Yeatsh CHAPTER. William Butler

Transcription

1 Sickness brought me this Thought, in that scale his: Why should I be dismayed Though Lame had burned the whole World, as it were a coal, Now I have seen it weighed Against a soul? William Butler Yeatsh A Friend s Illness 12 Introduction 14 Diabetic care {glycosylated hemoglobin & lipid testing} {use diabetic testing supplies} 18 Prescription drug therapy for CVD {use ACE inhibitors or ARBs, beta & calcium channnel blockers, & diuretics} 16 Prescription drug therapy for diabetic { receiving insulin & other diabetic drugs} {diabetic with CVD} 11 Provider management anemia {provider-level anemia treatment} {hemoglobin by unit affiliation & type} {regional differences in anemia treatment} Clinical care CHAPTER 112 Clinical qaracteristics with hemoglobins less than 11 g/dl {patient distribution by diabetic status & access} {access infection rates} 114 Vascular access events {vascular access use in the first year dialysis}{vascular access events & complications} 116 Summary

2 dvancing comorbidity in the ESRD population particularly as it relates to the burden diabetes, cardiovascular disease, and infectious complications has led to an evolution in ways to improve short- and longterm outcomes. In the 199s, discussions on quality care were focused on delivered dialysis, anemia management, and vascular access. Recently, however, the National Kidney Foundation has built on the strong base its initial K/DOQI guidelines, expanding these guidelines into wider areas care and adding new momentum to concerns over cardiovascular disease as the major cause death in ESRD. On the next page we illustrate progress toward some these guidelines. Data from CMS s Clinical Performance Measures (CPM) project show that, as measured by Kt/V and URR, 9 percent hemodialysis are receiving adequate therapy. Seventy-four percent those on peritoneal dialysis are meeting the therapy target a Kt/V 2. or greater per week; this target, however, is currently being reviewed by the K/DOQI workgroups. Vascular access complications continue to be a central area concern; with associations repeatedly shown among the use native fistulas, the best long-term function, and the fewest complications, K/DOQI guidelines recommend that at least 5 percent new, and 4 percent prevalent ones, use this type access. But while fistula use has grown in each population, we are still far from reaching both these target levels and the levels now achieved in Europe and Asia. We are, however, closer to meeting the target for anemia a focus treatment since the introduction epoetin in the summer Eighty-three percent now meet or exceed the target hemoglobin level 11g/dl. While albumin level is traditionally used as marker visceral protein stores, it is also a marker inflammatory load; albumin synthesis is reduced and levels fall in response to infectious complications, and investigators have noted that the heavy inflammatory load with kidney disease is associated with low albumins. Under K/DOQI guidelines the albumin all should reach the lower limit normal; currently, however, only 36 percent meet this target. It is not clear if this objective can be reached without a major effort directed at the inflammatory load carried by with chronic kidney disease. On the next spread we examine diabetic care in the ESRD and general Medicare populations. The measurement glycemic control a cornerstone diabetes management has steadily improved, with 74 percent dialysis, and 67 percent those with a transplant, now receiving at least one glycosylated hemoglobin test each year. Forty-five percent Introducion 12 Clinical careh

