Providing patient-centered comprehensive care to long-term cancer survivors may lead to reduced total healthcare expenditures. Check out our website’s new table/figure pop-up feature! Click on the name of a table or figure in the text to see it in your browser.
Objectives: To evaluate the economic outcomes associated with patient perceptions of patient-centered medical home (PCMH) characteristics among long-term cancer survivors in the United States.
Study Design: A retrospective analysis of the 2008 to 2012 Medical Expenditure Panel Survey.
Methods: A nationally representative sample of adult long-term cancer survivors (≥3 years since diagnosis) was categorized into either patient-centered care (PCC) or non-PCC groups based on responses to PCMH model hallmark attributes of “comprehensive care,” “whole-person orientation,” and “accessible care.” The positive perception of all 3 attributes was defined as PCC. The patient perceptions, as well as patient characteristics, were measured at year 1 (baseline), with a propensity score model to balance baseline characteristics. Adjusted total healthcare utilization and healthcare expenditures in 2014 US$ at year 2 (follow-up) were compared between the PCC and non-PCC groups.
Results: A total of 4288 long-term cancer survivors were identified, with a mean (SD) age of 65.2 (13.8) years. The PCC group was associated with a reduction in mean adjusted healthcare expenditures at follow-up (savings of $1596 per cancer survivor; P = .020). These findings are driven by lower odds of hospitalization (odds ratio, 0.81; 95% CI, 0.66-0.99; P = .035) and lower hospitalization-related healthcare expenditures (PCC: $3323; 95% CI, $2727-$3918; non-PCC: $4912; 95% CI, $4039-$5785; P = .002) associated with PCC among the population who were 65 years and older. The whole-person orientation attribute had a major impact on reduced healthcare expenditures.
Conclusions: The positive patient perception of PCMH characteristics was associated with reduced healthcare expenditures in adult long-term cancer survivors.
Am J Manag Care. 2018;24(9):e292-e299Takeaway Points
Cancer is the second-leading cause of mortality in the United States as of 2014, superseded only by heart disease, according to the CDC.1 Despite this, cancer survivorship rates have increased in recent years due to advances in early detection and treatment.2 A “cancer survivor,” as defined by the National Coalition for Cancer Survivorship, is a person who has been diagnosed with cancer, from the time of initial diagnosis through the remainder of the person’s lifetime.3 As of January 2016, there are an estimated 15.5 million cancer survivors in the United States, and this number is expected to increase to 20.3 million by 2026 and to 26.1 million by 2040.2
As the number of cancer survivors in the United States continues to climb over the next decade, specialized long-term care—including, but not limited to, chemotherapy long-term and late-effects management, follow-up visits, and preventive screenings—may be necessary to ensure continuity of remission and prevention of secondary relapse, due to each cancer survivor’s unique medical and social histories. Moreover, cancer survivors are at risk for the development of comorbid conditions associated with poor survival.4 This much-needed specialized long-term care often presents an economic challenge for cancer survivors, with some opting to forgo necessary treatments due to their financial concerns.5,6 Not only does the high cost of providing specialized long-term care present an economic challenge for the cancer survivors, it can create a high economic burden for society as well. One study estimated the annual excess economic burdens of cancer survivorship among recently diagnosed and previously diagnosed cancer survivors aged 18 to 64 years to be $16,213 and $4427, respectively, with excess medical expenditures making up the largest share of the economic burden.7 Because of this, it may be necessary to further explore ways to improve the affordability of providing specialized long-term care for cancer survivors, as the current fee-for-service payment system still being used in many oncology practices is costly and ineffective.8
One solution to the affordability issue of specialized long-term care among cancer survivors may be through providing patient-centered comprehensive care. Based on a patient-centered medical home (PCMH) model, this care is guided by a set of joint principles, which include a personal physician, a physician-directed medical practice, whole-person orientation, coordinated and/or integrated care, quality and safety, enhanced access to care, and payment.9 The PCMH model encourages patient-centered healthcare services among providers with a triple aim of improved health, improved quality, and controlled cost.10 Although results from previously published studies have demonstrated the benefits of utilizing a PCMH model for the treatment of other chronic conditions, such as diabetes, asthma, and hyperlipidemia,11-15 or in active cancer patients,16-19 the economic benefits of its use among long-term cancer survivors have yet to be determined.
