ESRD provide

rs

192 Introduction 19

4 Dif

erences

in pro

vider

growth

{dia

lysis

trea

tmen

ts} {u

nit lo

catio

n &

gro

wth

} {af

filia

tion

& p

rofit

stat

us}

196 Patient characteris

tics b

y unit aY

liatio

n {patie

nt dem

ogra

phics &

clin

ical

par

amet

ers }

{pre

vale

nt p

atie

nts,

diag

nose

s, &

mod

alit

y}

198 Provider compliance with K/DOQI g

uidelines

{anem

ia tr

eatm

ent}

{vasc

ular a

cces

s} {s

erum

alb

umin

, cre

atin

ine,

& ch

oles

tero

l} {p

atie

nt &

staf

f vac

cina

tion

s}

200 Reuse practices {trends} {re

use by ESRD network

} {ger

micides

& ESRD net

works}

{clin

ical p

aram

eter

s by r

euse

germ

icid

e }

20

2 Su

mm

ary

CHAPTER

Rossettih

Does the road wind up-hill all the way?Yes, to the very end.Will the day’s journey take the whole long day?From morn to night, my friend....Shall I meet other wayfarers at night?Those who have gone before.Then must I knock, or call when just in sight?They will not keep you standing at that door.

Shall I find comfort, travel-sore and weak?Of labour you shall Ind the sum.Will there be beds for me and all who seek?Yea, beds for all who come.

“Uphill”

Christina

192 ESRD providersh

xpansion of the ESRD program, particularly of the dialysis population,

has required providers to keep pace by increasing the number of units in

which patients receive services. Reimbursement for dialysis services, how-

ever, is capitated on a per treatment basis, with provisions for staff and supplies, overhead,

and ancillary support services. This capitation places continued pressure on providers to

reduce costs, particularly since rates of payment have changed little since their implementa-

tion in 1982. The consolidation of dialysis providers into large chains, and the transition of

these providers from non-profit to for-profit status, are both consistent with the need for

increased capital to build dialysis units; it has become increasingly difficult for non-profit

and independent units to thrive.

In Chapter Twelve we provide detailed information on the costs associated with dialysis

services, while in this chapter we present data on recent trends in the dialysis infrastructure

and on how these changes relate to provider profit status and chain status, growth in the

number of dialysis treatments, and differences in the geographic distribution of providers

across ESRD networks, states, and Health Service Areas.

Starting with 2002, facility profit status is now determined from the CMS Dialysis Facil-

ity Compare data, as this information is no longer collected in the CMS Annual Facility

Survey. Overall, the number of total providers increased 6 percent, within the 4–8 percent

annual increase noted since 1996. The total number of hemodialysis treatments grew 27

percent between 1998 and 2002, while the number of peritoneal dialysis treatments fell 21

percent over the same period.

Dialysis chains—defined by the USRDS as corporations owning 20 or more freestanding

units that are located in more than one state—continue to grow, and own the majority of

units in the southern and southeastern states. Non-chain units are more widely dispersed

across the country, particularly in the Upper Midwest and the northeastern states. By ESRD

network, there are dramatic differences in unit ownership; in Network 2 (New York), for

example, only 20 percent of units are owned by chains, compared to 82 percent in Network

8 (Alabama, Mississippi, and Tennessee). Most of the recent growth in the number of both

units and patients has occurred in the large chains—Fresenius, Gambro, DaVita, Renal Care

Group, Dialysis Clinics Inc., and National Nephrology Associates.

Unit density varies considerably, from a mean of 0.83 units per 100,000 people in Califor-

nia and rural areas to 4.2 units per 100,000 people in the southern and central states.

Introducion

90 91 92 93 94 95 96 97 98 99 00 01 02

stin

ufo

reb

muN

0

500

1,000

1,500

2,000

2,500

3,000

3,500Freestanding for-profit

Hospital facility Transplant & dialysis centerTransplant center

Freestanding non-profit Hospital center

Chapt

er hi

ghlig

hts

193H2004 USRDS Annual Data Report

{11.1} Counts of dialysis & transplant units, by CMS certification typedata are obtained from CMS’s annual End-Stage Renal DiseaseFacility Survey, CMS Independent Renal Facility CostReports, & the CMS “Dialysis Facility Compare”website. The leveling out of the number offreestanding, for-profit units in 2002 isdue to the reclassification of unitsin the survey.