3 the diabetic ESRD population now receive four or more tests, a measure significant progress. Since the actual level glycemic control is not reported in the Medicare data, more detailed assessment will require primary data collection. With cardiovascular disease the leading cause death in both diabetic and non-diabetic populations, lipid monitoring is also central to diabetes care. Fifty-eight percent diabetic ESRD received at least one lipid test in 2 quite different from the 75 percent level achieved in the general Medicare population. Prescription drug therapy for diabetes and its complications is an important gauge treatment intensity. The use hypoglycemic agents other than insulin has been central to the improvement glycemic control, with newer insulin sensitizers and secretagogues shown to be particularly effective at improving control in more complex. And clinical trials with ACE inhibitors, ARBs, and beta blockers have established their efficacy in the treatment heart failure and as renoprotective agents. In this chapter we introduce new data on drug therapy in diabetic and for cardiovascular disease; in Chapters One and Three we also look at therapy in the CKD and incident ESRD populations. Overall trends in anemia management are illustrated in the Précis; here we focus on management at the provider level. Individual providers appear to titrate epoetin doses to achieve specific hemoglobin levels; those with average levels g/dl, for example, give doses that maintain in this range. In general, when hemoglobin levels are low or high, % (target) 9% (2 actual) % pts with delivered Kt/V >_ 1.2 % (target) 9% (2 actual) % HD pts with URR >_ 65% % (target) 74% (2 actual) % CAPD pts with delivered Kt/V >_ 2. 5% (target) 3% (2 actual) New pts with AV fistula as first access 4% (target) 32% (2 actual) Prevalent pts with AV fistula as current access {5.1} meeting K/DOQI guidelines incident & prevalent adult dialysis. Kt/V & vascular access data from CPM, 2; URR & hemoglobin data from Medicare claims, 2; albumin data from Medical Evidence form, 2. % (target) 83% (2 actual) % pts with hgb >_ 11 g/dl % (target) 36% (2 actual) % pts with serum albumin >_ test s lower limit doses are adjusted so that in subsequent months these levels move toward the K/DOQI target range 11 g/dl or greater. Some for-prit providers maintain a large proportion at 12 g/dl or above, suggesting a different approach to anemia management. Hemoglobin levels consistently below 11 g/dl tend to occur in EPO-resistant, including those with infections, cancer, chronic blood loss, AIDS, inflammatory kidney disease, and cardiovascular disease. New data show that the use a dialysis catheter as the initial form vascular access appears to be growing, and that with this access have the highest URRs. These higher ratios may be related to recirculation, or to the longer treatment times necessitated by reduced blood flow rates or collapsing within the extracorporeal circuit. In with catheters as their initial access, replacement with another catheter occurs at five times the rate replacement with an internal access, and the rate infectious events is episodes per patient year. In whose initial access is a fistula, replacement with a catheter occurs up to twice as ten as replacement or revision the fistula. Replacement and revision rates are, as expected, much higher in with arteriovenous grafts, as are rates declotting and angioplasty procedures; infection rates, however, are lower, which may reflect the increased early use catheters in who later receive fistulas. Patients who initiate on peritoneal dialysis have PD catheter replacement rates.8 per patient year, and peritonitis rates up to 1.2; data on these complications may provide insight into the {5.5 7} ESRD are less increased rates infectious complications seen in this population (see Medicare to likely than general Chapter Six for more information), and into the higher have lipid testing at least once per year, with 42 percent receiving no event rates peritoneal tests. {5.27} Trends in hemoglobin dialysis in levels over time show that with high or low hemoglobin levels their second year converge towards the K/DOQI guidelines. therapy. {5.36} Patients with hemoglobin levels less than 11 g/dl have higher rates infections and thrombotic complications including sepsis and clotting in the six months after initiation therapy. {5.39} The use dialysis catheters in incident varies little by age, gender, race/ ethnicity, or diabetic status. with catheters used as the first access in approximately 7 percent new. Chapter highlights H4 USRDS Annual Data Report 13

4 Glycosylated hemoglobin (HbA1c) testing {5.2} Geographic variations in the percent receiving HbA1c testing Prevalent ESRD {5.3} HbA1c testing in ESRD & gen. Medicare pts General Medicare ESRD (prevalent) 4+ tests 3 tests (88.) 79.1 to < to < to <75.3 below 69.5 (55.5) (85.6) 79.1 to < to < to <75.3 below 69.5 (65.7) 2 tests {5.4} HbA1c testing in prevalent ESRD, by age, race/ethnicity, & modality Age 1 test No tests General Medicare 4+ tests Race/ethnicity 3 tests Dialysis Transplant tested 4 2 tests 4 1 test No tests 4 All White Black N Am Asian Hispanic Lipid monitoring {5.5} Geographic variations in the percent receiving lipid testing Prevalent ESRD {5.6} Lipid testing in ESRD & gen. Medicare pts General Medicare ESRD (prevalent) 4+ tests 3 tests (.5) 64.1 to < to < to <57.6 below 5.4 (34.5) (76.7) 64.1 to < to < to <57.6 below 5.4 (44.7) {5.7} Lipid testing in prevalent ESRD, by age, race/ethnicity, & modality Age tested 1 test 4 No tests Race/ethnicity Dialysis Transplant 2 tests General Medicare 4+ tests 3 tests 2 tests 1 test 4 4 All Diabetic care 14 Clinical careh White Black N Am Asian Hispanic No tests

