a death certificate, cases with missing or unknown tumor stage data, and cases with an additional cancer diagnosis reported in the registry within 1 year of the index diagnosis. Because Medicaid enrollment can be transient [11], we only included patients who had at least 12 months of Medicaid enrollment during the 15-month postdiagnosis period, so that most of their service claims could be captured. These exclusions followed a protocol that we described previously in another paper [10]. We further excluded patients who were eligible to receive both Medicare and Medicaid benefits, because complete treatment claims for these patients were not available. See Figure 1 (online version only) for the numbers of patients involved at each stage of the selection process.

Definition and Measurement of Variables
Cancer treatments. Breast cancer treatments (surgery, chemotherapy, radiation therapy, and hormone therapy) were identified using the codes listed in Appendix 1 (online version only). Binary variables—which were not mutually exclusive—were created to define receipt of each of these types of treatments at any time during the study period.

Health care visits associated with chemotherapy-related AEs (dependent variable). Because chemotherapy is the principal source of AEs in the breast cancer patient population, we specifically focused on these AEs. For the purpose of this study, we included moderate-to-severe chemotherapy-related AEs experienced by breast cancer patients as reported in other studies [7, 12]. Following the approach of Hassett and colleagues [7], we grouped these AEs into 8 categories: abnormal electrolyte levels or dehydration; constitutional symptoms and nonspecific symptoms associated with chemotherapy; nausea, emesis, and diarrhea; infection and fever; malnutrition; anemia and transfusion of red blood cells; neutropenia or thrombocytopenia; and deep vein thrombosis or pulmonary embolism. The codes for these AEs are listed in Appendix 1 (online version only).

It is important to note that these AEs are not 100% specific to the receipt of chemotherapy and may occur in breast cancer patients who did not receive chemotherapy. In order to be consistent with and allow comparisons to the published literature, we used the algorithm developed by Hassett and colleagues [7] to define this type of health care utilization, and we use the term ”chemotherapy-related AEs” in our study; however, we also attempted to parse out the occurrence of these events in patients who did not receive chemotherapy. Patients diagnosed with stage 0 breast cancer were included in our analysis because we observed that a significant proportion (nearly 22%) of these patients received chemotherapy during the study period; thus, they were likely to have experienced clinically relevant AEs.

A patient was considered to have experienced a chemotherapy-related AE when an inpatient, outpatient, or ED claim indicated that any of the above-mentioned AEs had been reported using a relevant procedure code, ICD 9-CM code, or diagnosis-related group (DRG) code. Because the first-listed diagnosis code on the claim does not necessarily represent the reason for the visit [13], mostly because of intercoder variations in practice, all available diagnosis code fields (up to 9) were used. Health care visits during which a patient received cancer treatments such as chemotherapy or radiation therapy were excluded. The primary outcome variables were the total number of visits to each of the 3 health care settings.

CCNC enrollment (primary independent variable). Because CCNC networks and primary care providers receive a per-member-per-month (PMPM) payment from Medicaid to coordinate the health and disease management needs of the enrolled population, patients were considered to be enrolled in the CCNC program (and thus to have a PCMH) when both the network and the provider management fees were paid on a monthly basis. We identified the PMPM payment using state-defined procedure codes (W9920 or W9921 for the provider and W9923 for the network) [14]. A binary variable representing patients who had any CCNC enrollment during the 15-month postdiagnosis period served as the primary independent variable.

Control variables. The sociodemographic control variables were sex, race (non-Hispanic white, non-Hispanic black, or other), rural/urban status of the county of residence, and whether or not Medicaid eligibility was due to blindness or disability.

Clinical control variables included cancer stage and comorbidity index, the latter of which was determined using a previously described [10, 15] modification of the National Cancer Institute Combined Index algorithm. Cancer stage was derived using the American Joint Committee on Cancer (AJCC) stage grouping, if possible; the Surveillance, Epidemiology, and End Results (SEER) summary staging was used if AJCC stage grouping was not available, or tumor, node, and metastasis (TNM) staging was used if neither AJCC stage grouping nor SEER summary staging was available [16].

Statistical Analysis
Bivariate analyses were performed to compare the characteristics and mean numbers of visits for chemotherapy-related AEs, by type of health care setting, for patients who had ever been enrolled in CCNC versus those who had never been enrolled in CCNC. Because of the lack of specificity of chemotherapy-related AEs, we also used bivariate statistics to compare the frequency with which patients experienced each type of AE and their chemotherapy status (ie, whether they received chemotherapy during the study period or not). Chi-square and Fisher’s exact tests were used for comparing categorical variables, and t tests were used for comparing continuous variables.

