The case for implementing a data-driven, systems-based and outcomes-responsive renal disease management clinical model in the pre-ESRD population.
ABSTRACT: The prevailing model for renal disease care focuses on acute illness with little interdisciplinary concurrent risk factor intervention. Compounding the problem, the reimbursement structure of U.S. medicine treats end-stage renal disease not as a failure of prevention, but as an externality. There is a potential for improvement in clinical outcomes and decreased cost of care by using a multidisciplinary care model. This article establishes the case for implementing a data-driven, systems-based and outcomes-responsive renal disease management clinical model in the pre-ESRD population.
It is a truism — whether it originated with Albert Einstein, Benjamin Franklin or Fred Zamberletti of the Minnesota Vikings — that the definition of insanity is doing the same thing over and over and expecting it to come out differently. Sadly, American nephrology lives in a paradigm fitting that definition. Year after year, the current model of renal care has us taking on patients one at a time, disconnected from colleagues in other disciplines whose inputs are integral to collaborative, connected care. We treat isolated diagnoses rather than disease systems. We treat acute illness without providing longitudinal health-directed care. The result is an alarming rise in the national costs of renal care, from onset of chronic kidney disease to end-stage renal disease.
However, there is a different model of care that could dramatically change that.
CKD, generally defined as a glomerular filtration rate of less than 60 milliliters per minute, affects up to 15 percent of the population over age 20. Between 2007 and 2014, as many as 16 percent of the adult participants in the National Health and Nutrition Examination Survey — or as many as 34 million people — outside institutions had evidence of CKD, with 15 million to 18 million in advanced stages (stages 3 and 4). The prevalence of the more-advanced stages of this condition had been rising along with the annual incidence of ESRD.1
Between 1990 and 2002, the incidence of ESRD increased a startling 162 percent for patients with diabetes.2 The National Institute of Diabetes and Digestive and Kidney Diseases recently has published statistics showing a slow national decline in ESRD incidence. “After rising steadily from 1980 to 2001,” it notes, “the incidence rate of ESRD leveled off in the early 2000s and has declined slightly since 2006.” However, in the same paragraph, the authors note: “Although the number of ESRD incident cases plateaued in 2010, the number of ESRD-prevalent cases continues to rise by about 21,000 cases per year.”3,4
Although better blood pressure and diabetes care might be responsible for the slight decrease in the overall number of patients with Stage 3 and worse renal function seen nation-ally, the sub-group declining toward ESRD at a rate of 2 to 5 milliliters per minute per year in our high-prevalence region is substantial, creating a future potential tsunami of ESRD patients. This is likely due to high-average body mass index, poor diet and associated hypertension, obesity and renal hyperfiltration. These patients, therefore, are disproportionately predisposed to progression to ESRD. Out of the 53,000-plus patients under primary care management at our medical center, more than 8,000 were found to be in stage 2 or higher. More than 4,000 were in stage 3 or higher. Some of those have up to five risk factors placing them at greatly increased risk for progression.
The risk factors for progression to ESRD are known and well-documented.5,6,7 The risk of progressive loss of renal function is elevated by starting from a baseline of decreased renal filtration, having diabetes, being overweight (BMI greater than 30), smoking, having a systolic blood pressure greater than 130, losing significant amounts of protein into the urine, and losing more than 3 to 5 milliliters per minute of renal filtering ability each year. There also are less-substantiated risk factors, such as metabolic acidosis and asymptomatic elevation of serum uric acid.7,8,9 All of these risk factors should be reasonably well-documented in robust EMRs.
Accurate prediction of renal failure risk should be possible within an enrolled patient base if there is coverage of that data over enough time to have a statistically valid sample of each risk factor.
Currently, renal disease management in the United States is fragmented. The prevailing commercial insurance cover-age model results in patients frequently changing health care plans. The lack of secure communication between different clinical EMRs also fragments the macro-database, rendering data available to any one nephrologist inadequate to allow proactive, data-driven, prevention-centered, systems-based and outcomes-responsive renal care over time periods that are commensurate with the evolution of renal disease. Attempts to deal with this enormous problem year after year have consisted of putting minor tweaks on an incident-illness-treatment model of care. This approach has been getting largely the same results each year for the past decade. That surely meets the definition of insanity.
ESRD/renal replacement therapy care is incredibly expensive. According to Medicare, U.S. expenditures for the treatment of chronic kidney disease exceed $48 billion a year. Hemodialysis costs a patient roughly $80,000 annually. Peritoneal dialysis costs roughly $60,000. Renal pre-transplant evaluation costs at least $20,000.10 The cost of transplant surgery varies by where it is performed and how it is estimated; it ranges from $143,500 to $260,000. First-year post-transplant follow-up costs range from $17,000 to $30,000.11 Therefore, every year, ESRD and the resulting dialysis or transplantation can be delayed by conservative outpatient management, accruing per-patient, per-year savings of between $17,000 and $90,000 across the high-risk population.