This study aimed to investigate the association between patient perceptions of PCMH characteristics of the healthcare they receive and economic outcomes. To guide the future design of a PCMH model among cancer survivors, it is important to understand how each PCMH characteristic of their healthcare is associated with overall healthcare costs. In this study, we evaluated total medical expenditures and healthcare utilization among long-term cancer survivors based on patient perceptions of healthcare in a nationally representative United States sample.
Study Design and Data Source
This study was a retrospective analysis of survey data acquired from the 2008 to 2012 Medical Expenditure Panel Survey (MEPS) Household Component longitudinal files (Panels 13-16). MEPS, published by the Agency for Healthcare Research and Quality, is a set of large-scale nationally representative surveys of families and individuals, in addition to medical providers and employers, across the United States.20 The goal of MEPS is to gather data on the usage of medical care and expenditures. Data are collected over 2 years in 5 rounds of in-person interviews.
The study population included adults 18 years and older who self-reported that they had ever been diagnosed with cancer, as indicated by the “CANCERY1” variable in the MEPS Household Component. Survey respondents were also asked about their cancer type and the age at which they received their cancer diagnosis. The type of diagnosed cancer was also identified using unique clinical classification coding variables from these self-reported data. Cancer survivors were excluded from this study if they did not have a usual care provider. Furthermore, because the study’s population of interest was long-term cancer survivors, cancer survivors were also excluded from this study if the survey was conducted within 3 years of their last cancer diagnosis.21
Patient Perceptions of PCMH Characteristics
Self-reported patient data were used to determine whether or not cancer survivors perceived their medical care as patient-centered comprehensive care. The definition of patient-centered comprehensive care was adopted from previously published studies by using a set of questions about a patient’s interactions with their usual source of care (USC), which focused on whether the patient received “comprehensive care,” “accessible care,” and care with a “whole-person orientation,” all 3 of which are hallmark attributes of a PCMH model.22-26 Cancer survivors who responded positively for all 3 hallmark attributes of a PCMH model at baseline (year 1) were categorized as the patient-centered care (PCC) group. Cancer survivors who did not respond positively to any of the PCMH characteristics or responded positively to only 1 or 2 characteristics were subsequently categorized as the non-PCC group. The list of questions is in the eAppendix (available at ajmc.com).
Total healthcare utilization and total healthcare expenditures were measured in the PCC group and the non-PCC group at baseline and at follow-up (year 2). Total healthcare utilization included hospitalizations, emergency department (ED) visits, office visits, primary care provider (PCP; specialties of family medicine and internal medicine) visits, and oncologist visits. Total healthcare expenditures included spending on hospitalizations, office visits, ED visits, and prescriptions in 2014 US$.
Descriptive statistics were used to summarize baseline characteristics in the PCC and non-PCC groups, with the baseline characteristics including age, age group, years since cancer diagnosis, gender, race/ethnicity, baseline survey year, region, marital status, education level, insurance type, perceived health, cancer type, comorbidities, and Elixhauser comorbidity score.27 Differences between the PCC group and the non-PCC group were compared using Student’s t test for continuous variables and χ2 test for categorical variables. Moreover, an inverse probability propensity score (PS) method was used to balance these baseline characteristics between the PCC group and the non-PCC group.28 Baseline characteristics with P <.25 (region, marital status, education level, insurance type, perceived health, comorbidity of hypertension, rheumatoid arthritis, depression, chronic pulmonary disease, and unknown skin cancer), as well as age, sex, baseline healthcare expenditures, and ED visits, were added to predict the PS. Generalized linear models with log-rank and gamma distribution were used to test the statistical significance of the total healthcare costs between the 2 groups. Recycled prediction methods were used to estimate the mean expenditures of each group after applying the PS model. Logistic regressions, with a response of 0 indicating no and 1 indicating yes, were used in conjunction with negative binomial models in the determination of the total healthcare utilization analysis. A priori subgroup analyses were conducted to examine whether the economic effects of PCC were consistent across the 2 subgroups of survivors either younger than 65 years or 65 years and older. Additionally, a secondary analysis was performed to investigate the association between each of the 3 PCMH model hallmark attributes and its potential impact on economic outcomes. A P value of <.05 was accepted for statistical significance. All statistical analyses were performed using SAS, version 9.4 (SAS Institute; Cary, North Carolina), and STATA, version 12 (STATA; College Station, Texas).