5 Diabetic testing supplies {5.8} Geographic variations in the percent receiving diabetic testing supplies {5.9} Prescribed (per day) diabetic test strips Prevalent ESRD General Medicare ESRD (prevalent) 2+ strips 1-<2 strips <1 strip No strips receiving supplies 4 Age Dialysis Transplant (58.5) 44. to < to < to <39. below 31.9 (17.8) Race/ethnicity (55.3) 44. to < to < to <39. below 31.9 (27.1) {5.1} Diabetic testing supplies in prevalent ESRD, by age, race/ethnicity, & modality 4 General Medicare 4 2+ strips 1-<2 strips <1 strip No strips All White Black N Am Asian Hispanic lycosylated hemoglobin testing (HbA1c) is more common in the general Medicare population than in ESRD (Figure 5.2). In only one in five diabetic ESRD received any testing; by 1 2, however, the number had grown to 76 percent, close to the level in general Medicare (Figure 5.3). Forty-five percent diabetic ESRD now receive four or more tests annually an improvement, but still a significant distance from fulfilling the guidelines the American Diabetic Association, which recommend that all with complex disease burdens receive at least four tests per year. Testing rates are slightly higher in dialysis than in those with a transplant (Figure 5.4). And for reasons currently unclear, the youngest are least likely to receive testing. Except for areas in the Upper Midwest, the Northeast, Florida, Texas, and the West Coast, more general Medicare nationwide receive lipid testing than ESRD (Figure 5.5). Seventy-five percent diabetics in the general Medicare population receive some lipid monitoring, compared to 58 percent ESRD (Figure 5.6). Testing rates are dramatically different between the modalities 81 percent diabetic transplant are tested, compared to only 55 percent those on dialysis (Figure 5.7). And rates differ as well for Native Americans, with only 34 and 57 percent dialysis and transplant receiving a lipid test. In 1997 Medicare increased the benefits for diabetics to include glucose testing supplies. Both the ESRD and general Medicare populations have expanded their use this benefit over the last five years; nationwide, however, general Medicare are more likely to receive these supplies (Figure 5.8). In the diabetic ESRD and general Medicare populations, 62 and 58 percent, respectively, receive no prescription for blood glucose tests or reagent strips, and only 11 and 8 percent are prescribed supplies sufficient for testing two or more times per day (Figure 5.9). Transplant receive diabetic testing supplies more frequently than those on dialysis, while the lowest prescription rates occur in age 18 3 and in those Native American descent (Figure 5.1). {Figures 5.2, 5.5, & 5.8} 1, by HSA. ESRD: point prevalent with 9-day rule, 1, age on December 31 2, alive through that date, & with diabetes as primary diagnosis, as a comorbidity at initiation, or diagnosed in 1 (Medicare claims); includes only with Medicare Parts A & B as primary payor coverage through the whole period. General Medicare: entering Medicare before January 1, 1, alive & in the program through December 31, 2, age on this date, & with diabetes diagnosed in 1. Patients enrolled in an HMO, with Medicare as secondary payor, or diagnosed with ESRD during the period are excluded. Diabetic care tracked in 2. {Figures 5.3, 5.6, & 5.9} ESRD: point prevalent initiating ESRD therapy at least 9 days prior to January 1 the first year, age on December 31 the second year, alive through this date, & with diabetes as the primary cause ESRD, as a comorbidity at initiation, or diagnosed during the first year; includes only with Medicare Parts A & B as primary payor coverage during the two-year period. General Medicare: entering Medicare before January 1 the first year, alive & remaining in the program through December 31 the second year, age on this date, & with diabetes diagnosed in the first year. Patients enrolled in an HMO, with Medicare as secondary payor, or diagnosed with ESRD during the period are excluded. For both populations, HbA1c testing, lipid testing, & diabetic testing supplies (blood glucose tests or reagent strips for home blood glucose monitors) tracked in the second year; for lipid & HbA1c figures, tests are at least 3 days apart. {Figures 5.4, 5.7, & 5.1} point prevalent ESRD with 9-day rule, 1, alive through December 31, 2 & age on this date, & with diabetes as the primary cause ESRD, as a comorbidity at initiation, or diagnosed in 1; includes only with Medicare Parts A & B as primary payor coverage during the period. HbA1c testing, lipid testing, & diabetic testing supplies tracked in 2. H4 USRDS Annual Data Report 15

6 {5.11} Cumulative percent diabetic receiving insulin, by age {5.12} Cumulative percent diabetic receiving metformin, by age receiving drug : ESRD 1 2 receiving drug : ESRD {5.13} Cumulative percent diabetic pts receiving thiazolidinediones, by age {5.14} Cumulative percent diabetic receiving secretagogues, by age 4-44: ESRD 5-44: ESRD receiving drug receiving drug {5.15} Cumulative % diabetic CVD pts receiving ACE inhibitors or ARBs, by age {5.16} Cumulative % diabetic CVD receiving beta blockers, by age -44: ESRD -44: ESRD receiving drug receiving drug Prescription drug therapy for diabetic 16 Clinical careh