We used a method proposed by Long and Freese [17] to select an appropriate analytic model with which to perform multivariate analyses for our continuous outcome measures (see Figure 2; online version only). We chose negative binomial regression for our analytic model, because it fit the observed data best, and it took into account the possibility of AEs occurring in patients who were not receiving chemotherapy [18].

Three separate regression analyses were performed to estimate health care utilization; outcomes were the number of inpatient visits associated with chemotherapy-related AEs, the number of outpatient visits associated with chemotherapy-related AEs, and the number of ED visits associated with chemotherapy-related AEs. We used an alpha level of 0.05 to determine statistical significance. Results are presented as average marginal effects (MEs), which represent the marginal change in the number of visits induced by changes in each independent variable [19]. For instance, for our primary independent variable (CCNC enrollment), the ME would indicate the increase or decrease in the number of inpatient admissions associated with chemotherapy-related AEs for patients who were enrolled in CCNC compared with those who were not enrolled in CCNC. Modeling was performed using Stata statistical software [20].

Results
A total of 570 breast cancer patients met the inclusion criteria and were included in the analysis. The average age of patients in the sample was 48.4 years (standard deviation [SD] = 8.9); the sample included an almost equal proportion of white patients (43.0%) and black patients (44.4%); and the majority of patients (65.3%) lived in an urban area. Fifty-four percent of patients in the sample were enrolled in CCNC for at least 1 month during the study period; among patients with any enrollment, the mean duration of enrollment was 10.8 months. Table 1 summarizes the demographic and clinical characteristics of patients by CCNC enrollment status. Enrollment in CCNC was more common among younger women (P<.001) and among those who were black (P=.005).

Nearly three-fourths (n = 412 [72.3%]) of the patients in the sample had at least 1 health care visit associated with a chemotherapy-related AE during the 15-month follow-up period; specifically, 19.1% had at least 1 inpatient admission, 69.6% had at least 1 outpatient visit, and 24.9% had at least 1 ED visit. The mean numbers of visits associated with chemotherapy-related AEs during the follow-up period were 0.30 for inpatient admissions (SD = 0.83; range, 0–8), 6.92 for outpatient visits (SD = 10.94; range, 0–75), and 0.46 for ED visits (SD = 1.26; range, 0–15). Patients enrolled in CCNC had fewer overall visits associated with chemotherapy-related AEs across all 3 health care settings, but the differences were not statistically significant in the bivariate analysis (see Figure 3). Table 2 shows the numbers and percentages of patients who experienced each type of chemotherapy-related AE during the study period, by chemotherapy status (received chemotherapy or did not receive chemotherapy).

Table 3 shows the average ME of CCNC enrollment and other covariates on health care utilization (inpatient, outpatient, and ED visits) associated with chemotherapy-related AEs. CCNC enrollment was associated with significantly fewer inpatient admissions for chemotherapy-related AEs (ME, –0.14; 95% confidence interval, –0.280 to –0.004). Our analysis did not find any significant association between CCNC enrollment and the number of outpatient or ED visits for chemotherapy-related AEs. Age at diagnosis and black race were significantly associated with fewer outpatient visits for chemotherapy-related AEs. Higher comorbidity score, receipt of chemotherapy, and receipt of surgery were significantly associated with a greater number of health care visits for chemotherapy-related AEs across all 3 health care settings.

Discussion
To our knowledge, this is the first study to investigate the role of PCMHs in managing chemotherapy-related AEs among women who were recently diagnosed with breast cancer. In this study, we describe the extent of health care utilization associated with chemotherapy-related AEs among early-stage breast cancer patients insured by Medicaid. We found that these patients commonly experienced many of the AEs typically associated with chemotherapy—including abnormal electrolyte levels, dehydration, nausea, emesis, and diarrhea—all of which can be proactively monitored and managed during and after chemotherapy in coordinated care settings such as a PCMH. CCNC’s patient-centered approach to providing transitional support and facilitating follow-up care in primary care settings gives it an advantage over other care delivery systems.

The proportions of patients seeking health care for chemotherapy-related AEs were higher in our study than in the study by Hassett and colleagues (19% versus 12% for hospitalizations; 25% versus 6% for ED visits) [7]. This could be because of differences between the 2 studies in population characteristics and/or duration of follow-up (15 months versus 12 months). Also, the Medicaid patient population has higher rates of hospitalization for conditions that can often be treated outside of the hospital or avoided altogether [21, 22], compared with Medicare populations that have been studied previously.

