The COTA Nodal Address Model

The COTA Nodal Address Model: A Novel Digital Classification System Identifies Variances in Cancer Care Cost to Support Value Based Care Delivery

by Andrew L. Pecora MD 1,2,3 Andrew Ip MD MS,3 Ching-Kun Wang MD,1 Stuart L. Goldberg MD,1,3 Andre H. Goy MD,1,3 Lili Brillstein, MPH, CEO,4 Glenn Pomerantz MD JD,5 Michael B. Atkins MD,6 and Donald M. Berwick MD7

1. COTA, Inc., New York NY 10005, USA; 2. Hackensack Meridian Health, Edison, NJ 08837, USA; 3. John Theurer Cancer Center at Hackensack University Medical Center, Hackensack, NJ 07601, USA; 4. BCollaborative; 5. Gateway Health, 444 Liberty Avenue, Pittsburgh, PA 15222; 6. Lombardi Comprehensive Cancer Center, MedStar Georgetown University Hospital, Washington, DC 20057, USA; 7. Institute for Healthcare Improvement, Cambridge, MA 02109, USA

Corresponding Author: Andrew L Pecora MD, John Theurer Cancer Center, 92 Second Street, Hackensack NJ 07601; phone 551-996-5900; email Andrew.Pecora@hackensackmeridian.org Keywords: Health care cost, International Classification of Diseases, Value-Based Health Insurance, Neoplasms
Grants or Research Funding: none

Abstract

Objective: We  have previously described a digital classification schema (Cota Nodal Address [CNA]) which incorporates validated prognostic elements. We propose to demonstrate that reducing variation in cancer care among similar patients might lead to decreased costs without affecting clinical benefit. Identifying variation is difficult under claims-based coding systems (i.e. ICD-10) that group cancer subtypes. We aim to use the CNA model to investigate variations in care.

Materials and Methods: We retrospectively categorized breast, colorectal, and lung cancer patients (1/2013 through 1/2016) using the CNA model. Variances in care, defined by bundling total cost, were descriptively identified in the sorted CNA subgroups. Costs were obtained from a health insurer, adjusted to Medicare rates, and validated against other insurers to be similar.

Results: A total of 4032 patients  were  sorted into 1,403 unique CNA disease states. The 10 most common CNAs for each disease contained a majority of each cancer (30% breast, 23% colorectal, 48% lung cancer). The average number of different care pathways  used  for each CNA group was 6.9 in breast, 4.9 in lung and 2.9 in colorectal cancer. Fewer than 10% of CNA groups were treated with a single pathway despite similar prognostic characteristics. The ranges of total cost of care for identical CNAs were wide in breast ($16,000-$88,000), colorectal ($25,000-$222,000) and lung cancer ($44,000-$424,000). The calculated savings potential by avoiding cost above the median  was $26,773 per patient.

Conclusions: Using the CNA model, variations in treatment were readily identified as well as the potential for reducing the number of care pathways and associated costs. Further investigation is required to validate the use of the CNA model to reduce costs in cancer care.

Implications For Practice

Identifying variances in cancer care cost is a major step towards supporting a value-based alternative payment model. The COTA Nodal Address (CNA), a new disease classification approach that incorporates key data elements, was able to identify significant variability in 4,000 breast, colon, and lung cancer patients. For these diseases, we were able to estimate conservatively a potential $26,000 savings per patient by not spending above the median cost. While further validation is needed, this real-world data shows the CNA model could improve value-based, data-informed decision making and ultimately drive healthcare costs down in the US and globally.

Introduction

Background and Motivation

Healthcare spending in the United States, including cancer care expenditures, is growing at unsustainable rates and often without clear association with outcome improvement.1,2 Although some of the increases are a result of new technologic advances, at least 20% of the nearly $3 trillion annual expenditure fails to enhance clinical outcomes and may be considered waste.3-11 Efforts to curb this rise have focused on value-based models that encourage clinicians to consider costs in their

“The COTA Nodal Address (CNA), a new disease classification approach that incorporates key dataelements, was able to identify significant variability in 4,000 breast, colon, and lung cancer patients. For these diseases, we were able to estimate conservatively a potential $26,000 savings per patient by not spending above the median cost.”

decision-making.5,6 Standardized treatment algorithms, such as specific care pathways that narrow treatment choices or payments for outcomes rather than process measures, are being increasingly employed.7,8 The Oncology Care Model is a prominent alternative payment model, as compared to traditional fee-for- service models, which hopes to bridge the gap to deliver quality cancer care based on defined meaningful measures.8

Identifying services, including confirmatory tests, with little to no likelihood of adding value might control costs without adversely affecting patient outcomes.9-13 In a majority of cancer patients, however, great variation exists in treatment choices based on presentation as shown in National Comprehensive Cancer Network (NCCN) guidelines. The International Statistical Classification of Disease and Related Health Problems (ICD-10) digital  coding system used in healthcare claims lacks clinical granularity in cancer (example: does not account for stage or genomics) to analyze for unwanted versus acceptable variation in care delivered.14 Consequently, analyses based on claims data often fail to provide insights encouraging providers to avoid low-value care choices.