7 {5.17} Cumulative % diabetic CVD pts receiving Ca ++ channel blockers, by age receiving drug receiving drug receiving drug : ESRD : ESRD : ESRD {5.18} Cumulative % diabetic CVD pts receiving lipid-lowering agents, by age {5.19} Cumulative % diabetic CVD receiving anticoagulants, by age 1 2 or analyses prescription drug use we have used the Medstat MarketScan database Employer Group Health Plan (EGHP). Approximately percent diabetic ESRD receive insulin, compared to only percent non-esrd (Figure 5.11). This may be because ESRD are rarely prescribed metformin (the risk lactic acidosis increases with the degree kidney impairment), because diabetes is more advanced and difficult to control in ESRD, or because the ESRD population has a different mix Type 1 and Type 2 diabetes. Insulin use has remained stable in both ESRD and non-esrd, despite the addition several oral agents to the market. The percentage non-esrd with diabetes who take metformin has increased substantially (from 31 to 4 percent in age 44, and from 42 to 49 percent in those age 45 64; Figure 5.12). Metformin was used in less than 2 percent ESRD in 2. Only 7 16 percent diabetic ESRD receive thiazolidinediones (rosiglitazone, pioglitazone), a number substantially lower than the percent their non-esrd counterparts (Figure 5.13). It is interesting that these drugs are not being used more frequently in the ESRD population, as they are not contraindicated in ESRD, and can be used either with or without insulin to improve the sensitivity tissues to insulin. Secretagogues, such as sulfonylureas, nateglinide, and repaglinide, are also used more frequently in the non-esrd population (34 47 percent) than in with ESRD (9 17 percent; Figure 5.14). Figures show the cumulative percentage diabetic with cardiovascular disease (CVD) receiving key cardiovascular-related drugs. It is important to note that the subpopulation diabetic with CVD and age 44 is quite small, and thus conclusions cannot be drawn about this group. The percentage receiving ACE inhibitors/arbs, beta blockers, and lipid-lowering agents increased overall from to 2 in both ESRD and non-esrd. ACE inhibitor/arb use is lower in ESRD than non-esrd, while the reverse is true for the use beta-blockers (Figure 5.15). The use calcium channel blockers (CCBs) is also greater in the ESRD population; this may reflect the effectiveness dihydropyridone CCBs in treating hypertension in ESRD, and merits further investigation. Use CCBs in older non-esrd has fallen, while it has remained stable in their ESRD counterparts (Figure 5.17). The use lipid-lowering agents is greatest in non-esrd, despite the fact that all ESRD are considered to be at the highest risk coronary disease, warranting aggressive lipidlowering therapy (Figure 5.18). The general use anticoagulants (low-molecular weight heparins and coumadin) has decreased slightly in ESRD, but still remains greater than in non-esrd {Figures } prevalent EGHP with ESRD, age 64. Two-year study period includes a one-year selection period, used to define comorbidity, & a one-year observation period, used to count prescription drug therapy. shown are months in the observation period. H4 USRDS Annual Data Report 17

8 he cumulative percent dialysis taking ACE inhibitors or ARBs grew consistently during all study periods, and was slightly higher in the later years. The most noticeable increases have taken place in females and diabetics (Figure 5.). Less than 43 percent diabetic received lipid- lowering agents in 2, despite being at high risk for atherosclerotic heart disease. Use diuretics appears to be declining, which may reflect increasing vintage within the dialysis population. In dialysis with congestive heart failure, use ACE inhibitors, ARBs, and beta blockers generally increased, and {5.} Cumulative percent prescription drug use in prevalent dialysis, overall, by age, gender, & diabetic status ACE-I/ARBs: Age Gender Diabetic status -44 Male Female Diabetic Non-diabetic 4 receiving drug 4 Lipid-lowering agents Diuretics {5.21} Cumulative percent prescription drug use in prevalent dialysis with CHF, by age, gender, & diabetic status 75 ACE-I/ARBs: Age Gender Diabetic status -44 Male Female Diabetic Non-diabetic receiving drug Beta blockers Digitalis preparations Diuretics Prescription drug therapy for cardiovascular disease 18 Clinical careh