In our study, North Carolina breast cancer patients who were enrolled in CCNC had fewer overall health care visits associated with chemotherapy-related AEs than did patients who were not enrolled in CCNC; this finding was clinically and statistically significant for inpatient admissions but not for outpatient visits or ED visits. Because inpatient admissions are costly to the North Carolina health care system and may represent more severe AEs, our findings suggest an important potential value of CCNC’s efforts to improve coordination of care and to ensure timely access to primary care providers. This finding is consistent with the general expectation that medical care is more often provided in the outpatient setting in a PCMH model, thus increasing outpatient utilization and decreasing inpatient admissions and ED use.

Interestingly, we observed a negative association between age at diagnosis and the number of AEs (ie, there were fewer AEs with older age). One possible explanation for this finding could be that younger breast cancer patients tend to present with more advanced tumors that require more aggressive treatment [23-25]; hence, these patients may be more likely to experience a chemotherapy-related AE that leads to 1 or more health care visits.

Like previous studies [6, 7], our study found a significantly positive association between the overall number of health care visits associated with AEs and several other factors: cancer stage, the presence of comorbidities, and receipt of chemotherapy. Our finding that a sizeable proportion of stage 0 patients received chemotherapy may indicate that these patients were initially diagnosed at stage 0 but later progressed to a more advanced stage, at which time they were deemed appropriate candidates for chemotherapy. Unfortunately, we cannot verify this hypothesis, because NCCCR data only record cancer stage at incident diagnosis. Other possibilities are that these stage 0 patients may have been found to have HER2-positive tumors, or they may have had high Oncotype Dx scores.

Our study has several limitations. First, Medicaid claims data include up to 9 diagnosis codes for each health care encounter, and some of the visits that we identified as being associated with a chemotherapy-related AE may have been only peripherally related to the AE; this could have led to a possible overestimation of health care utilization associated with chemotherapy-related AEs. In the absence of patient- or provider-reported data about the visit, we have no way of knowing whether this is the case. Second, in the absence of patients’ medical records, we did not have detailed contextual information about communication with and recommendations of providers, or about any sentinel events leading up to and possibly influencing the visit. Third, because of unavailability of data, we were unable to include patients with stage III breast cancer, who are also suitable candidates for chemotherapy and may therefore experience chemotherapy-related AEs. Finally, because this was an observational study, we cannot draw any conclusions about the causality of these associations.

Despite these limitations, our study provides important and timely information about the extent of health care utilization associated with chemotherapy-related AEs, and it suggests that care management programs such as CCNC may reduce the number of health care visits associated with chemotherapy-related AEs. In the context of an aging population and rising cancer-related health care costs, these findings have important implications for estimating health care demand and for containing health care costs in North Carolina [26].

Providing high-quality cancer care for economically disadvantaged populations is challenging for several reasons, including gaps in communication, poor access to primary care, and lack of post-treatment coordination of care [27]. In addition, disadvantaged patients may be less able to self-manage their condition and to meet their post-treatment health care needs [28]. The importance of innovative interventions that can bridge these gaps, facilitate optimum care of breast cancer patients, and maintain quality of care cannot be overemphasized under current economic conditions, given the increasing demand for care and shortages of providers. PCMHs are equipped to improve communication and to provide care management and coordination of care [29, 30]; thus, they can help providers to meet these challenges more effectively when caring for breast cancer survivors.

PCMHs have proven to be effective in improving outcomes for patients with chronic diseases such as asthma, diabetes, and cardiovascular conditions [14, 31-33]. As we have reported elsewhere [10], breast cancer survivors who were enrolled in CCNC were significantly more likely to receive guideline-concordant follow-up care. By virtue of their patient-centered primary care approach, CCNC and other PCMH programs hold huge potential for reducing costs (by proactively managing treatment-related AEs in a primary care setting) and for reducing preventable inpatient admissions and ED visits. The results of this study, although based on North Carolina Medicaid data, provide relevant information to providers and policymakers and should encourage PCMH initiatives nationwide. In this study we showed that any CCNC enrollment during the 15-month postdiagnosis period may decrease health care utilization associated with chemotherapy-related AEs, but future studies could investigate how the duration and intensity of CCNC participation affect health care utilization among cancer patients. Future research also should continue to explore the extent to which PCMHs can help to coordinate survivorship care and to monitor and manage late effects of cancer treatments.