COTA Nodal Address Approach

By grouping patients with similar characteristics together, variations in care become more apparent. We therefore developed a novel, precise digital classification system, the COTA Nodal Address (CNA), to cohort cancer patients accurately using evidence-based prognostic criteria and facilitate the application of precision medicine.15 The CNA is a digital classification system that incorporates clinical and disease-specific elements known to influence treatment decision and clinical outcomes. CNAs are divided into six numeric digit groups that characterize the field of medicine, general category of oncology, specific cancer type, phenotype, therapeutic intent, and progression track (see Figure 1). Elements used to derive the CNA phenotype include evidence-based patient specifics (e.g. age, sex, race, performance  status) and disease-specific parameters (e.g. stage, genomics).

An element is selected after  suggestion by expert peer review and confirmation by supporting evidence to validate its individual inclusion. For example, the CNA for breast cancer includes:

  • Age
  • Gender
  • Race
  • Menopausal status
  • ECOG performance status
  • Histologic subtype
  • Estrogen receptor status
  • Progesterone receptor status
  • OncotypeDx value
  • Breast cancer index
  • Her2neu oncogene status
  • Tumor size
  • Lymph node status
  • Lympho-vascular invasion status
  • BRCA-1 or BRCA-2 status
  • PALB-2 mutation status
  • Metastatic disease sites
  • Overall co-morbidity

Since each element may have multiple options, the number of possible CNA combinations for

breast cancer exceeds 500 million; by clustering common phenotype elements, the number of observed CNAs has been reduced dramatically (e.g. less than 1000 observed in over 10,000 breast cancer patient database).

In this study, we aim to investigate if the CNA model can identify variations in cancer care by comparing actual total costs to an idealized treatment bundle cost of care.  The  eventual  goal is to use the CNA model to support value‑based oncology care models and deliver more individualized care. We hypothesize that more accurate grouping of cancer patients by the CNA will help to show variation in cancer care costs across a large hospital system. We used a real-world dataset to measure the extent of variations and cost implications in treatment choices among patients with breast, colorectal, and lung cancer sharing identical CNAs.

Methods

A retrospective review of the health records of consecutive breast, colorectal, and lung cancer patients diagnosed between January 2013 and June 2016 treated at Hackensack Meridian Health institutions (16-hospital system in New Jersey) was performed. Each patient was retrospectively characterized using the CNA schema.

Total cost of care and cost by category of utilization was obtained for the patients insured by Horizon Blue Cross Blue Shield of New Jersey (HBCBSNJ). To standardize the cost of care, all costs of services provided were adjusted to Medicare fee schedules. To expand the sample size, we examined what therapy regimen was given to each patient, regardless of other insurance sources (no payer provided  data). Since the cost of the therapy is more than the chemotherapy regimen alone, we approximated the additional costs of delivery. To do this our disease experts listed the ancillary tests (based on NCCN guidelines), number of visits, and other utilizations that would be needed to deliver the regimen. We calculated the Medicare fee schedule for this idealized regimen/delivery pathway, and these costs and utilizations were assigned to the patient. As confirmation of the validity of this idealized costing, we examined the total cost variability between patients having HBCBSNJ and other insurers, noting marked similarities (data not shown).

To facilitate estimates of a prospective payment for an episode of cancer care model, multidisciplinary experts organized specific care pathway (SCP) groupings, referred to as bundles, to broadly account for all aspects of care.16 Twenty-five bundles (7 breast, 7 colorectal, 11 lung cancer), each containing a variety of NCCN guideline care pathways including surgical, chemotherapy, and radiation therapies (2,088 potential options; including 1,047 breast, 487 colorectal, 554 lung cancer) were developed.

Taking one example, Breast Cancer Bundle 1 contains all patients with Stage I-III disease who are hormone receptor positive but HER2/neu oncogene negative. This bundle has seven general care categories (hormone suppression only, endocrine therapy with non-anthracycline chemotherapy, endocrine therapy with anthracycline chemotherapy, endocrine therapy with radiation therapy, endocrine therapy with non-anthracycline chemotherapy with radiation, endocrine therapy with anthracycline chemotherapy with radiation, and endocrine therapy with surgery).

Within each general care category (GCC), there are many SCPs consisting of NCCN-approved, individual chemotherapy and/ or hormonal agent combinations (example, see Figure 2). The bundle time period starts on the date of pathology diagnosis and ends 12 months later for early stage (I-III) adjuvant/ neo-adjuvant bundles and 6 months later for advanced/metastatic bundles. For example, two identical patients (same CNA) with early stage, hormone receptor positive/her2neu oncogene negative breast cancer might receive either endocrine therapy alone or non-anthracycline chemo-endocrine therapy. Since breast bundle 1 contains all patients who are hormone receptor positive/her2neu oncogene negative, both patients would be included in bundle 1 analyses. The two patients with the same CNA classification would, however, be assigned to different general care categories (endocrine versus non-anthracycline based chemotherapy) and different SCPs (tamoxifen versus anastrozole). Descriptive statistics were used to identify total cost of care as well as variation in care comparing real-world data set and an idealized treatment bundle, with both cohorts standardized to the Medicare fee schedule.

COTA reviewed the health records and cost data and assigned the CNAs under business associate agreements for the purpose of ongoing healthcare operations from both HMH and HBCBSNJ. No data was specifically requested from either entity, or any patient, for the conduct of this study. Protected health information was removed, and the data was placed into the COTA Observational Database (whose structure has been reviewed by the Western Institutional Review Board, Puyallup WA). The de-identified data set was utilized for all analyses, in aggregate, without patient identifiers per waiver of consent for secondary research.

