End Stage Renal Disease
End stage renal disease (ESRD) is the final stage in Chronic Kidney Disease (CKD). CKD is defined in different stages based on the Glomerular Filtration Rate (GFR); normal GFR is typically > 90ml/min. Patients with a GFR < 15ml/min are designated as CKD Stage 5; patients with CKD stage 5 who are dependent on renal replacement therapy (RRT) AKA dialysis are classified as having ESRD₁.
An estimated 650,000 Americans undergo dialysis to treat ESRD₂. All patients with ESRD are eligible to enroll in Medicare (regardless of age)₃. Patients with ESRD have a number of comorbidities including: hypertension, electrolyte abnormalities, anemia, coronary/peripheral/cerebral arterial disease, congestive heart failure, poor nutrition and a higher risk of infections. These risks were highlighted during the recent COVID-19 pandemic; the ESRD population had the highest hospitalization rates due to COVID-19 among Medicare beneficiaries₄.
Although ESRD patients represent approximately 1.2 percent of the Medicare population, given their comorbid conditions, they account for 6.3 percent of Medicare spending - totaling more than $30 billion₅. The per-person, per-year spending on an ESRD patient averages nearly $80,000₆!
Prior Payment Models
Traditional payment models have focused on a fee-for-service model which pays nephrologists, dialysis companies, and hospitals on a per encounter basis. Given this model, there were no incentives for coordination of care to keep this population healthier over a longer period of time. However, given the chronic nature of the disease, multiple comorbidities, and high number of healthcare touch points (ex: 3x/week of in-person dialysis sessions and/or 1x/mo dialysis clinic visit), a value based care model is ideal in this patient population.
In 2015, through the Center of Medicare & Medicaid Services’ (CMS) Comprehensive ESRD Care (CEC) initiative, the first Accountable Care Organizations (ACOs) were created specifically for ESRD patients. The ESRD Seamless Care Organizations (ESCOs) brought dialysis facilities and nephrologists together to try to coordinate care. Dialysis centers and nephrologists continued to be paid in a fee-for-service model but ultimately participated in a shared-savings risk-adjustment model₈. In 2017, the 2 year evaluation of the program found mixed results. The model reduced Medicare spending by $68 million, but because of inaccurate spending projections leading to overpaid shared savings, Medicare experienced aggregate net losses of $46 million₉.
A leading theory for why the model did not meet its potential is that it focused on ESRD patients only and centered around dialysis providers, NOT nephrologists. Without engaging, educating, and preparing patients upstream before they transition to ESRD, it is difficult to make meaningful changes after kidney failure occurs. Similarly, the core competencies and financial incentives of dialysis facilities are oriented around fee-for-service treatment reimbursements, which makes it challenging to justify investments in upstream care or holistic care management₁₀.
Future Payment Models
Given the mixed results with ESCOs, CMS developed the next value-based care iteration, Kidney Care Choices (KCC), to focus on a broader and more upstream point of care including: CKD stage 4, CKD stage 5, dialysis, kidney transplantation, and post-transplantation related care₁₂.
Unlike ESCOs, the coordination of care in the KCC models is centered around the nephrologist.
KCC offers 2 subtype models: Kidney Care First (KCF), a capitated payment model and the Comprehensive Kidney Care Contracting (CKCC), an ACO-like model. These models are set to run from 2020 to 2023 with an option to extend the model till 2025₁₃.
The payment model for the KCF has 4 main components:
- The Adjusted Monthly Capitation Payment (AMCP) for ESRD patients
- The CKD quarterly capitated payments for CKD Stage 4 and 5 patients (CKD QCP)
- The kidney transplant bonus (KTB)
- The performance based adjustment (PBA)
Moving away from the fee-for-service model, the capitated payment model will provide a predictable amount of income that can allow practices to innovate on the services provided. Given that a practice must have at least 500 CKD patients and 200 ESRD patients to be eligible, at a minimum, each practice would receive $815,000/yr in capitated payments ($2200/year/pt for ESRD patients & ~$750/year/pt for CKD stage 4 and Stage 5 patients).