The care delivered by providers has been a focus of the

USRDS for the last three years. Clinical results by provider

vary widely. DaVita, for instance, appears to have the great-

est percent of patients with hemoglobins of 12 g/dl and higher,

and epoetin doses that are among the highest as well. Levels

of dialysis therapy, in contrast, as measured by urea reduc-

tion ratio (URR), show that, 86 percent of hemodialysis pa-

tients across the country meet the K/DOQI URR target of

≥65 percent. The distribution of patients by vascular access

is similar across chains, with patients treated by National

Nephrology Associates most likely to have a synthetic graft.

And patients receiving therapy in units owned by Renal Care

Group are considerably more likely to receive a pneumococ-

cal pneumonia vaccination than those in other units.

The CDC has monitored hemodialyzer reuse for many

years through its annual National Surveillance of Dialysis-

Associated Diseases in the United States. Survey data from

2002 were not available in time for this ADR; we present,

then, data from the 2001 survey, which show a decline in the

percent of units reusing dialyzers. Reuse practices vary widely

nationwide. The most striking geographic differences are re-

lated to the practice of using bleach as a cleaning agent. In

Louisiana, for example, two-thirds of dialysis units use bleach

in their reuse process, while in the Great Plains states bleach

is used in less than 3 percent of all units. Interestingly, units

that do not reuse their dialyzers appear to have the lowest

proportion of patients with hemoglobin levels of 12 g/dl and

above; this may be related to the patient population in these

units, and merits further investigation. Generally, there is

little difference by reuse practice or germicide use when look-

ing at hemoglobin levels, epoetin doses, serum albumin lev-

els, and delivered dialysis therapy.

Provider-level information presented here can be exam-

ined in conjunction with data in Chapter Six on hospitaliza-

tion and mortality ratios (pages 134–137), in Chapter Seven

on transplantation ratios (page 145), and in Chapter Twelve

on PMPM costs for clinical services (page 211). After ad-

dressing basic demographic differences, morbidity and mor-

tality ratios are similar across providers. And there are clear

differences in resource utilization between providers, yet out-

comes on a gross scale are similar.

With attention being paid to the K/DOQI guidelines, many

basic elements of dialysis care have improved in recent years.

To improve patient outcomes even further, new and more

detailed assessments may be required. The impact of

diabetes, insulin resistance, and the heavy inflamma-

tory load carried by both the chronic kidney dis-

ease and dialysis populations, for instance, each

merit investigation. And while effective inter-

ventions are yet to be defined, observa-

tional studies support hypotheses that

inflammation may be an impor-

tant area to consider for clini-

cal trials. These efforts

would include current

trials addressing the

efficacy of daily

dialysis.

{11.2} In most

networks the

number of perito-

neal dialysis treat-

ments has fallen or re-

mained relatively stable,

though growth has occurred in

Networks 9, 10, and 18. {11.5} The

number of patients receiving treat-

ment in chain-affiliated units has in-

creased six-fold since 1993, while the number

treated in non-chain units fell after peaking in the

early 1990s, and has remained stable during the past

four years. {11.12} While the dialysis chains, particularly

Fresenius, have grown significantly, there has been little

change in the number of non-chain and hospital-based units.

{11.19} Renal Care Group appears to have the highest rates of vaccina-

tion for influenza and pneumococcal pneumonia, and Gambro among the

lowest rates. {11.21} Dialyzer reuse grew during the 1990s, but since 1997 has

fallen slightly.

194 ESRD providersh

)sn

oillim

ni(st

nemtaert

retnec-

nI

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.01998

2002

ESRD network

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

)sd

nasu

oht

ni(st

nemtaerT

0

5

10

15

20

25

30

35

40

Hemodialysis

Peritoneal dialysis

89 2 0 89 2 0 89 20 89 20 89 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 2 0 8 9 2 0 8 9 20 89 20

s tin

ufot

ne creP

0

20

40

60

80

100

Chain-affiliated

Non-chain

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Units addedUnits unchanged

Units dropped

etween 1998 and 2002, the num-

ber of in-center hemodialysis treat-

ments increased 28 percent (Figure

11.2). Growth ranged from 16 percent in

Networks 5 and 8 to nearly 37 percent in

Networks 16 and 17—the northwestern

states and northern California. Changes in

the number of peritoneal dialysis treatments

have been less consistent nationwide.