A total of 4288 adult patients were identified as long-term cancer survivors. At baseline, 1883 (43.9%) were categorized into the PCC group, while the remaining 2405 cancer survivors (56.1%) were categorized into the non-PCC group (Table 1 [part A and part B]). For each of the characteristics associated with the 3 hallmark attributes of a PCC (comprehensive care, whole-person orientation, and accessible care), a low of 79% to a high of 99% positive response rate was observed, depending on the characteristic (eAppendix). These positive responses allowed for the differentiation of the PCC group from the non-PCC group.
Overall, the mean (SD) age of a long-term cancer survivor was 65.2 (13.8) years, and the mean (SD) years since cancer diagnosis were 10.1 (10.0). Approximately 75.6% identified themselves as white and 57.4% were female. In general, cancer survivors in the PCC group were more likely to have private insurance, perceive their health as excellent, and have comorbid hypertension, and less likely to have comorbidities of depression and chronic pulmonary disease, compared with cancer survivors in the non-PCC group.
After applying the PS, the weighted sample showed good balance in baseline characteristics between the PCC and non-PCC groups with small standardized differences. Differences in healthcare utilization between the PCC and non-PCC groups were not statistically significant after applying the PS model (Table 2). Despite this finding, the PCC group, compared with the non-PCC group, had a trend of lower odds of hospitalization and ED visits, as well as decreases in utilization of hospitalizations, ED visits, office visits, and PCP visits.
The overall crude mean total healthcare expenditures were similar at baseline between the PCC group ($11,193; 95% CI, $10,122-$12,265) and the non-PCC group ($11,783; 95% CI, $11,055-$12,512; P = .357) (Table 3). At follow-up, the PCC group had significantly lower crude mean total healthcare expenditures ($11,208; 95% CI, $10,204-$12,212) than the non-PCC group ($13,316; 95% CI, $12,463-$14,169; P = .002). Considering baseline differences between the 2 groups using the PS model, the PCC group had significantly lower adjusted mean total healthcare expenditures ($11,433; 95% CI, $10,430-$12,437) than the non-PCC group ($13,020; 95% CI, $12,166-$13,873), yielding cost savings of $1587 per cancer survivor (P = .020) (Figure).
Subgroup and Secondary Analyses
Subgroup analyses showed that cost savings were from those 65 years and older. Among cancer survivors younger than 65 years, there were no statistical differences in total healthcare expenditures, but a significant decrease in prescription expenditures in the PCC group ($2580; 95% CI, $2152-$3008) compared with the non-PCC group ($3309; 95% CI, $2858-$3760) yielded a cost savings of $729 per cancer survivor in that subgroup (P = .023) (Table 4).
In the 65 years and older subgroup, the PCC group was associated with lower odds of hospitalization compared with the non-PCC group (odds ratio [OR], 0.81; 95% CI, 0.66-0.99; P = .035). Moreover, there were statistically significant differences in adjusted total healthcare expenditures between the PCC group ($11,918; 95% CI, $10,881-$12,954) and the non-PCC group ($14,382; 95% CI, $13,187-$15,576; P = .002). Hospitalization-related healthcare expenditures were also lower in the PCC group ($3323; 95% CI, $2727-$3918) than in the non-PCC group ($4912; 95% CI, $4039-$5785) in this subgroup analysis (P = .002) (Table 4). Other types of costs were not statistically different between the PCC and non-PCC groups.
The secondary analysis indicated that of the 3 hallmark attributes of the PCMH model (ie, comprehensive care, whole-person orientation, accessible care), the whole-person orientation attribute was mainly responsible for the reductions observed in the PCC group’s healthcare utilization and healthcare expenditures (P = .006). On the contrary, the comprehensive care attribute (P = .906) and the accessible care attribute (P = .905) did not show any significant impact. From these findings, the whole-person orientation attribute was further analyzed to determine which specific healthcare utilization and/or expenditure types were affected by it. It was revealed that the PCC group had lower adjusted mean total healthcare expenditures ($11,666; 95% CI, $10,891-$12,440) than the non-PCC group ($13,572; 95% CI, $12,405-$14,739), yielding cost savings of $1906 per cancer survivor in this subgroup (P = .006). Furthermore, the whole-person orientation attribute was associated with decreased odds of ED visits (OR, 0.79; 95% CI, 0.68-0.92; P = .003) and decreased healthcare expenditures associated with ED visits (cost savings of $96 per cancer survivor; P = .024) (Table 4).