Results

Subject Characteristics and Assignment of CNAs

Health records of 4,893 consecutive cancer patients were reviewed from 72 medical, surgical and radiation oncologists, including 26 academic physicians at academic/tertiary care centers and 46 at community-based centers. The records of 4,032 (82%) patients contained the required elements to assign a CNA. Of these 4032 patients, 604 (15%) were insured by HBCBSNJ. Next, 1,403 unique CNAs were assigned for the 4,032 patients (581 breast, 534 colorectal, 288 lung cancer CNAs). The CNA distributions were not even, with the top 10 patient volume CNAs from each disease containing an over-represented population (30% breast, 23% colorectal, 48% lung cancer) indicating that each disease had a substantial population of patients with common CNA types and then a tail of rarer disease subtypes. Furthermore, 76% of the patients were assigned to a CNA with more than one patient.

Variation in Treatment Patterns

Physicians utilized 22 of the 25 (88%) available bundle groupings and 842 of the 2,088 (40%) SCPs. Among patients who shared identical CNA characteristics there was extensive variation in treatment selection. The average number of total SCPs used for each CNA was greatest in breast cancer (6.9),  followed  by lung cancer (4.9) and colorectal cancer (2.9) (see Table 1 and Table 2). Variation in SCP selection was lowest for CNA cohorts with the greatest number of patients per CNA (“common types”). It was uncommon for physicians to agree on a single SCP for patients sharing the same CNA, with one treatment being chosen in 4% breast, 9% colorectal, and 6% lung cancer CNAs. The extent of variation was similar for academic and community-based physicians and between patients insured by HBCBSNJ and other carriers (data not shown). Long term clinical outcomes were not assessed, but since nearly all patients received NCCN-concordant therapies, the care was considered “appropriate.”

Variations in Cost of Care

For HBCBSNJ patients (n=604), the total cost of cancer related care claims paid was $47,447,575 (adjusted to a Medicare fee schedule). In CNAs with three or more BCBSNJ patients (n=272), the total of cancer-related claims paid was $21,014,052 and the mean total cost of care per patient was $77,257; lung cancer was greatest (mean) per patient ($99,624), followed by breast cancer ($73,759) and colorectal cancer ($52,446). The sum of the total cost of cancer related care claims paid for all BCBSNJ patients above  the median total cost of care  in each CNA cohort was $6,913,325, representing 33% of total dollars spent above the median of the total $21,014,051 claims paid for CNAs with three or more patients. The total dollars spent above the median was similar (31%) when CNA groups

with only two patients were included in the analysis (n=318). Overall, the rate of total dollars spent above the median was greatest for lung cancer (39%) followed by breast cancer (32%) and colorectal cancer (16%).

Total cost of care varied between patients assigned identical CNAs as physicians selected multiple SCPs. The range of total cost of care  in patients with identical CNAs using different SCPs was substantial for breast ($16,000- $88,000), colorectal ($25,000-$222,000) and lung ($44,000-$424,000) cancer (Table 3). Total cost of care varied for all three cancer types when patients with identical CNAs were treated using the same SCP (Table 3), largely due to differences in ancillary services such as imaging, laboratory tests, supportive care drugs, hospitalizations and emergency room visits (data not shown). To identify the key drivers of cost, we compared the average cost of drugs and services per patient whose total cost of care was at the median to patients who had total costs of care above the median, similar to a previous analysis of variations in lifetime healthcare cost. The largest drivers of increased total cost of care included hospitalization/ emergency room costs, supportive care drugs (e.g., anti-emetics, growth factors), and radiation therapy. Among patients with breast cancer, the average hospital/emergency room related costs for patients at the median was $20,551 which increased to $59,859 for those above the median. Similar rises in radiation-therapy ($15,231 to $22,602), supportive drugs ($8731 to $10564), laboratory ($5785 to $6828), surgery ($3369 to $6555), chemotherapy ($5359 to $6465), E&M ($2964 to $3945) and imaging ($1585 to $1973) were noted.

Cost Savings Potential:

Adjusted to the Medicare fee schedule, the actual mean total cost of care per patient of HBCBSNJ patients at the median, for all CNAs with three or more patients, was $70,549. The mean total cost of care was $124,420 per patient among the HBCBSNJ patients whose to Using a single care pathway associated at the median cost of care in place of those above the median, the estimated calculated potential total savings per patient was $29,083.

Total cost of care was above the median total cost of care. Approximately $53,871 more was spent per HBCBSNJ patient on roughly half of patients whose total cost of care was above the median. By avoiding cost above the median, the calculated potential for total costs of care savings for the HBCBSNJ group was $26,773 per patient. The mean total cost of care for the group below the median was $38,585 per HBCBSNJ patient. For the 1,318 non-HBCBSNJ patients, there was also wide variation of care.