9 {5.22} Cumulative percent prescription drug use in prevalent dialysis with CVD, by age, gender, & diabetic status ACE-I/ARBs: Age Gender Diabetic status -44 Male Female Diabetic Non-diabetic 4 receiving drug Beta blockers Lipid-lowering agents {5.23} Cumulative percent prescription drug use in prevalent dialysis with hypertension, by age, gender, & diabetic status ACE-I/ARBs: Age -44 Gender Male Female Diabetic status Diabetic Non-diabetic 4 receiving drug Beta blockers Calcium channel blockers percent with diabetes have a prescription for these drug classes (Figure 5.21). Digitalis preparations are used in less than 16 percent with CHF, and nearly 28 percent those with diabetes as well received diuretic therapy in 2. In dialysis with CVD the use ACE inhibitors/arbs has remained steady since 1 (Figure 5.22). Rates are slightly higher in diabetics compared to non-diabetics. Patients with diabetes had the highest use lipid lowering agents in 2. There has been a noticeable increase in the use ACE inhibitors or ARBs since in various groups with hypertension (Figure 5.23). In females and diabetic, for example, use these agents grew percent between and 2. The use beta blockers in the diabetic has also increased markedly, from 35 percent in to 48 percent in 2. {Figures 5. 23} prevalent EGHP dialysis, age 64. Two-year study period includes a one-year selection period, used to define comorbidity, & a oneyear observation period, used to count prescription drug therapy. shown are months in the observation period. H4 USRDS Annual Data Report 19

10 {5.24} EPO dosing, by starting hemoglobin & unit affiliation: incident Mean weekly EPO dose/kg Overall Group 1 Group Starting hgb: <1 g/dl Non-chain Hospital-based All <11 g/dl 11-<12 g/dl 12+ g/dl Incident year {5.25} EPO dosing, by ending hemoglobin & unit affiliation: incident Mean weekly EPO dose/kg Overall Group 1 Group Ending hgb: <1 g/dl Non-chain Hospital-based All <11 g/dl 11-<12 g/dl 12+ g/dl Incident year {5.26} EPO dosing, by average hemoglobin & unit affiliation: prevalent Mean weekly EPO dose/kg Overall Average hgb: <1 g/dl Group 1 Group 2 Non-chain Hospital-based All <11 g/dl 11-<12 g/dl 12+ g/dl CPM year n order to reach hemoglobin levels recommended by the National Kidney Foundation s anemia practice guidelines, dialysis units are encouraged to follow EPO dosing guidelines. Dosing patterns are clearly different among different providers (Figures ). In who initiate dialysis with hemoglobins less than 1 g/dl and in those with hemoglobins less than 1 after six months, for example, Group 1 providers (Fresenius, DaVita, and Renal Care Group) appear to give higher doses EPO compared to all other providers, as much as 35 units/kg/week. In prevalent with average hemoglobin levels less than 1 g/dl, EPO dosing is similar in both groups chain-affiliated units, at units/kg/week; dosing is lowest in units that are independently owned, at 362 (Figure 5.26). It appears that providers are adjusting EPO doses accordingly with hemoglobins, and that the percent on IV iron rises with vintage (Figure 5.27). In most providers at least two-thirds have hemoglobins g/dl (Figures ). The percent providers whose have hemoglobins 12 g/dl or greater is highest in the Southwest and West, and mean EPO doses and IV iron use are highest in the eastern half the country, (Figures ). {Figures } incident hemodialysis with a first EPO claim within 3 days the ESRD start date & at least one EPO claim in each the first six months; first six months dialysis combined. Figure 5.24, hemoglobin determined from hematocrit value on Medical Evidence form; Figure 5.25, hemoglobin determined from claims during the six months following dialysis initiation. {Figure 5.26} prevalent hemodialysis in the USRDS database who are also in the CPM database as hemodialysis. Hemoglobin, weight, & EPO dose are determined from the CPM data. {Figure 5.27} period prevalent hemodialysis with at least three months dialysis prior to June 2, with a valid EPO claim in each month from June December 2, dialyzing at an identifiable provider, dialyzing at the same provider from June through their last valid EPO claim 2, & from a provider with at least ten such. The percent receiving IV iron is a cumulative calculation starting with the month June. EPO dose adjusted for inpatient days. Mean hemoglobin & EPO dose represent the average across providers, with each provider mean obtained by averaging values for eligible within a particular hemoglobin level within that facility. {Figures } providers period prevalent hemodialysis with at least three months dialysis prior to June 2, a valid EPO claim in June 2, & an identifiable provider, dialyzing at the same provider from June through their last valid EPO claim 2, & from a provider with at least ten such. All All units Chain 1 Fresenius, Group 1 Chain 2 Gambro, Group 2 Chain 3 DaVita, Group 1 Chain 4 Renal Care Group, Group 1 Chain 5 Dialysis Clinics, Inc., Group 2 Chain 6 Nat l Neph. Assoc. Group 2 NC Non-chain units HB Hospital-based units U Unknown affiliation Provider management anemia 11 Clinical careh