The long-term cancer survivor population is more likely to utilize more healthcare resources and accumulates more total medical expenditures over time than adults without a history of cancer, due to their complex medical history and continuous need for rigorous preventive screening, follow-up visits, chemotherapy long-term and late-effect management, and specialized care to prevent secondary relapse of their cancer.2,29 In addition, a higher risk for developing comorbid conditions and higher treatment costs further burden the patients. There is an urgent need for a more cost-effective method of managing this patient population.8,30-35 A patient-centered comprehensive care model has the potential to provide more benefits to cancer survivors than to other patient cohorts, and this study tried to establish an association between patient perceptions of PCMH characteristics and total medical expenditures to further guide the future design of a PCMH model.
The findings of this study suggest that providing patient-centered comprehensive care may mitigate total healthcare utilization and subsequently reduce total healthcare expenditures among long-term cancer survivors in the United States. In the overall analysis, although there were no significant differences in healthcare utilization, there were significant reductions in adjusted mean healthcare expenditures at follow-up between the PCC group and the non-PCC group. Subgroup analyses suggest that the majority of the benefits of PCC were seen in the elderly population. In the subgroup of those 65 years and older, cancer survivors in the PCC group had 19% lower odds of hospitalization compared with the non-PCC group, and the savings were mostly from hospitalization-related healthcare expenditures. In the subgroup of those younger than 65 years, the benefits stemmed from reductions in prescription-related healthcare expenditures. These findings suggest that providing patient-centered comprehensive care might be more effective in the elderly population of cancer survivors than in the younger population. Elderly patients usually have more comorbidities and are likely to experience more hospitalizations or ED visits; therefore, patient-centered comprehensive care may play a bigger role in reducing overall healthcare costs. In this study, the population 65 years and older had a higher number of hospitalizations and ED visits compared with the population younger than 65 years, which might have driven our findings. These findings are also supported by previous literature that has suggested that reduction of hospital readmission rates and lowering of inpatient expenditures were more pertinent to the older adult population.26,36 However, this study was not able to further explain the reasoning behind these findings.
The secondary analysis performed in this study revealed that of the 3 hallmark attributes of a PCMH, the whole-person orientation attribute contributed the most to the reduction in total healthcare expenditures in long-term cancer survivors, whereas the comprehensive care and accessible care attributes were not associated with savings. This may be due to a ceiling effect for the comprehensive care and accessible care attributes, because their positive responses reached 98.3% and 93.7%, respectively, making it more difficult to assess if they impacted the economic outcomes.
The impact of the whole-person orientation attribute on reducing healthcare utilization and expenditures may be explained by the idea that as clinicians empower cancer survivors to become more engaged in the decision-making process of their own care, the patients become more responsible and adherent to those treatments. As a result, unnecessary healthcare expenditures stemming from treatment nonadherence are reduced. This finding was supported by further evaluating the whole-person orientation attribute in the secondary analysis, which showed significant decreases in total healthcare utilization, specifically in the odds of ED visits, in addition to decreases in ED visit—related healthcare expenditures and total healthcare expenditures.
