Validation of the Center for Medicare and Medicaid Services algorithm for eligibility for dialysis.

Abstract

The Center for Medicaid and Medicare Services (CMS) has recently revised their end-stage renal disease (ESRD) Medical Evidence Report, Medicare Entitlement, and Patient Registration CMS 2728 Form. The modified algorithm calls for the use of formulae to estimate glomerular filtration rate (GFR). The new criterion is defined as estimated GFR of less than 20 ml/min per 1.73 m(2). GFR is either estimated by Schwartz formula (C(SCH)) in children or Modification of Diet in Renal Disease formula (C(MDRD)) in adults. The purpose of this communication is to test the validity of the new CMS GFR algorithm in detecting children who need renal replacement therapy. We evaluated two cohorts of children. Group I included single-center data from 626 (125)I-iothalamate clearance studies (C(IO)) that were compared with the simultaneous estimation of GFR by C(SCH). Group II included data on 659 children from the patient incidence registry obtained from the ESRD Network of Texas between February 1996 and October 2003. In group I there were 76 children (76 C(IO)) with C(SCH) less than 20 ml/min per 1.73 m(2) of whom 50 (67%) had C(IO) less than 15 ml/min per 1.73 m(2). Of children with C(IO) less than 15 ml/min per 1.73 m(2), 62% had a C(SCH) less than 20 ml/min per 1.73 m(2). The ability of C(SCH) greater than 20 ml/min per 1.73 m(2 ) to predict C(IO) greater than 15 ml/min per 1.73 m(2 )(negative predictive value) is 0.95. The number of children who were started on dialysis in Texas within the study period was 659 (group II). The mean C(SCH)+/-SD was 10.8+/-7.7 ml/min per 1.73 m(2). Of the patients who were initiated on dialysis, 94% had C(SCH) less than 20 ml/min per 1.73 m(2). The results were sustained when race, gender, age range, and type of diagnosis were considered. The new CMS algorithm provides a good negative predictive estimate of GFR less than 15 ml/min per 1.73 m(2).