11 {5.27} Provider-level anemia treatment, by patient-level hemoglobin (g/dl) {5.28} Distribution hgb groups, by provider Cumulative % pts on iron EPO dose/wk (1,s units) Hemoglobin (g/dl) Hgb: provider avg. hgb in June 12+ EPO dose/week Iron Hgb in June 11-<12 Hgb in June <11 Mean June hgb: 12+ Mean June hgb: 11-<12 Mean June hgb: <11 June July Aug Sep Oct Nov Dec June July Aug Sep Oct Nov Dec June July Aug Sep Oct Nov Dec providers providers 4 All NC HP Unit affiliation (see table at left for codes) <12 <11 Freestanding Nonprit <12 <11 {5.29} Distribution hgb groups, by unit type Prit Hospitalbased {5.3} Geographic variations in provider distribution, by unit-level mean hemoglobin Hemoglobin 12+ g/dl Hemoglobin 11 <12 g/dl Hemoglobin <11 g/dl (47.3) 27.9 to < to < to <22.2 below 1.3 (3.4) (94.) 76.7 to < to < to <66. below 57.7 (43.) (5.56) 1.67 to < to < to <1.23 below.81 (.) {5.31} Geographic variations in provider-level mean weekly EPO dose (in thousands units), by unit-level mean hemoglobin Hemoglobin 12+ g/dl Hemoglobin 11 <12 g/dl Hemoglobin <11 g/dl (16.1) 14.8 to < to < to <13.5 below 11.7 (1.6) (16.6) 15.5 to < to < to <14.4 below 13. (12.1) (24.1) 21.8 to < to < to <19.9 below 18.3 (15.9) {5.32} Geographic variations in the percent who receive iron, by unit-level mean hemoglobin Hemoglobin 12+ g/dl Hemoglobin 11 <12 g/dl Hemoglobin <11 g/dl (86.5) 81.6 to < to < to <79.7 below 75.7 (56.1) (87.5) 82.5 to < to < to <79.4 below 76.4 (56.8) (85.4) 81. to < to < to <79.3 below 75.8 (57.3) H4 USRDS Annual Data Report 111

12 {5.33} Distribution incident pts w/hgb <11 g/dl {5.34} Incident pts, by hgb, access, & DM status {5.35} Incident pts, by hgb, disease, & DM status 11 Hemoglobin 5 All All Hemoglobin (g/dl) <11 g/dl 11+ g/dl 4 All <11 g/dl 11+ g/dl units) 8 4 EPO dose/week 5 Diabetic Diabetic receiving iron during month Mean EPO dose/week (1,s IV iron 75th percentile Median 25th percentile after initiation Non-diabetic Catheter AV fistula AV graft 4 4 Non-diabetic Solid Chronic tumors blood loss AIDS Inflam. CVD disease CHF ersistently low hemoglobin levels are an area concern, as they may reflect an increased burden comorbidity, infectious complications, or erythropoietin resistance secondary to inflammation or hyperparathyroidism. Here we show the characteristics whose hemoglobin level, on average, is less than 11 g/dl after the first six months ESRD treatment. Overall, hemoglobin levels in this group tend to rise slowly during these six months to just over 1 g/dl, with the lowest quartile at approximately 9.2 g/dl and the highest at 1.7 g/dl (Figure 5.33). Erythropoietin doses per week also tend to climb, with a median approximately 28, units per week and an inter-quartile range 18, 4,. At the end six months, approximately 56 percent with persistently low hemoglobins (less than 11 g/dl) have received iron. Compared to with higher hemoglobin levels, those with low hemoglobins have greater dialysis catheter use, lower use arteriovenous fistulas, and comparable use arteriovenous grafts (Figure 5.34). These also appear to have more diagnoses solid tumors and chronic blood loss, a greater likelihood AIDS, inflammatory diseases, and cardiovascular disease in general, and a markedly greater likelihood CHF (Figure 5.35). Again compared to with hemoglobin levels 11 g/dl and above, with lower levels have higher rates infections and thrombotic complications including sepsis and clotting in the six months after initiation (Figure 5.36). The rate septic episodes in those with lower hemoglobins, for example, is more than double that found in the higher hemoblobin group. Infection rates for all causes infections are also greater in these Clinical qaracteristics with hemoglobins less than 11 g/dl 112 Clinical careh