The significant reductions in total healthcare expenditures and total healthcare utilization among the PCC group in this study, especially the reduced odds of hospitalization and reductions in hospitalization-related expenditures, were mostly consistent with findings from other similarly designed studies evaluating PCMHs. Findings of a study by Cuellar and colleagues suggested that a PCMH model, by the third year of its inception, resulted in lower rates of healthcare utilization and subsequently lower observed total healthcare expenditures.37 Likewise, a study by Cole and colleagues found diminished healthcare expenditures, as well as fewer hospitalizations, among Louisiana primary care clinics that have adopted the PCMH model in the management of their chronically ill patients.38 One study by Kohler and colleagues reached a different conclusion and found that patients with breast cancer in North Carolina actively receiving treatment in a PCMH model had higher total healthcare expenditures after initial diagnosis via increased total healthcare utilization, possibly due to greater access to their primary and specialty care teams compared with those who were not in a PCMH model.39
However, it is important to note that our study findings should be understood as the economic benefits associated with patient perceptions of their care instead of direct effects of implementation of a structured PCMH model. The definition of PCC relied on cancer survivors’ self-reported data; cancer survivors were considered to have received patient-centered comprehensive care if they provided positive responses to statements correlated with common characteristics of a PCMH model, rather than actual enrollment in a PCMH program. This patient perception may be affected by various factors, including characteristics of patients, individual physicians, or the practices in which patients received care. Although this study does not provide direct evidence from the implementation of a PCMH model, the indirect method of measurement may provide an advantage because it allows us to examine unique patient perspectives and compare each of the 3 hallmark attributes of the PCMH model and how each attribute can affect overall economic outcomes.
Limitations and Strengths
This study has several limitations. The PCMH attributes were potentially determined by both primary care and oncology care because the survey respondents did not specify whether their USC was their primary care provider or an oncologist. Therefore, study findings may not serve as evidence of patient perception of oncology care. Another limitation of this study is possible bias from unobserved confounding variables, as commonly seen with other observational studies. In this study, unobserved confounding variables, such as disease severity, may have affected the study findings rather than the PCC. However, observed confounding variables were adjusted with a PS model to maintain balance between the PCC group and the non-PCC group of cancer survivors. Other possible limitations of this study are nonresponse bias or recall bias, which are common in survey-type studies. Therefore, the analysis had to utilize only data that were readily available.
Despite these limitations, this study also has various strengths. To date, there have been a limited number of studies on the effects of implementing a PCMH model for patients with cancer and even fewer studies examining its effects on long-term cancer survivors. The findings derived from this study may be useful in the development of clinical programs for the care of long-term cancer survivors or in aiding the design of future PCMH models for managing other costly chronic conditions. In addition, this study utilized panel data to investigate the association between patient perceptions of PCMH characteristics at year 1 and economic outcomes at year 2, thereby minimizing limitations that would arise from a cross-sectional study when evaluating outcomes. In addition, a majority of cancer survivors (80%) did not change their responses to the questions on having received care attributable to PCMH characteristics at year 2, which suggests a continuation of similar types of care provided throughout the year. Furthermore, the unique patient perspectives obtained from a nationally representative sample may allow the results to be further applied and generalized to the broader US population.
The positive patient perception of PCMH characteristics was associated with reductions in mean total healthcare utilization and mean total healthcare expenditures among long-term cancer survivors. Future studies should further investigate the economic benefits of implementing a PCMH model for long-term cancer survivors through an interventional study design, in order to gain a better understanding of the origins of cost savings.Author Affiliations: Department of Pharmacy Practice and Administration, Western University of Health Sciences College of Pharmacy (JA, AL), Pomona, CA.
Source of Funding: None.
Prior Presentation: Part of the study findings were presented as a poster at the Academy of Managed Care Pharmacy Managed Care & Specialty Pharmacy Annual Meeting 2017, Denver, Colorado, March 27-31, 2017.
Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (JA); acquisition of data (JA); analysis and interpretation of data (JA, AL); drafting of the manuscript (JA, AL); critical revision of the manuscript for important intellectual content (JA, AL); statistical analysis (JA); provision of patients or study materials (JA); administrative, technical, or logistic support (JA); and supervision (JA).
Address Correspondence to: JaeJin An, BPharm, PhD, Department of Pharmacy Practice and Administration, Western University of Health Sciences College of Pharmacy, 309 E Second St, Pomona, CA 91766. Email: email@example.com.REFERENCES
1. U.S. Cancer Statistics data visualization tool. CDC website. gis.cdc.gov/Cancer/USCS/DataViz.html. Published June 2018. Accessed July 18, 2018.
2. Bluethmann SM, Mariotto AB, Rowland JH. Anticipating the “silver tsunami”: prevalence trajectories and comorbidity burden among older cancer survivors in the United States. Cancer Epidemiol Biomarkers Prev. 2016;25(7):1029-1036. doi: 10.1158/1055-9965.EPI-16-0133.