Figure 11.3 illustrates growth in the

number of chain-affiliated and non-chain

dialysis units. The Mid-Atlantic region and

East Coast have seen the greatest increase

in chain-affiliated units, while growth in

non-chain units is more widespread, with

the highest concentrations occurring in the

northern half of the country.

In approximately half of the ESRD net-

works the proportion of units that are

chain-affiliated has remained relatively

stable, while Networks 8, 9, 10, 12, and 13

saw an increase in that proportion between

1998 and 2002 (Figure 11.4). The greatest

growth—from 16 to 46 percent—occurred

in Network 16. While the proportion of

chain-affiliated units doubled in Network

2 (New York), this network continues to

have the greatest percentage of indepen-

dently owned units, at almost 80 percent.

Figure 11.5 illustrates the parallel growth

of unit and patient counts, as well as thedramatic changes occurring in the compo-

sition of the ESRD program since 1988.

The number of patients receiving treatment

in chain-affiliated units, for instance, has

increased six-fold since 1993, while the

number treated in non-chain units fell af-

ter peaking in the early 1990s, and has re-

mained stable during the past four years.

From 1998 to 2002 overall growth in unit

and patient counts was quite similar, at 21–

24 percent, but there are considerable varia-

tions across the country (Figure 11.6). In

Networks 3, 4, and 10, for example, the in-

crease in the number of units has been far

higher than that in the number of patients.

In the western states of Networks 16, 17,

and 18, in contrast, growth in the number

of patients has far outpaced that of new units.

Three-quarters of ESRD patients are

treated in freestanding, for-profit dialysis

units (Figure 11.7). Growth in the for-profit

sector generally parallels that seen with

chain affiliation (Figure 11.9). In Networks

6, 7, 13, and 14, nearly 90 percent of units

are for-profit, while in Network 2 almost

{11.2} Dialysis treatments, by ESRD network & modality

Diferences in provider growth

Chain-affiliated & non-chain units

{11.3} Unit growth between 1998 & 2000

{11.4} Unit distribution, by ESRD network

Chain-affiliated units Non-chain units

195H2004 USRDS Annual Data Report

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

0

700

1,400

2,100

2,800

0

50

100

150

200

0

700

1,400

2,100

2,800

3,500

)sd

nasu

oht

ni(st

neitapf

ore

bmu

N

0

50

100

150

200

250Profit status

Chain status

88 90 92 94 96 98 00 02

stin

ufo

reb

muN

0

1,000

2,000

3,000

4,000

0

50

100

150

200

250

300Freestanding/hospital-based

Chain: units

Chain: patients

Profit: units

Profit: patients

FS: units

FS: patients

HB: units

HB: patients

Non-profit: unitsNon-profit: patients

ESRD network

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

2002-8991,eg

nahc

tne cre P

0

10

20

30

40

50

89 20 89 20 89 20 89 2 0 8 9 2 0 89 2 0 89 20 89 20 89 20 89 20 89 20 89 2 0 89 2 0 89 2 0 89 2 0 89 2 0 89 20 8 9 20 89 20

st in u f

otne creP

0

20

40

60

80

100

89 20 89 20 8 9 2 0 8 9 2 0 89 20 8 9 20 8 9 20 8 9 20 8 9 20 8 9 20 89 2 0 89 2 0 89 2 0 89 2 0 89 2 0 89 20 8 9 2 0 89 20 89 20

s t inu f

otn ec re P

0

20

40

60

80

100

96 97 98 99 00 01 02

stneita

pfot

nec reP

0

20

40

60

80

100

2.60+ (4.21)1.89 to <2.601.47 to <1.89 No units

1.09 to <1.47below 1.09 (0.83)

Transplant & dialysis center

Hospital facility

Hospital center

Freestanding non-profit

Freestanding for-profit

Units

Patients

Non-chain:

patients

units

Non-chain:

Unknown

Non-profit

Profit

Hospital-based

Freestanding

{11.5} Unit & patient counts, by unit type {11.6} Percent change in the number of units & patients, 1998 to 2002, by ESRD network

{11.9} Distribution of for-profit & non-profit units, by ESRD network

{11.10} Distribution of freestanding & hospital-based units, by ESRD network

{11.8} Units per 100,000 population, 2002{11.7} Pt distribution, by CMS certification type

70 percent continue to be operated on a

non-profit basis.The number of units available per

100,000 population differs widely across the

country, with a five-fold difference between

upper and lower quintile averages (Figure

11.8). In the country as a whole, 81 percent

of units are freestanding (Figure 11.10). In

the southeastern states of Networks 6, 7, and

8 this number reaches more than 90 per-

cent, while Network 2 (New York) has the

largest proportion of hospital-based units, at

49 percent.