Discussion

Identifying low-value healthcare expenditure options can prove difficult without proper risk stratification. These difficulties are compounded in complex diseases, such as cancer, where multiple disease subtypes require individualized therapeutic strategies. Given that the current ICD-10 classification schema utilized in claims analyses fails to capture the granularity required for cancer risk stratification, we developed the CNA, which captures the relevant elements required to make clinical decisions and used CNA in conjunction with disease-specific SCPs

“Healthcare spending in the United States, including cancer care expenditures, is growing at unsustainable rates and often without clear association with outcome improvement. Although some of the increases are  a result of new technologic advances, at least 20% of the nearly $3 trillion annual expenditure fails to enhance clinical outcomes and may be considered waste.”

and bundles for our analyses.15 In this study we applied the CNA schema to a real-world cohort of breast, colorectal, and lung cancer patients and were able to identify significant variations  in cancer treatment care selection among patients with identical characteristics. Although a variety of treatments may be necessary for the multitude of different disease states, the number of appropriate choices should narrow after grouping patients into similar clinical/ biologic relevant cohorts, similar to NCCN guidelines. This study not only demonstrates the great variances in treatments and potential cost implications when patients with similar characteristics are treated in a real-world cohort, but also spurs the hypothesis that it may be possible to identify unwarranted variances at scale. Comparing “apples to apples” allows an evaluation of the outcomes and costs of similar patients in parallel.

Extensive variation was observed in the selection of SCPs for all three common cancers examined. Notably the highest variations in choices occurred among the CNAs with the fewest patients (uncommon subtypes) suggesting that a lack of standardization in treatment may be more prevalent in rarer CNA types. CNA cohorts with the greatest patient volumes had the least variations (incidence ranging from 6 to 25% for the most common CNA of each disease). Some variation may be appropriate for factors not captured by the CNA and/or patient preferences or social determinants of health, but the extent of unexplained variances observed is nonetheless concerning and warrants further investigation. We hypothesize breast and lung cancer cohorts had a higher median number of treatment pathways chosen compared to colon  due  to the more complex decision-making pathways available based on hormone/mutation/ histology variables.

Variation in treatment selection among patients with identical CNA  characteristics was associated with significant variations in total cost of care. Even when the same SCP was used for patients with identical CNAs, we noted considerable variations in total cost of care, indicating that other types of utilization (physician services, hospital services, laboratory services, supportive care medications) greatly affect overall cost variability. This likely explains the increased cost of care for lung cancer patients given the higher number of metastatic patients relative to colon or breast. Thus, adherence to a predetermined treatment pathway alone, without accounting for ancillary issues, will not achieve the full potential of reduction in cost sought in value-based programs.

Current trends towards payment reform encourage standardization of care through increasing use of guidelines. However, expert derived guidelines might also be flawed by including newer, more expensive options that marginally improve outcomes, if at all.17 It should be noted that the NCCN guidelines list multiple treatment options (with highly variable  costs) for similar patient types, frequently without benefit of comparative data on progression-free or overall survival outcomes. The NCCN  evidence block strategies are an important step in providing opinions regarding the relative efficacy and costs of the various approved regimens. In this study we did not attempt

to assess the “appropriateness” of treatment selections or adherence to NCCN guidelines.

The United States healthcare system is evolving from volume-based reimbursement to value-based methods, particularly in cancer care where costs are increasing, often absent a commensurate improvement in clinical outcomes.18,19 Value-based efforts to date, however, have not uniformly resulted in lower utilization or cost.20 Our stratification methodology suggests that there is the potential for two levels of reduction in total cost of care without adversely affecting clinical outcomes within risk-based value reimbursement programs. First, prospective assignment of the CNA may ensure that all relevant diagnostic and prognostic variables, including co-morbidities, performance status, and genomics which influence treatment are collected. This information is critical for optimizing choice of SCPs, by facilitating evidence-based treatment selection (and avoidance of expensive but less efficacious treatments).21 Second, once an optimal SCP is selected, it becomes feasible to limit cancer care related discretionary (non-emergent) expenditures to the median within that pathway. We have already worked with HBCBSNJ to pilot approved treatment bundles directly to

“By grouping patients with similar characteristics together, variations in care become more apparent. We therefore developed a novel, precise digital classification system, the COTA Nodal Address (CNA), to cohort cancer patients accurately using evidence-based prognostic criteria and facilitate the application of precision medicine.”

the clinician to systematically remove possible unnecessary expenses. The CNA schema provides a “lens” through which past patterns of variation in SCPs are observed, and thus, might be used to prospectively avoid unwarranted variations. With conservative assumptions we estimated a range up to $26,000 per patient (33% of total mean spend per patient using a Medicare fee schedule) that might be prevented by using CNAs to guide SCPs selection and avoid excess care within a SCP. An example of meaningful cost variance includes two patients with identical CNA that received the same chemotherapy agents but at differing intervals (weekly versus every three weeks) leading to 20% greater costs of care with the weekly therapy due to office visit, laboratory, supportive care and other non-chemotherapy charges (data not shown).

Our retrospective study has several potential limitations. A main limitation includes not analyzing warranted vs unwarranted variation, which would likely need a prospective study design. The clinical and treatment data was derived from a single hospital network, although New Jersey is a diverse state and hospitals included academic/community and metropolitan/rural sites. The study also involved limited numbers of patients over a 3.5-year timeframe, with changes in standard therapies over time potentially influencing variances. Approximate imputed costs were used for idealized bundles of care, which can be improved upon by analyzing actual patient data from large claims database compared to common CNAs. We focused on common cancers and suspect that we may be underestimating the variances in choices  for less common tumor types where consensus guidelines are not as available. Our study also did not evaluate outcomes for the different care pathways, although the CNA schema certainly can facilitate this. Thus, we cannot be assured that the cost savings projections are aligned  with optimal outcomes, although most of the care choices were part of the current NCCN guidelines and would be deemed clinically appropriate by most physicians and insurance providers. Lastly, we acknowledge that the CNA model, as a combination of validated prognostic variables, requires further validation in larger multivariate analyses.