Acknowledgments
The authors would like to thank and recognize their expert advisory committee, consisting of Anne Braswell, Jonathan Fischer, Bo Gamble, Eleanor Greene, Linda Kinney, Patrick Maguire, Brenda McCants, Rachel Raab, LaSonia Roberts-Melvin, and Thea Monet. The authors also would like to thank Annette DuBard and Hy Muss for their insights into Community Care of North Carolina and cancer survivorship care.

Financial support. This work was supported by a Health-e-NC pilot grant from the University Cancer Research Fund. S.B.W. was supported by an Agency for Healthcare Research and Quality Comparative Effectiveness Research Career Development Award, 1-K-12 HS019468-01 (principal investigator, Weinberger; scholar: Wheeler). In addition, this study was supported by the Integrated Cancer Information and Surveillance System (ICISS), a UNC Lineberger Comprehensive Cancer Center resource funded by the state of North Carolina through the University Cancer Research Fund.

Potential conflicts of interest. All authors have no relevant conflicts of interest.

References
1. North Carolina Central Cancer Registry. Statistics and reports. Cancer. Annual reports. NC cancer incidence rates 1999–2011. NC cancer mortality rates 1999–2011. North Carolina State Center for Health Statistics Web site. http://www.schs.state.nc.us/data/cancer.cfm. Accessed May 14, 2014.

2. Hewitt M, Greenfield S, Stovall E, eds; Institute of Medicine; National Research Council. From Cancer Patient to Cancer Survivor: Lost in Transition. Washington, DC: National Academies Press; 2006.

3. Bates DW, Spell N, Cullen DJ, et al. The costs of adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study Group. JAMA. 1997;277(4):307-311.

4. Raschetti R, Morgutti M, Menniti-Ippolito F, et al. Suspected adverse drug events requiring emergency department visits or hospital admissions. Eur J Clin Pharmacol. 1999;54(12):959-963.

5. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1997;277(4):301-306.

6. Du XL, Osborne C, Goodwin JS. Population-based assessment of hospitalizations for toxicity from chemotherapy in older women with breast cancer. J Clin Oncol. 2002;20(24):4636-4642.

7. Hassett MJ, O’Malley AJ, Pakes JR, Newhouse JP, Earle CC. Frequency and cost of chemotherapy-related serious adverse effects in a population sample of women with breast cancer. J Natl Cancer Inst. 2006;98(16):1108-1117.

8. Rosenthal TC. The medical home: growing evidence to support a new approach to primary care. J Am Board Fam Med. 2008;21(5):427-440.

9. Wilhide S, Henderson T. Community Care of North Carolina: a provider-led strategy for delivering cost-effective primary care to Medicaid beneficiaries. Washington, DC: American Academy of Family Physicians; 2006.

10. Wheeler SB, Kohler RE, Goyal RK, et al. Is medical home enrollment associated with receipt of guideline-concordant follow-up care among low-income breast cancer survivors? Med Care. 2013;51(6):494-502.

11. Seifert R, Kirk G, Oakes M. Enrollment and Disenrollment in MassHealth and Commonwealth Care. Boston, MA: Massachusetts Medicaid Policy Institute; 2010. http://chapter58.bluecrossfoundation.org/sites/default/files/download/publication/2010_4_21_disenrollment_mh_cc_0.pdf. Accessed May 14, 2014.

12. Shapiro CL, Recht A. Side effects of adjuvant treatment of breast cancer. N Engl J Med. 2001;344(26):1997-2008.

13. Senathirajah M, Owens P, Mutter R, Nagamine M. Special Study on the Meaning of the First-Listed Diagnosis on Emergency Department and Ambulatory Surgery Records. Healthcare Cost and Utilization Project (HCUP) Methods Series. Report #2011-03. Washington, DC: US Agency for Healthcare Research and Quality; 2011. http://www.hcup-us.ahrq.gov/reports/methods/2011_03.pdf. Accessed July 31, 2013.

14. Domino ME, Humble C, Lawrence WW Jr, Wegner S. Enhancing the medical homes model for children with asthma. Med Care. 2009;47(11):1113-1120.

15. Klabunde CN, Lanier D, Nadel MR, McLeod C, Yuan G, Vernon SW. Colorectal cancer screening by primary care physicians: recommendations and practices, 2006–2007. Am J Prev Med. 2009;37(1):8-16.

16. Wheeler SB, Wu Y, Meyer A-M, et al. Use and timeliness of radiation therapy after breast-conserving surgery in low-income women with early-stage breast cancer. Cancer Invest. 2012;30(4):258-267.

17. Long JS, Freese J. Predicted probabilities for count models. Stata J. 2001;1(1):51-57. http://www.stata-journal.com/sjpdf.html?articlenum=st0002. Accessed May 14, 2014.