3. Defining cancer survivorship. National Coalition for Cancer Survivorship website. canceradvocacy.org/news/defining-cancer-survivorship. Published July 24, 2014. Accessed August 17, 2017.
4. Søgaard M, Thomsen RW, Bossen KS, Sørensen HT, Nørgaard M. The impact of comorbidity on cancer survival: a review. Clin Epidemiol. 2013;5(suppl 1):3-29. doi: 10.2147/CLEP.S47150.
5. Yabroff KR, Lund J, Kepka D, Mariotto A. Economic burden of cancer in the United States: estimates, projections, and future research. Cancer Epidemiol Biomarkers Prev. 2011;20(10):2006-2014. doi: 10.1158/1055-9965.EPI-11-0650.
6. Weaver KE, Rowland JH, Bellizzi KM, Aziz NM. Forgoing medical care because of cost: assessing disparities in healthcare access among cancer survivors living in the United States. Cancer. 2010;116(14):3493-3504. doi: 10.1002/cncr.25209.
7. Guy GP Jr, Ekwueme DU, Yabroff KR, et al. Economic burden of cancer survivorship among adults in the United States. J Clin Oncol. 2013;31(30):3749-3757. doi: 10.1200/JCO.2013.49.1241.
8. Patel K, Thoumi A, Nadel J, O’Shea J, McClellan M. Transforming oncology care: payment and delivery reform for person-centered care. Am J Manag Care. 2015;21(5):388-393.
9. American Academy of Family Physicians; American Academy of Pediatrics; American College of Physicians; American Osteopathic Association. Joint principles of the patient-centered medical home. American Academy of Family Physicians website. aafp.org/dam/AAFP/documents/practice_management/pcmh/initiatives/PCMHJoint.pdf. Published March 2007. Accessed August 17, 2017.
10. Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27(3):759-769. doi: 10.1377/hlthaff.27.3.759.
11. McGinley EL, Gabbay RA. The impact of new payment models on quality of diabetes care and outcomes. Curr Diab Rep. 2016;16(6):51. doi: 10.1007/s11892-016-0743-5.
12. 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. doi: 10.1097/MLR.0b013e3181adcc65.
13. Beadles CA, Hassmiller Lich K, Viera AJ, Greene SB, Brookhart MA, Weinberger M. Patient-centered medical homes and oral anticoagulation therapy initiation. Med Care Res Rev. 2014;71(2):174-191. doi: 10.1177/1077558713510563.
14. Beadles CA, Farley JF, Ellis AR, et al. Do medical homes increase medication adherence for persons with multiple chronic conditions? Med Care. 2015;53(2):168-176. doi: 10.1097/MLR.0000000000000292.
15. Buchanan JR, Liang Q, Walsh C, Manolis CH, Perkins SE. Cholesterol medication adherence improvement in patient centered medical home primary care practices. Value Health. 2013;16(3):A292. doi: 10.1016/j.jval.2013.03.1513.
16. Green BB, Anderson ML, Chubak J, et al. Colorectal cancer screening rates increased after exposure to the patient-centered medical home (PCMH). J Am Board Fam Med. 2016;29(2):191-200. doi: 10.3122/jabfm.2016.02.150290.
17. Sprandio JD. Oncology patient-centered medical home. J Oncol Pract. 2012;8(suppl 3):47s-49s. doi: 10.1200/JOP.2012.000590.
18. Page RD, Newcomer LN, Sprandio JD, McAneny BL. The patient-centered medical home in oncology: from concept to reality. Am Soc Clin Oncol Educ Book. 2015:e82-e89. doi: 10.14694/EdBook_AM.2015.35.e82.
19. Cox JV, Sprandio JD, Barkley R. Understanding and surviving the transition to value-based oncology. Am Soc Clin Oncol Educ Book. 2013:e361. doi: 10.1200/EdBook_AM.2013.33.e361.
20. Medical Expenditure Panel Survey (MEPS). Agency for Healthcare Research and Quality website. ahrq.gov/cpi/about/otherwebsites/meps.ahrq.gov/index.html. Published August 2013. Accessed August 19, 2017.
21. Frick KD, Snyder CF, Herbert RJ, et al. Relationship between quality of comorbid condition care and costs for cancer survivors. J Oncol Pract. 2016;12(6):e734-e745. doi: 10.1200/JOP.2015.006098.