As an application of a more granular clinical cohort-generation methodology, the CNA model holds promise to reduce total costs of care for the population while improving outcomes for the individual through greater optimization in care and identification of unwarranted variances. Further investigation is needed to validate the CNA model in identifying and explaining unwarranted variance in care, as compared to ICD schema. Prospective CNA assignment could eliminate the need for precertification and prior authorizations and thereby enable “precision payment” to match “precision medicine,” thus facilitating success of ongoing value-based reimbursement programs.

Conclusion

Using real world data on over 4,000 colon, lung, and breast cancer patients, a novel classification schema (CNA) was able to identify variations in cancer care cost. Using the CNA schema containing key oncology data elements, this proof-of-concept, while needing further validation and correlation with clinical outcomes, demonstrates a quantifiable model to estimate variances in cancer care in granular cohorts of cancer patients. Despite the increasing complexity of oncology care compared to other specialties, we argue that the CNA approach should encourage a long-term goal of achieving value-based cancer care. In the future, we will continue to investigate if the CNA approach can form the backbone of a bundled payment model in oncology.

REFERENCES 

  1. Berwick D, Nolan TW, Whittington J. The Triple Aim: care, health, cost. Health AFF. 2008; 27(3): 759-769 
  2. Gordon N, Stemmer SM, Greenberg D, et al. Trajectories of injectable cancer drug cost after launch in the United States. J Clin Oncol. 2018; 36(4): 319-325
  3. Berwick D, Hackbarth AD. Eliminating waste in US Health Care. JAMA. 2012; 307(14): e1-e4.
  4. Papanicolas I, Woskie L, Jha, AK. Healthcare spending in the United States and other high-income countries. JAMA. 2018; 319(10): 1024-1039
  5. Wang CJ, Cheng SH, Wu JY, et al: Association of a bundled payment program with cost and outcomes in full-cycle breast cancer care. JAMA Oncol. 2017; 3(3): 327-334
  6. Wong MD, Anderson R, Sherbourne CD, et al. Effects of cost sharing on care seeking and health status: Results from the medical outcomes study. Am J Public Health. 2001; 91(11): 1889-1894 
  7. Marsland T, Robbins G, Marks A, et al. Reducing cancer cost and improving quality through collaboration with payers: A proposal from the Florida Society of Clinical Oncology. J Oncol Pract. 2010; 6(5): 265-269. 
  8. Mullangi S, Schieicher S, Freely T. Outcomes measurements in value-based payments: Viewpoint. JAMA Oncol. 2017; 3(6): 1019-1020 
  9. Kelley R. Where can $700 billion in waste be cut annually from the US healthcare system? Thompsons Reuters: 2009. http://www.ncrponline.org 
  10. Elbner CE, Hussey PS, Ridgely MSet al. Controlling healthcare spending in Massachusetts: An analysis of options. Santa Monica, CA: RAND Corp; 2009. 
  11. Schoen C, Guterman S, Shih A, et al. Bending the curve: Options for achieving savings and improving value in US health spending. New York NY: Commonwealth Fund; 2007. 
  12. Thorpe KE, Ogden LL. The foundation that health reform lays for improved payment, care coordination and prevention. Health AFF (Millwood). 2010; 29(6): 1183-1187. 
  13. Delaune J, Everett W. Waste and inefficiency in the US healthcare system. Cambridge, MA: New England Healthcare Institute: 2008. 
  14. International Classification of Disease, Tenth Revision, Clinical Modification (ICD-10-CM). https:/www.cdc.gov/nchs/icd/icd10cm.htm 
  15. Pecora AL, Norden AD, Hervey J, et al: Development of a precise, clinically relevant, digital classification scheme for cancer. JCO Clinical Cancer Informatics. 2018; 10.1200/CCI. 18.00006 
  16. About the NCCN clinical practice guidelines in oncology (NCCN guidelines). http://www.nccn.org/professionals/ 
  17. Wagner J, Marquart J, Ruby J, et al: Frequency and level of evidence in recommendations by the National Comprehensive Cancer Network guidelines beyond approvals by the US food and Drug Administration: retrospective observational study. BMJ. 2018;360: K668 
  18. Basch E. Evaluating alternative payment models in oncology. JAMA Oncol. 2017; 3(3): 327-334 
  19. Spinks T, Ensuring quality cancer care: A follow-up review for the Institutes of Medicine 10 recommendations for improving the quality of cancer care in America. Cancer 2012; 15; 118 (100): 2571-2582 
  20. Lam MB, Figueroa JF, Zheng Z, et al. Spending among patients with cancer in the first 2 years of accountable care organization participation. J Clin Oncol. 2018; 36(29): 2955-2960 
  21. Gutierrez M, Choi K, Lanman R, et al: Genomic profiling of advanced non-small cell lung cancer in community settings: Gaps and Opportunities. Clin Lung Cancer. 2017;18(6):651-659 

