3041 — Unexplained and Medically Unjustified Trend toward Starting Dialysis at Higher Levels of Kidney Function Found Outside and Inside the VA
Hebert P, Perkins M, Lemon JM, Liu CF, and O'Hare A, Northwest Center for Outcomes Research in Older Adults, VA Puget Sound Healthcare System, Seattle;
Nationwide patients are starting dialysis at increasingly high levels of kidney function as indicated by the patient’s estimated glomerular filtration rate (eGFR). This is troubling because no evidence suggests early initiation of dialysis is beneficial to the patient, dialysis is costly to payers, and additional dialysis time is an extreme burden to patients. The extent to which early dialysis initiation is driven by its financial benefits to nephrologists and dialysis facilities, 80% of which are operated by national for-profit chains, is not known. Our goal was to determine whether similar trends were present within the VA, where no such financial incentives exist.
We assembled a cohort of all VA patients who initiated chronic dialysis in the VA from 2000-2008. We estimated each patient’s most recent level of kidney function prior to initiation of dialysis using three common methods: Modification of Diet in Renal Disease (MDRD), Chronic Kidney Disease Epidemiology (CKD-EPI), and the Cockroff-Gault (CG) equations. MDRD and CKD-EPI estimate GFR from patient’s age, gender, race, and serum creatinine; Cockroff-Gault also considers weight. Trends in eGFR at initiation among Veterans were compared to those for all US dialysis patients using aggregated data from the US Renal Data System (USRDS)—a national registry for end-stage renal disease—that were directly standardized to the VA population. Best-fit linear regression with bootstrapping was used to determine slope of the time trend.
12,840 patients had serum creatinine measured within 45 days of dialysis. The mean GFR at dialysis, between 2000 and 2008, increased from 10.6 to 12.1 (MDRD), 10.0 to 11.6 (CKD-EPI) and 14.8 to 18.4 (CG), respectively. The annual trends in eGFR as measured by C-G, MDRD, and CKD-EPI, were +0.45 mL/min/yr (95% CI, 0.017 to 0.89), +0.20 (CI, -0.21 to 0.60) , and +0.19 (CI, -0.21 to 0.60) per year respectively. In the USRDS, increased from 8.5 to 10.6 across the time period. The trend standardized to the VA population was +0.26 (CI, 0.15 to 0.37) and did/not differ significantly from the trends in MDRD, CKD-EPI, or C-G eGFR in the VA. Slopes of each time trend were non-zero with 95% confidence limit that did not include zero.
Trends toward initiation of chronic dialysis at higher levels of kidney function within the VA parallel those observed in the wider population of dialysis patients.
The absence of financial incentive did not insulate the VA from the burdensome and expensive trend toward earlier initiation of dialysis. More careful evaluation of dialysis initiation practices among Veterans is needed.