18. Greene WH. Accounting for excess zeros and sample selection in Poisson and negative binomial regression models. NYU Working Paper No. EC-94-10. New York, NY: New York University Stern School of Business; 1994. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1293115. Accessed July 31, 2013.

19. Williams R. Using the margins command to estimate and interpret adjusted predictions and marginal effects. Paper presented at: Stata Conference; July 14–15, 2011; Chicago, IL. http://ideas.repec.org/p/boc/chic11/9.html. Accessed May 15, 2014.

20. Stata [computer program]. Version 21.1. College Station, TX: StataCorp; 2012.

21. Weissman JS, Gatsonis C, Epstein AM. Rates of avoidable hospitalization by insurance status in Massachusetts and Maryland. JAMA. 1992;268(17):2388-2394.

22. Falik M, Needleman J, Wells BL, Korb J. Ambulatory care sensitive hospitalizations and emergency visits: experiences of Medicaid patients using federally qualified health centers. Med Care. 2001;39(6):551-561.

23. Halpern MT, Bian J, Ward EM, Schrag NM, Chen AY. Insurance status and stage of cancer at diagnosis among women with breast cancer. Cancer. 2007;110(2):403-411.

24. Gajdos C, Tartter PI, Bleiweiss IJ, Bodian C, Brower ST. Stage 0 to stage III breast cancer in young women. J Am Coll Surg. 2000;190(5):523-529.

25. Partridge AH, Ruddy KJ, Kennedy J, Winer EP. Model program to improve care for a unique cancer population: young women with breast cancer. J Oncol Pract. 2012;8(5):e105-e110.

26. Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future of cancer incidence in the United States: burdens upon an aging, changing nation. J Clin Oncol. 2009;27(17):2758-2765.

27. Sprandio JD. Oncology patient-centered medical home and accountable cancer care. Community Oncol. 2010;7(12):565-572.

28. Chen JY, Diamant AL, Thind A, Maly RC. Determinants of breast cancer knowledge among newly diagnosed, low-income, medically underserved women with breast cancer. Cancer. 2008;112(5):1153-1161.

29. Berenson RA, Hammons T, Gans DN, et al. A house is not a home: keeping patients at the center of practice redesign. Health Aff (Millwood). 2008;27(5):1219-1230.

30. Friedberg MW, Coltin KL, Safran DG, Dresser M, Schneider EC. Medical home capabilities of primary care practices that serve sociodemographically vulnerable neighborhoods. Arch Intern Med. 2010;170(11):938-944.

31. Lewis J, Hoyt A, Kakoza RM. Enhancing quality of primary care using an ambulatory ICU to achieve a patient-centered medical home. J Prim Care Community Health. 2011;2(4):234-239.

32. Maeng DD, Graf TR, Davis DE, Tomcavage J, Bloom FJ Jr. Can a patient-centered medical home lead to better patient outcomes? The quality implications of Geisinger’s ProvenHealth Navigator. Am J Med Qual. 2012;27(3):210-216.

33. Auger KA, Kahn RS, Davis MM, Beck AF, Simmons JM. Medical home quality and readmission risk for children hospitalized with asthma exacerbations. Pediatrics. 2013;131(1):64-70.


Ravi K. Goyal, MS, BPharm health outcomes scientist, RTI Health Solutions, Research Triangle Park, North Carolina; research associate, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Stephanie B. Wheeler, PhD, MPH faculty member, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill; assistant professor, Department of Health Policy and Management, UNC Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Racquel E. Kohler, MSPH doctoral student and predoctoral fellow, Department of Health Policy and Management, UNC Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Kristen H. Lich, PhD, MHSA assistant professor, Department of Health Policy and Management, UNC Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Ching-Ching Lin, MHS doctoral student and predoctoral fellow, Department of Health Policy and Management, UNC Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Katherine Reeder-Hayes, MD, MBA assistant professor, Division of Hematology and Oncology, School of Medicine, University of North Carolina at Chapel Hill; faculty member, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Anne-Marie Meyer, PhD facility director, Integrated Cancer Information and Surveillance System (ICISS), Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Deborah K. Mayer, PhD, RN, AOCN, FAAN director of cancer survivorship, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill; associate professor, UNC School of Nursing, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.

Address correspondence to Dr. Stephanie B. Wheeler, 135 Dauer Dr, 1103C McGavran-Greenberg Hall, CB# 7411, Chapel Hill, NC 27599 (stephanie_wheeler@unc.edu).