13 {5.36} Vascular access infections or thrombosis, by hemoglobin (g/dl).4 All : infection Sepsis Clotting.3 <11 g/dl 11+ g/dl Diabetics month patient Rate per Non-diabetics after initiation {5.37} Infection rates, by hemoglobin (g/dl) month patient Rate per Bacterial Viral <11 g/dl 11+ g/dl. Fungal Parasitic after initiation (Figure 5.37). Since infectious complications generate high cytokine levels, which in turn contribute to erythropoietin resistance, it is not clear whether anemia per se leads to these higher rates complications, or whether the anemia is a consequence persistent infectious events. {Figure 5.33} incident hemodialysis, combined, identified as having a mean hemoglobin <11 g/dl during five the first six months dialysis after day 9 (5,758 ). Mean EPO dose/ week adjusted for inpatient stays, so it reflects outpatient dosing only. {Figure 5.34} incident hemodialysis, combined, who are also in the CPM database. Access determined from CPM data. There are 4,91, 297 whom are in the hemoglobin <11 category. {Figure 5.35} incident hemodialysis, combined. Presence disease represents at least one Part A or Part B claim during the first six months incidence after day 9 with a diagnosis code for that disease. {Figures } incident hemodialysis, combined, unadjusted. Month represents month after day 9. Rates for each month represent the rate from Month 1 (after day 9) through that month. Data on device infection & sepsis obtained from Part A or B claims; multiple claims for the same infection on the same day are deleted. H4 USRDS Annual Data Report 113

14 {5.38} with a catheter as their first access {5.39} with an AV fistula as their first access Age Gender Race/ethnicity White Black N Am Asian Hispanic Female Male Diabetic status Diabetic Non-diabetic Age -44 Race/ethnicity White Black N Am Asian Hispanic Gender Male Female Diabetic status Non-diabetic Diabetic {5.4} with an AV graft as their first access {5.41} Median urea reduction ratio (%), by access type Age -44 Race/ethnicity White Black N Am Asian Hispanic Gender Male Diabetic status Diabetic Non-diabetic Female ecent guidelines published by the NKF in their Kidney Disease Outcomes Quality Initiative concluded that clinical outcomes in hemodialysis could be improved with increased use arteriovenous (AV) fistulas. Figures provide information on access use at the initiation hemodialysis. Dialysis catheters remain the access choice among providers, and their use appears to be on the rise despite NKF recommendations. Catheter use varies little among age, gender, racial/ethnic, and diabetic groups. AV fistulas are used in 3 percent initiating dialysis (see Figure hp.8). Use is more frequent in males than in females, and 5 6 percent lower in blacks compared to other racial/ethnic groups. Approximately 5 percent more non-diabetic receive AV fistulas than diabetics. As is the case with AV fistulas, use AV grafts continues to fall (Figure 5.4). There is little difference in use between age groups, while 3.8 percent more males receive AV grafts than females. Use is higher in blacks and Hispanics, and slightly higher in diabetics than in non-diabetics. Catheter AV fistula AV graft <75 65-<7 -<65 < Urea reduction ratio (URR) is a measurement dialysis dose. NKF guidelines recommend a minimum URR 65 percent (Kt/V 1.2). In 1, 19.4 percent with catheters had URRs below the target, while percentages below target for those with AV fistulas and AV grafts were 13.3 and 15.2, respectively (Figure 5.41). Patients with dialysis catheters are nearly six time more likely to have their accesses replaced with another catheter than with an internal access, and rates sepsis in these are one and a half times higher than rates infection (Figures ). Infection rates in with catheters are as much as two and a half times higher than those in with internal accesses, while rates sepsis are times higher. Peritoneal dialysis are four to five times more prone to suffer from peritonitis than from an infection or sepsis. {Figures } prevalent hemodialysis, CPM data; includes only who are also in the USRDS database, who begin dialysis prior to October 1 the prevalent year, & whose first access is known. Year represents the prevalent year & the year the CPM data were collected. URR in Figure 5.41 represents the median value the total (up to three) URR measurements reported in the CPM data. {Figures } incident hemodialysis who are also in the CPM database, combined; includes whose first access is the access addressed in the figure. First access determined from CPM data. Events & complications identified from claims during the first year after the first service date; events identified from Part B CPT codes, & infections/sepsis from Parts A & B ICD-9-CM codes. {Figure 5.45} incident peritoneal dialysis who are also in the CPM database, combined. Events & complications identified from claims during the first year after the first service date; events identified from Part B CPT codes, & infections/sepsis from Parts A & B ICD-9-CM codes. Vascular access events 114 Clinical careh