Since ESRD costs in the United States are not charged to pre-ESRD primary-care organizations, potential savings from delaying or preventing ESRD currently become an economic externality. Rather than being accounted as the cost of failed prevention, ESRD costs are subsumed by the federal government. Accordingly, the potential systemwide cost savings of preventing progression are not reflected in the primary care system. This results in it are neither incentivized nor structured to make maximal efforts to delay CKD progression.
Researchers have shown that multidisciplinary care for ESRD patients on dialysis confers a dramatically lower hazard ratio (0.48) for death.12 In Taiwan and Canada, there also is highly encouraging evidence of similar benefits for pre-ESRD patients, including a decreased rate of renal decompensation, decreased ESRD costs and better quality of life for patients exposed to multidisciplinary and integrated care applied be-fore dialysis.13,14,15,16 Notably, it appears that overall economic benefit is realized when the cost of renal replacement care is borne by the same system that provides pre-ESRD primary care. This both incentivizes and documents the rewards of prevention-centered care.
Addressing risk factors early in the disease process (stages 2-3), when the number of remaining kidney filters sharing the workload is higher, holds the greatest promise of slowing or arresting functional decline. However, in one of the authors’ renal clinics, patients have stabilized for several years with as few as 17 to 20 milliliters per minute of remaining renal function following intensive education and risk-factor management. This avoids or significantly delays the need for renal replacement therapy.
The Veterans Health Administration system has a number of factors that make it an ideal testing ground for integrated, prevention-centered, multidisciplinary pre-ESRD care. The veteran population nationally is aging, and veterans tend to have multiple CKD risk factors. Furthermore, there is a high prevalence of chronic back pain with prolonged excessive use of non-steroidal anti-inflammatory medications that are toxic to the kidneys.17
The VHA has a single EMR, called VistA, designed by clinicians for patient care delivery. It is connected nationally by a web-based, secure, digital superstructure known as the Decision Support System. That system permits any VA provider to access all data inside in the VA system for any patient from any site across time. [Editor’s note: The VA announced earlier this year that it will move to an EHR system that integrates with one currently being used by the Department of Defense.] Using that rich DSS database, we have obtained data on the prevalence of stage 2, 3 and 4 chronic kidney disease within our medical center, the James H. Quillen/Mountain Home VA Medical Center, in Mountain Home, Tennessee, for the enrolled population between fiscal years 2013 and 2015.
For these three years, the prevalence has been fairly constant at 4,000 patients in stages 3-5 out of a total population of approximately 53,000 enrollees, but the incident progression to ESRD has been rising. That data has been parsed by renal disease risk factors, provider enrollment teams and geo-graphic location. The number of veterans in each CKD stage is known by multiple demographic elements, including residence and nearest VA clinic.
The VHA has decades-long continuing “ownership” of costs and outcomes of more than 7 million lives under care, including dialysis and transplantation costs. It is therefore effectively a “single-payer” system similar to that in Taiwan, where a landmark study of prevention-centered, multidisciplinary care was performed. This uniquely positions it to benefit from the same comprehensive, multidisciplinary disease management strategy.
Other large, nongovernmental health care systems with similar organizational characteristics should be able to benefit economically as well, although perhaps to a lesser degree (see Figure 1).
Research from Taiwan documented a 30 percent decrease in the annual incidence of ESRD from this type of nephrology care structure.16 The model reported in this research was implemented in a single-payer environment and delayed onset of dialysis, thereby averting substantial costs of care. In ad-dition to delaying the onset of ESRD, the study group also promoted using alternatives to hemodialysis that were more cost-effective and promoted a better quality of life, such as renal transplantation and peritoneal dialysis. There was reduced reliance on catheters for dialysis and smoothing of the transition to the treatment modality most suitable for the patient. This reduced mortality and morbidity, and made the care more cost-effective.
One of the authors of this paper found in his Wake Forest practice that by using strong educational programs and a sound home training program, he was able to put 33 percent of his ESRD patients on peritoneal dialysis over a two- to three-year period. This was accompanied by a marked improvement in patient quality of life and reduced rehospitalization rates. In contrast, only about 5 percent of ESRD patients are on peritoneal dialysis in our medical center’s region.
THE PROPOSED MODEL
Our VA medical center’s executive leaders have approved establishing a multidisciplinary model for renal disease management upon completion of a facility renovation project that, as a side effect, will create a near-perfect physical site of care for the clinic. The proposed model follows the one outlined in Taiwanese research. It requires the integrated (and co-located) efforts of a renal dietitian, a nurse educator (for reality-orientation and motivation), a pharmacist (for medication reconciliation and review of all non-prescribed OTC medications and dietary supplements), a data manager, a clinical social worker, a dedicated renal transplant coordinator, LIP primary care managers, dialysis nursing (for pre-need orientation and motivation) and nephrologists.