About the Authors

Stuart L. Goldberg, M.D.Stuart L. Goldberg, M.D., serves as the Chief, Division of Value and Outcomes Research at the John Theurer Cancer Center and Associate Professor of Medicine at the Hackensack Meridian School of Medicine. Previously he was a co-founder and Chief Medical Officer/ Chief Scientific Officer of COTA. His research efforts, which include over 200 publications and a dozen book chapters, were recognized by the Association of Community Cancer Centers with the 2015 David King Clinical Scientist Award and he has been awarded Top Doc status by Castle Connolly. Dr. Goldberg earned his medical degree at the Pennsylvania State University in 1986. After internal medicine residency and hematology/oncology fellowships at the George Washington University, he served as a visiting fellow in bone marrow transplantation at the Fred Hutchinson Cancer Research Center in Seattle. He subsequently was the first dedicated fellow in bone marrow transplantation at the Mayo Clinic followed by serving as a full time faculty member at Temple University School of Medicine in the blood and marrow transplant program. He joined the John Theurer Cancer Center in 1998 and helped establish the adult unrelated bone marrow transplant program with a special interest  in supportive care issues. In 2004 he assumed the role of Chief, Division of Leukemia before stepping down to join COTA. In 2018 he utilized his big data skills to establish the John Theurer Cancer Centers Outcomes Division and he has been active in leading the Hackensack Meridian Health Network Covid-19 observational database.

C.K. Wang, M.D.C.K. Wang, M.D. is a medical oncologist who practiced in the Dallas/Fort Worth area from 2006 to 2017. During that time, he held multiple management and leadership positions including managing partner physician with Dallas Oncology Consultants, P.A., Director of Oncology with Medical Clinics of North Texas, P.A./ USMD and Cancer Program Chairman with the USMD Hospital in Arlington. Dr. Wang joined IBM Watson Health in 2017 as a Clinical Adoption Specialist and subsequently served as the Acting Deputy Chief Health Officer for Watson Health Oncology/Genomics and the Global Oncology Leader for the Watson Health Clinical Adoption Team. Dr. Wang joined COTA in 2019 as Senior Medical Director of clinical and medical operations and assumed the role of Chief Medical Officer in early 2020.

Donald M. Berwick, MD, MPPDonald M. Berwick, MD, MPP is President Emeritus and Senior Fellow at the Institute for Healthcare Improvement (IHI), an organization that Dr. Berwick co-founded and led as President and CEO for 19 years. He is one of the nation’s leading authorities on health care quality and improvement. In July, 2010, President Obama appointed Dr. Berwick to the position of Administrator of the Centers for Medicare and Medicaid Services (CMS), which he held until December, 2011. A pediatrician by background, Dr. Berwick has served as Clinical Professor of Pediatrics and Health Care Policy at the Harvard Medical School, Professor of Health Policy and Management at the Harvard School of Public Health, and as a member of the staffs of Boston’s Children’s Hospital Medical Center, Massachusetts General Hospital, and the Brigham and Women’s Hospital. He has also served as vice chair of the U.S. Preventive Services Task Force, the first “Independent Member” of the Board of Trustees of the American Hospital Association, and chair of the National Advisory Council of the Agency for Healthcare Research and Quality. He is an elected member of the American Philosophical Society, the American Academy of Arts and Sciences, and of the National Academy of Medicine (formerly the Institute of Medicine). Dr. Berwick served two terms on the IOM’s governing Council and was a member of the IOM’s Global Health Board. He served on President Clinton’s Advisory Commission on Consumer Protection and Quality in the Healthcare Industry. He is a recipient of numerous awards, including the 1999 Joint Commission’s Ernest Amory Codman Award, the 2002 American Hospital Association’s Award of Honor, the 2006 John M. Eisenberg Patient Safety and Quality Award for Individual Achievement from the National Quality Forum and the Joint Commission on Accreditation of Healthcare Organizations, the 2007 William B. Graham Prize for Health Services Research, the 2007 Heinz Award for Public Policy from the Heinz Family Foundation, the 2012 Gustav O. Lienhard Award from the IOM, and the 2013 Nathan Davis Award from the American Medical Association.  In 2005, he was appointed “Honorary Knight Commander of the British Empire” by Her Majesty Queen Elizabeth II, the highest honor awarded by the UK to non-British subjects, in recognition of his work with the British National Health Service. Dr. Berwick is the author or co-author of over 200 scientific articles and six books. He also serves now as Lecturer in the Department of Health Care Policy at Harvard Medical School.

Andre Goy, M.D.Andre Goy, M.D. is HMH Physician in Chief, Hackensack Meridian Health Oncology Services and Chairman and Chief Physician Officer of John Theurer Cancer Center (“JTCC”) at Hackensack University Medical Center as well as Academic Chairman – Oncology at the Hackensack Meridian Health School of Medicine.

HMH Oncology Program encompasses 18 hospitals system including John Theurer Cancer Center (JTCC) at Hackensack University Medical Center (HUMC), a member of the National Cancer Institute (NCI)- designated Lombardi Comprehensive Cancer Center Consortium. JTCC provides the largest oncology program in New Jersey and is also a member of the Memorial Sloan Kettering – Hackensack Meridian Health partnership. Dr.  Goy is also professor and Founding chair of the Department of Oncology at the Hackensack Meridian School of Medicine at Seton Hall University as well as Professor of Medicine at Georgetown University in Washington, DC.