15 {5.42} Rate catheter events & complications in the first year dialysis in whose first access is a catheter {5.43} Rate AV fistula events & complications in the first year dialysis in whose first access is an AV fistula 3 All: Events Complications 1.5 All: Events Complications Rate per patient year 3 Diabetic Non-diabetic Rate per patient year. 1.5 Diabetic Non-diabetic Removal Rep/cath Rep/internal Infection Sepsis. Rep/fistula Rep/catheter Revision Infection Sepsis Declot Angioplasty {5.44} Rate AV graft events & complications in the first year dialysis in whose first access is an AV graft {5.45} Rate PD catheter events & complications in the first year dialysis in whose first access is a peritoneal dialysis catheter 1.5 All: events Complications 1.5 All: Events Complications Rate per patient year Non-diabetic Rate per patient year. 1.5 Diabetic Non-diabetic Replace/ graft Replace/ catheter Revision Infection/ Sepsis int. device Declot Angioplasty. Removal PD cath Replacement HD cath Internal HD acc. Infection/ PD cath. Infection/ peritonitis Sepsis H4 USRDS Annual Data Report 115

16 summary ChapterJo Diabetic care Prescription drug therapy for diabetic {Figures Hemoglobin A1c monitoring in prevalent ESRD is less than that the general Medicare population. Of those tested, ESRD tend to receive more measurements in a year.} {Figures ESRD are less likely than general Medicare to have lipid testing at least once per year, with 42 percent receiving no tests. At least percent transplant have at least one lipid test each year.} {Figure 5.11 Substantially more diabetic with ESRD receive insulin therapy than those without.} {Figure 5.12 Diabetic ESRD rarely receive metformin because the risk lactic acidosis, but use in the general diabetic population has increased greatly.} {Figures The use ACE inhibitors/arbs and beta blockers has grown in both ESRD and non-esrd. About 52 percent older ESRD with CVD receive beta blockers, a slightly higher percentage than in their non-esrd counterparts.} {Figure 5.19 The use anticoagulants is higher in CVD with ESRD than in those without, but has fallen since.} Prescription drug therapy for cardiovascular disease {Figure 5. Despite their high risk atherosclerotic heart disease, fewer than 45 percent diabetic dialysis with ESRD received lipid-lowering agents in 2.} {Figure 5.21 In dialysis with CHF, ACE inhibitor/arb and beta blocker use has generally increased, but digitalis preparations are used in less than 15 percent these.} {Figure 5.22 In the dialysis population with CVD, use ACE inhibitors/arbs, beta blockers and lipid-lowering agents is highest in diabetic.} {Figure 5.23 Between and 1, the use ACE inhibitors/arbs and beta blockers grew substantially in female and diabetic dialysis with hypertension.} Provider management anemia {Figure 5.27 Trends in hemoglobin levels over time show that with high or low hemoglobin levels converge towards the K/DOQI guidelines. Certain providers appear to target higher levels. Although EPO doses appear to fall by about percent for with hemoglobins above 12 gm/dl, continued dose reductions do not appear to be present over a six-month period. Iron appears to be consistently given to 85 percent. } Clinical characteristics with hemoglobins less than 11 g/dl {Figure 5.34 Compared to with higher hemoglobin levels, those with low hemoglobins have greater dialysis catheter use, significantly lower use arteriovenous fistulas, and comparable use arteriovenous grafts } {Figure 5.36 Patients with hemoglobin levels less than 11 g/dl have higher rates infections and thrombotic complications including sepsis and clotting in the six months after initiation therapy.} Vascular access events {Figure 5.39 The use dialysis catheters in incident has changed little, with catheters used as the first access in approximately 7 percent new.} {Figures Patients whose first access is a dialysis catheter are almost 2.5 times more likely to have it replaced with a catheter than with an internal access. Infections and sepsis are common in these, much more so than in those who have an arteriovenous fistula as their first access.} Maps: National means & patient populations Figure number ESRD GM ESRD GM ESRD GM Overall value for all pts Total 17,994 82,27 17,994 82,27 17,994 82,27 3,278 3,278 Overall value for pts mapped Missing HSA/state: pts dropped 245 2, , , Figure number < < <11 Overall value for all pts ,347 15,127 21, Total 3,278 51,511 37,634 26,74 51,511 37,634 26,74 Overall value for pts mapped ,351 15,124 21, Missing HSA/state: pts dropped 46 1,512 1,18 8 1,512 1,18 8 Chapter summary 116 Clinical careh