Vascular surgeons (for hemodialysis access planning) and general surgeons (for peritoneal dialysis assessment and planning) will be closely involved for all enrolled patients, but separately located from the renal disease clinic. EMR data on all patients of the health care organization will be mined and parsed in terms of the number and severity of risk factors, and patients then will be prioritized by number and severity of risk factors. Once identified and prioritized, the highest-risk patients will be offered enrollment in the single co-located clinic.
The criteria for consideration will be the number and severity of risk factors, ordered by the literature-supported strength of those risk factors:
- Absolute estimated GFR M of less than 60 milliliters per minute.
- Age older than 65.
- Rate of loss of GFR (dGFR/dt) of more than or equal to 5 milliliters per minute per year.
- Diabetes mellitus with hemoglobin A1c greater than or equal to 7.5 percent.
- Hypertension (systolic greater than 140, diastolic greater than 90).
- Proteinuria (random urine protein-to-creatinine ratio greater than 0.3, or urinary albumin-to-creatinine ratio of greater than 30 milligrams per gram).
- Smoking (by history, or with positive urine nicotine metabolites).
- Obesity (BMI greater than 30).
- History of regular use of over-the-counter or pre-scribed nonsteroidal anti-inflammatory medication.
Patients visiting a nephrologist or primary care physician in this clinic will be seen by other team members at nephrologist visits and also at dedicated visits. Since education, medication reconciliation, dietary counseling and social work expert resources would be available via nonphysician team members, follow-up appointments with a nephrologist could be as short as 20 minutes. The clinic’s data manager prospectively will search the digital database across significant time periods to identify those with one or more risks for CKD progression as early as stage 3. All team members then will have access to a current patient-level summary of that data. The number and intensity of the risks as well as the current renal functional status will be tracked in real time.
In virtually all cases, this clinic’s care would supplement and not replace that of a primary care provider who would attend to the patient’s nonrenal health problems. This clinic structure would allow three nephrologists seeing outpatients eight hours a day to have 7,200 encounters each over 300 clinic days a year, or 21,600 encounters a year. Even if all 4,000 veterans in stage 3 or worse in our medical center’s catchment area were to see a nephrologist, that would allow for five annual visits per veteran. However, because the stable stage 3 patients would not require repeated nephrologist en-counters in such a multidisciplinary clinical setting, there would be sufficient capacity for those showing evidence of declining function to be seen even more frequently.
Iterative reviews will be conducted of the risk factors and renal function profiles of the clinic patients and compared to those same indices of patients with similar risk profiles who were unwilling or unable to participate in the program. This will permit assessment of the effectiveness of the clinical model (see Figure 2) based on real-time data. Further sub-group analysis by risk factor will allow detection of differential effects of risk factors on chronic kidney disease progression. The primary end points will be the rate of GFR loss in milliliters/minute and the annual incidence of ESRD. The cost effectiveness of the clinical model will be continually reassessed, and the level of clinic emphasis on individual risk factors will be adjusted accordingly.
Total labor plus additional hardware costs of the above co-located multidisciplinary clinic model using prevailing federal full-time equivalent costs are estimated be between $1.1 million and $1.8 million a year. Real data on incidence and prevalence of ESRD in our facility’s catchment area were used to construct a cost-savings model to explore this model’s economic viability (see Table 1).
Our cost projection was assumed to start in fiscal year 2016 for modeling purposes. In fiscal year 2015, only six of 124 ESRD patients at our medical center were on peritoneal dialysis, or 4.8 percent. The calculations were based on the observed year-on-year rise in ESRD incidence of six to 12 patients. That was the year all existing dialysis contracts were renewed, and data did not distinguish new from old contracts.
To stress the model, we assumed the incremental addition of dialysis patients would be at the low end of that range, and that the annual per-patient, per-year cost of hemodialysis care in the contracted VHA environment, including medications provided by the VHA over and above those given at the dialysis clinics, would be $80,000. The VHA’s build-vs.-purchase model at the time of projection was $48,048 per patient per year for dialysis, which is less than the commonly quoted national average. To stress the model further, it was the figure employed in the projections. Because no official VHA build-vs.-buy cost for peritoneal dialysis was published at the time the projection was run, it was assumed that the total cost of peritoneal dialysis care would be $60,000 per patient per year, quoted frequently as a national average.
For projection purposes, it was assumed that intensive multidisciplinary renal disease management care would be capable of delaying 20 percent of the incident patients annually. This is 10 percent less than was observed in the Taiwan study. As a separate projection, it was modeled that rigorous use of peritoneal dialysis whenever clinically appropriate could place 15 percent of the incident ESRD patients on lower-cost dialysis.