22. Bowdoin JJ, Rodriguez-Monguio R, Puleo E, Keller D, Roche J. Associations between the patient-centered medical home and preventive care and healthcare quality for non-elderly adults with mental illness: a surveillance study analysis. BMC Health Serv Res. 2016;16(1):434. doi: 10.1186/s12913-016-1676-z.
23. Jones AL, Cochran SD, Leibowitz A, Wells KB, Kominski G, Mays VM. Usual primary care provider characteristics of a patient-centered medical home and mental health service use. J Gen Intern Med. 2015;30(12):1828-1836. doi: 10.1007/s11606-015-3417-0.
24. Beal A, Hernandez S, Doty M. Latino access to the patient-centered medical home. J Gen Intern Med. 2009;24(suppl 3):514-520. doi: 10.1007/s11606-009-1119-1.
25. Jerant A, Fenton JJ, Franks P. Primary care attributes and mortality: a national person-level study. Ann Fam Med. 2012;10(1):34-41. doi: 10.1370/afm.1314.
26. Stockbridge EL, Philpot LM, Pagán JA. Patient-centered medical home features and expenditures by Medicare beneficiaries. Am J Manag Care. 2014;20(5):379-385.
27. Thompson NR, Fan Y, Dalton JE, et al. A new Elixhauser-based comorbidity summary measure to predict in-hospital mortality. Med Care. 2015;53(4):374-379. doi: 10.1097/MLR.0000000000000326.
28. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. doi: 10.1080/00273171.2011.568786.
29. Short PF, Moran JR, Punekar R. Medical expenditures of adult cancer survivors aged <65 years in the United States. Cancer. 2010;117(12):2791-2800. doi: 10.1002/cncr.25835.
30. Schrag D. The price tag on progress—chemotherapy for colorectal cancer. N Engl J Med. 2004;351(4):317-319. doi: 10.1056/NEJMp048143.
31. Tangka FK, Trogdon JG, Richardson LC, Howard D, Sabatino SA, Finkelstein EA. Cancer treatment cost in the United States: has the burden shifted over time? Cancer. 2010;116(14):3477-3484. doi: 10.1002/cncr.25150.
32. Woodward RM, Brown ML, Stewart ST, Cronin KA, Cutler DM. The value of medical interventions for lung cancer in the elderly: results from SEER-CMHSF. Cancer. 2007;110(11):2511-2518. doi: 10.1002/cncr.23058.
33. Howard DH, Kauh J, Lipscomb J. The value of new chemotherapeutic agents for metastatic colorectal cancer. Arch Intern Med. 2010;170(6):537-542. doi: 10.1001/archinternmed.2010.36.
34. Wong YN, Meropol NJ, Speier W, Sargent D, Goldberg RM, Beck JR. Cost implications of new treatments for advanced colorectal cancer. Cancer. 2009;115(10):2081-2091. doi: 10.1002/cncr.24246.
35. Elkin EB, Bach PB. Cancer’s next frontier: addressing high and increasing costs. JAMA. 2010;303(11):1086-1087. doi: 10.1001/jama.2010.283.
36. Stranges PM, Marshall VD, Walker PC, Hall KE, Griffith DK, Remington T. A multidisciplinary intervention for reducing readmissions among older adults in a patient-centered medical home. Am J Manag Care. 2015;21(2):106-113.
37. Cuellar A, Helmchen LA, Gimm G, et al. The CareFirst patient-centered medical home program: cost and utilization effects in its first three years. J Gen Intern Med. 2016;31(11):1382-1388. doi: 10.1007/s11606-016-3814-z.
38. Cole ES, Campbell C, Diana ML, Webber L, Culbertson R. Patient-centered medical homes in Louisiana had minimal impact on Medicaid population’s use of acute care and costs. Health Aff (Millwood). 2015;34(1):87-94. doi: 10.1377/hlthaff.2014.0582.
39. Kohler RE, Goyal RK, Lich KH, Domino ME, Wheeler SB. Association between medical home enrollment and health care utilization and costs among breast cancer patients in a state Medicaid program. Cancer. 2015;121(22):3975-3981. doi: 10.1002/cncr.29596.