Dr. Goy is widely known for his work in lymphoma, particularly in drug development in mantle cell lymphoma and other aggressive lymphomas.  His research interest also includes the identification of prognostic markers and better ways to stratify patients with lymphoma. Dr. Goy has served on the steering committee for the NCI for lymphoma and on the Scientific Advisory Board of the Lymphoma Research Foundation. Dr. Goy and his team have been involved with the NCI for almost a decade in the development of new forms of immunotherapy, including checkpoint inhibitors in lymphoma and CAR T-cell therapy, leading to the first approval of CAR-T cells in lymphoma, in November 2017.

Dr. Goy believes that a true revolution is happening in cancer medicine, which will be driven by cell-based therapies /immunotherapy, precise medicine, earlier detection, analytics and connectivity. The bioinformatics platform developed at JTCC – called COTA – uses a unique classification system compressing all relevant individual patient data into a usable code (CNA). This digital architecture allows us to track individual outcome, optimize decisions in cancer care, and hence reduce adverse variance in order to improve outcome at the individual level, while lowering costs for the population along the way and supporting value-based care initiatives.

Dr. Goy has published more than 180 peer reviewed papers and serves as reviewer for many journals in the field including the New England Journal of Medicine, the Journal of Clinical Oncology, and Blood, the American Society of Hematology Journal, among others. He has been the co-chair of the Global Council on the Future of Health and Healthcare for the World Economic Forum since 2015. He believes that we are at an inflection point in medicine due to unprecedented combination of rapid progress in science and discovery, but also technology which will help reshape both health and healthcare delivery.

Dr. Goy received his medical degree from University Joseph Fourier in Grenoble, France before completing his training in hematology oncology in Paris at the Faculté  of Medicine Cochin Port Royal. While there he also received a Master’s degree in Tumor Immunology from Pasteur Institute and Curie University Paris VI as well as a Master’s degree in Experimental Oncology from University Kremlin Bicetre. Dr. Goy was on faculty at Memorial Sloan-Kettering and M.D. Anderson Cancer Center before joining John Theurer Cancer Center in 2005. Dr. Goy has been leading the Lymphoma Program at JTCC, he became Chair and Director of JTCC in 2012, and has been appointed as Physician in Chief for Hackensack Meridian Health, the largest, most comprehensive and truly integrated health care network in New Jersey, with 18 hospitals from Bergen to Ocean counties.

Andrew Ip, MDAndrew Ip, MD, MS specializes in hematology/medical oncology, treating lymphoma and multiple myeloma. He joined the John Theurer Cancer Center’s Outcomes and Value Research division, specializing in big data and data science for the cancer center and Hackensack Meridian Health. He received his medical degree from Jefferson Medical College and his residency and fellowship training at Emory University School of Medicine, where he was chief resident and chief fellow. He also completed a Masters of Science in Clinical Research to prepare for his career in outcomes work.

Michael B. Atkins, M.D.Michael B. Atkins, M.D. is an internationally recognized leader in translational and clinical research. He began his career at Tufts Medical Center before moving to Beth Israel Deaconess Medical Center and being appointed Professor at Harvard Medical School where he served as Deputy Chief of the Division of Hematology/Oncology and leader of the Biologic Therapy and Cutaneous Oncology Programs, as well as Co-PI of the Harvard Skin Cancer SPORE, leader of the Dana Farber Harvard Cancer Center Kidney Cancer Program and Director of the DF/HCC Kidney Cancer SPORE. In 2012, he moved to Georgetown where he is the Deputy Director of the Georgetown Lombardi Comprehensive Cancer Center and William M. Scholl Professor and Vice Chair of the Department of Oncology. He leads the Lombardi Immunotherapy Initiative  and  institutional Pilot Grants. His current research focuses on immunotherapy for melanoma and RCC and biomarkers for response and toxicity. He has published over 450 original research and review articles and 4 books and has lectured extensively on these topics. He is past president of the Society for Immunotherapy of Cancer and  currently co-Chair of the  Melanoma  Research Foundation Scientific Advisory Council.

Glenn PomerantzGlenn Pomerantz is Senior Vice President of Health Services at Gateway Health with extensive experience in financial discipline and driving performance in the healthcare industry. Dr. Pomerantz oversees Care Management, Clinical Innovation, Pharmacy, Quality/Stars, Provider Relations and Network Management at Gateway.

Dr. Pomerantz has most recently served as Chief Medical Officer at Blue Cross Blue Shield of Minnesota and has held medical leadership roles at Horizon Blue Cross Blue Shield of New Jersey, Cigna, and Aetna. Throughout his career, Dr. Pomerantz has come to oversee all medical and care management strategies at Blue Cross Blue Shield of Minnesota. His expertise includes developing value-based reimbursement models and medical management.

Dr. Pomerantz began his career as National Medical Director at Cigna and Horizon Blue Cross Blue Shield of New Jersey before joining Blue Cross Blue Shield of Minnesota and Gateway Health. He also served as Instructor of Medicine at Harvard Medical School. He earned his M.D., from the Miller School of Medicine at the University of Miami and his J.D., from Boston University School of Law.

Andrew L. Pecora, MD, FACP, CPEAndrew L. Pecora, MD, FACP, CPE – Chief Executive Officer, Outcomes Matter Innovations, LLC

A nationally recognized expert in care transformation, Andrew L. Pecora, MD, FACP, CPE, is at the forefront of the health-systems improvement movement. As Chief Executive Officer of Outcomes Matter Innovations (OMI), he is a national advocate for making medical practices more efficient and profitable, and at the same time, improving patient care. Dr Pecora, one of Modern Healthcare’s 2019 “50 Most Influential Clinical Executives,” is dedicated to pairing practices with artificial intelligence (AI) and health technologies to increase efficiencies, reduce costs and enable physicians to better do what they do best – heal patients.