The predicted estimated cost savings are significant, and could repay the increased staffing cost of the multidisciplinary clinic relatively quickly.
This radical redesign model for renal care in the VA was presented in February 2017 at the VHA-Booz Allen Hamilton Renal Innovation Summit in Washington, D.C. It is under consideration, with funding being a limiting consideration.
In a reimbursement system where the cost of ESRD is treated as the cost of failed prevention, dialysis and transplantation expenses are no longer economic externalities. They become a metric for assessing the adequacy of the system to delay or avert a potentially preventable, expensive disease.
There is currently one published report of such a system in Taiwan, which achieved a 30 percent decline in the annual incidence of ESRD, so there is supporting literature for the potential for significant ESRD care cost savings through creation of an intensive, data-driven, prevention-focused and outcomes-directed multidisciplinary renal disease management clinic. These cost savings could pay for the upfront costs of creating the model, and reduction of ESRD incidence would result in improved patient quality of life.
Organizations where the necessary robust EMR and longitudinal patient care conditions exist should consider embracing a multidisciplinary model that prospectively identifies and enrolls patients at elevated risk. A data-driven, systems-based, and outcomes-responsive multidisciplinary renal disease management clinical model should be piloted in the United States, and we are proposing such a model within the VHA system.
Terrence Jay O’Neil, MD, is nephrologist emeritus at the James H. Quillen/Mountain Home VA Medical Center and a clinical professor at East Tennessee State University’s Quillen College of Medicine.
William Scott Moore, MD, is a nephrologist and medical director of the James H. Quillen/Mountain Home VA Medical Center dialysis unit, and a clinical professor at East Tennessee State University’s Quillen College of Medicine.
- Prevalence of CKD in the General Population / Prevalence of CKD Stages, 1988-1994k versus 1999-2006 versus 2007-2014. CDC. At nccd.cdc.gov/ckd/detail.aspx?Qnum=Q8.
- Incidence of End-Stage Renal Disease Among Persons With Diabetes —United States, 1990-2002. CDC, MMWR 54(43);1097-1100
- Kidney Disease Statistics for the United States; NIDDK Health information/Health Statistics, December 2016
- Incident and Prevalent Counts By Quarter. United States Renal Data System, April 2017. At usrds.org/qtr/default.aspx.
- Tangri N, Stevens LA, Griffith J, et al., A Predictive model for Progression of Chronic Kidney Disease to Kidney Failure, JAMA, 2011; 305 (15): 1553-1559
- Hunsicker LG, Adler S, Caggiula A, et al., Predictors of the progression of renal disease in the modification of diet in renal disease study; Modification of Diet in Renal Disease Study Group. Kidney International, 1997; 51: 1908-1919
- Kovesdy CP, Coresh J, Ballew S, et al., Past decline versus current eGFR and subsequent ESRD risk, J Am Soc Nephrol, 2016; 27: 2447-2455
- Chen W and Abramowitz MK, Metabolic acidosis and the progression of chronic kidney disease, BMC Nephrology, 2014; 15: 55
- Johnson RJ, Nakagawa T, Jalal D, et al., Uric acid and chronic kidney disease: Full Review, Nephrol Dial Transplant. 2013 Sep;28(9):2221-8
- Honeycutt AA, Segel JE, Zhou X, et al. Medical Costs of CID in the Medicare Population. J Am Soc Nephrol, 2013; 24(9): 1478-1483
- Kidney Transplant Cost. Cost Helper Health. At health.costhelper.com/kidney-transplant.html
- Wolfe RE, et al., “Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation and recipients of a first cadaveric transplant”, N Engl J Med 1999; 341:1725-30]
- Ravani P, Marinangeli G, Tancredi M and Malberti F, Multidisciplinary chronic kidney disease management improves survival on dialysis, J Nephrol, 2003 Nov-Dec;16(6):870- 7
- Bayliss EA, et al., Multidisciplinary Team Care May Slow the Rate of Decline in Renal Function, Clin J Am Soc Nephrol. 2011 Apr; 6(4): 704–710.
- Chen PM, et al., Multidisciplinary Care Program for Advanced Chronic Kidney Disease: Reduces Renal Replacement and Medical Costs, Am J Med (2015) 128, 68-76.
- Curtis BM, Ravani P, Malberti F, et al., The short- and long-term impact of multi-disciplinary clinics in addition to standard nephrology care on patient outcomes, Nephrol Dial Transplant. 2005; 20: 147-154
- Walter H. Hörl, Nonsteroidal Anti-Inflammatory Drugs and the Kidney, Pharmaceuticals 2010, 3, 2291-2321