Additionally, the KCC models introduce the first bonus incentive to tie nephrology practice incentives with the overall health system. Given the mortality and cost benefit for kidney transplant over dialysis ($39,939/quality adjusted life years [QALY]) vs. $72,476/QALY), the KTB introduces a financial incentive for nephrologists whose patients receive a kidney transplant. In total, practices can earn $15,000 for each viable transplant₁₅. The bonus payments incrementally increase over time: Year 1 - $2,500, Year 2 - $5,000, and Year 3 - $7,500.
Finally, the PBA creates an upward or downward adjustment to the AMCP and CKD QCP based on the practice’s yearly performance. Practices are graded both on their relative performance compared to their peers and their overall quality improvement over time. A KCF practice’s revenue could be adjusted by -20% to +20% depending on their PBA rankings.
Quality Metrics in Nephrology
As more nephrology practices participate in the KCC models, there will be an increased need for robust data acquisition, processing, and analysis. One area that is primed for more growth & analysis is the delay of ESRD progression - a proposed quality metric by CMS. For the first time, practices would be reporting how well they are taking care of their CKD populations. Given the chronic nature of the disease and many different healthcare touchpoints (lab, imaging, path, & hospital data) - creating a well defined and usable database will be critical for reporting and quality improvement.
Machine Learning in Nephrology
Similar to Citiizen for Oncology, a Nephrology focused end-to-end solution would utilize different components of data refinery such as Natural Language Processing, Optical Character Recognition, Machine Vision & Machine Learning to collect and aggregate data to create clinically meaningful databases₁₇ ₁₈.
The V² approach
We are currently working with a Nephrology group that is participating in the CKCC model to build out their CKD progression database. Our most pressing problem right now is building out the OCR and NLP capabilities to read and output the myriad of lab reports. The overall structure of lab reports for a given test (ex: comprehensive metabolic panel) may be relatively similar across all labs, however, minor variability (ex: BUN/Creatinine vs BUN/Cr vs BUN/Creat) create a long tail of data that needs additional post-processing to be usable.
As value-based care within nephrology continues to grow, we believe clinical data science will be critical to appropriately treat populations at scale and improve the delivery of CKD care - ultimately keeping patients healthier in a cost-conscious manner.
References:
1. Chen TK, Knicely DH, Grams ME. Chronic Kidney Disease Diagnosis and Management: A Review. JAMA. 2019 Oct 1;322(13):1294-1304. doi: 10.1001/jama.2019.14745. PMID: 31573641; PMCID: PMC7015670.
2. https://adr.usrds.org/2020/end-stage-renal-disease/1-incidence-prevalence-patient-characteristics-and-treatment-modalities
3. https://www.cms.gov/Medicare/Coordination-of-Benefits-and-Recovery/Coordination-of-Benefits-and-Recovery-Overview/End-Stage-Renal-Disease-ESRD/ESRD
4. www.cms.gov/newsroom/press-releases/cms-proposes-medicare-payment-changes-support-innovation-and-increased-access-dialysis-home-settinghttps://
5. United States Renal Data System: Medicare expenditures for persons with ESRD. In: 2015 USRDS Annual Data Report: Epidemiology of kidney disease in the United States, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2015. Available at: http://www.usrds.org.
6. https://www.cdc.gov/kidneydisease/basics.html
7. https://www.kidneyx.org/
8. https://innovation.cms.gov/innovation-models/comprehensive-esrd-care
9. https://www.ajkd.org/article/S0272-6386(20)30752-6/pdf
10. https://strivehealth.com/the-evolution-of-value-based-care-in-nephrology/
11. https://twitter.com/nikillinit/status/1360280931408445445?s=20
12. https://innovation.cms.gov/innovation-models/kidney-care-choices-kcc-model
13. https://innovation.cms.gov/innovation-models/kidney-care-choices-kcc-model
14. https://innovation.cms.gov/files/x/kcc-kcf-infographic.pdf
15. D. A. Axelrod, M. A. Schnitzler, H. Xiao et al., “An economic assessment of contemporary kidney transplant practice,” American Journal of Transplantation, vol. 18, no. 5, pp. 1168–1176, 2018.
16. https://innovation.cms.gov/files/slides/kcc-kcf-intro-slides.pdf
17. https://www.ciitizen.com/research/
18. https://outofpocket.health/p/ciitizen-and-the-patient-data-marketplace
Vedant Acharya
Medical Student - 4th Year
University of Miami Miller School of Medicine
UPenn Interventional Radiology Residency Class of 2027
@vedantacharya20