Dr. Pecora came to OMI from Hackensack Meridian Health, where he was President, Physician Enterprise and Chief Innovation Officer. A certified hematologist and oncologist – and one of the world’s leading experts in blood and marrow stem cell transplantation, cellular medicine and immunology research – Dr. Pecora played a pivotal role in the recruitment of leading oncologists, initiation of innovative research and clinical trials, expansion of patient care services and introduction of new, state-of-the-art technologies to shift administrative burdens from healthcare professionals.

While at Hackensack Meridian Health, he created and expanded the John Theurer Cancer Center, named one of the United States top 50 cancer centers – the only New Jersey institute with this designation – and played a key role in the success of multiple Hackensack Meridian strategic partnerships, including those with Memorial Sloan Kettering Cancer Center and Georgetown Lombardi Cancer Center.

Dr. Pecora was instrumental in the John Theurer Cancer Center approval from the National Cancer Institute as a research consortium member of the NCI-approved Georgetown Lombardi Comprehensive Cancer Center Consortium, awarded May 2019. This esteemed designation recognizes providers for their scientific leadership, community engagement, and the depth and breadth of their cancer research. The collaboration is one of just 16 NCI-designated cancer consortia. Along with innovating in clinical settings, Dr. Pecora is a noted leader in building physician alliances. He was instrumental in the creation and development of Regional Cancer Care Associates (RCCA), one of the nation’s largest oncology physician networks. Dr. Pecora is also the founder and executive chairman of COTA, a healthcare software analytics company guiding value-based care delivery for payers and providers. A medical innovator, he co-founded PCT Caladrius, LLC: a company focused on development, manufacturing and delivery of cell-based therapies, later sold to Hitachi.

Dr. Pecora has been awarded the prestigious BioNJ Dr. Sol J. Barer Award for vision, innovation and leadership; the ASCO Cancer Foundation Research Award and Gallo Award for outstanding cancer research; tapped as one of America’s Top Doctors, listed as a “Who’s Who in the World;”; and selected as the PM360 Uber ELITE

honoree. His award-winning research has been published in numerous peer-review medical journals and funded  by organizations including the National Cancer Institute, AHEPA Research Foundation and biopharma giants Amgen and Merck.

Dr. Pecora is a Professor of Medicine and Oncology at Georgetown University and University of Medicine and Dentistry of New Jersey. He is Associate Dean for Technology and Innovation at Hackensack Meridian School of Medicine at Seton Hall University. He received his medical degree from the University of Medicine and Dentistry of New Jersey and completed his fellowship in hematology/oncology at Memorial Sloan Kettering Cancer Center.

Lili BrillsteinLili Brillstein is a leading advocate for Episodes of Care/Bundled Payments, with a global reputation for successfully advancing and implementing value-based care models. She was the strategic force behind one of the most progressive Episodes of Care models in the US. Lili founded BCollaborative in July of 2019, which provides strategic advisory services to boards and c-suite stakeholders across the healthcare industry, helping to craft strategy and engagement in specialty care value based models. She works with providers, payers, pharma, start-ups and others to help advance the move from fee for service to value based care.

Lili was formerly the Director of Specialty Care Value Based Models for Horizon Blue Cross Blue Shield of New Jersey, and, under her leadership, built the largest, most progressive and most collaborative Episodes of Care program for commercially insured  patients  in the country.

Lili is a passionate advocate of Episodes of Care/Bundled Payments as a strategy to successfully migrate from fee for service to quality- & value-based models that rewards providers for excellent outcomes and patient experience, while reducing the overall cost of healthcare. One of the hallmarks of her perspective is the criticality of beginning with models that do not include a risk transfer to providers until all of the stakeholders have worked together in a retrospective, no risk model and achieved success. The goal is to cultivate functional partnerships aligned around the patient and improve their experiences and outcomes, and ultimately reduce unnecessary care and waste.

Lili is a guest lecturer on episodes of care/bundled payments at the Harvard Business School working with Michael Porter and team, and has co-authored several peer-reviewed articles on the subject of Episodes. In addition, Lili has served as an Advisor to CMS on bundled payments, and is on the Advisory Boards of the US Women’s Health Alliance and the Quality Cancer Care Alliance; both national coalitions focused on advancing value based care to improve quality and cost of care delivery. She is also an Adjunct Associate Professor at The Rutgers School of Pharmacy, and a member of the Board of Directors for the NJ Coalition to End Domestic Violence.

Lili’s expertise at cultivating functionally collaborative relationships between payers and providers across the full continuum of health care has allowed not only the pillars of the triple aim to be achieved, but the spirit of the relationships between the parties to shift from one of adversaries to one of collaborators.

Disclosure of Potential Conflicts of Interest

Andrew L. Pecora, C. K. Wang, Stuart L. Goldberg, Andre H. Goy, Donald M Berwick: Ownership and/or equity interest in COTA Inc.

Lili Brillstein, MPH, CEO, BCollaborative, No conflicts

Michael B. Atkins: No conflicts Andrew Ip: No Conflicts

Funding: none

NEXT