{Figures 11.2–10} data obtained from CMS’s annualEnd-Stage Renal Disease Facility Survey, CMS Inde-pendent Renal Facility Cost Reports, & the CMS “Di-alysis Facility Compare” website. {Figures 11.8} dataalso obtained from estimates of the United States 2002census, based on the 2000 census. x {Figure 11.2}Transient treatments, which account for less than 1percent of all treatments, are not included. Hemodialy-sis includes outpatient hemodialysis & hemodialysistraining treatments; peritoneal dialysis includes out-patient IPD treatments & IPD, CAPD, & CCPD train-ing treatments. {Figure 11.8} 2002, by HSA, unadjusted.

The CDC did not conduct a survey in 1998, & data for2002 were not available as this book went to press. Fig-ure 2.38, in Chapter Two, contains a map of the ESRDnetworks; a list of network contacts can be found onpage 230 of Appendix A.

196 ESRD providersh

)sraey(e

ganae

M

56

58

60

62

64

elamef

tnecreP

40

42

44

46

48

50Age Gender: Female

stneita

pfo

tnecreP

0

20

40

60

80

100

citebai

dt

necr eP

34

36

38

40

42

44

46

Race

Diabetic status: Diabetics

stneita

pfo

tnecreP

0

20

40

60

80

100

)ld/

g(ni

bol

go

meH

9.6

9.8

10.0

10.2

10.4

10.6

Primary diagnosis

Modality

stneita

pfot

necreP

0

20

40

60

80

100

stneita

pfot

necreP

0

20

40

60

80

100

stneita

pfo

tnecreP

25

30

35

40

All 1 2 3 4 5 6 NC HB U

stneita

pfo

tnecre P

45

50

55

60

65

70

Unit affiliation (see table at right for codes)

All 1 2 3 4 5 6 NC HB U

/m2 )

gk(IM

B

25

26

27

28

29

All 1 2 3 4 5 6 NC HB U

eGFR

(ml/

min

/1.7

3 m

2 )

9.0

9.5

10.0

10.5

11.0

Hispanic ethnicity

Mean hemoglobin at initiation Percent receiving EPO at initiation

Percent with albumin < test’s lower limit Mean BMI at initiation Mean eGFR at initiation

Hispanic-MexicanHispanic-otherNon-HispanicOther/unknown

Hemodialysis

Peritoneal dialysis

White

Black

Native American

Asian

Other/unknown

Diabetes

Hypertension

Glomerulonephritis

Cystic kidney

Other/unknown

{11.11} Characteristics of incident dialysis patients, by unit affiliation, 2002

Patient qaracterisics, by unit aYliation

197H2004 USRDS Annual Data Report

Unit affiliation (see table below for codes)

1 2 3 4 5 NC HB0

200

400

600

800

1,000

1,200 Number of units Number of patients (in thousands)

1 2 3 4 5 NC HB0

10

20

30

40

50

60

70

80

)sraey(e

ganae

M

48

50

52

54

56

58

60

62

selameft

necreP

40

42

44

46

48Age Gender: Female

s tnei ta

pfo

tnecreP

0

20

40

60

80

100

citebai

dtnecreP

24

28

32

36

40

44Race Diabetic status: Diabetics

Unit affiliation (see table below for codes)

All 1 2 3 4 5 6 NC HB U

stneita

pfo

tnecr eP

0

20

40

60

80

100

All 1 2 3 4 5 6 NC HB U

stneita

pfo

tnecreP

0

20

40

60

80

100Primary diagnosis Modality

Hemodialysis

Peritoneal dialysis

White

Black

Native American

Asian

Other/unknown

Diabetes

Hypertension

Glomerulonephritis

Cystic kidney

Other/unknown

1998

2002

he mean age of the incident ESRD

population is 62.3, and with the

exception of patients in facilities

owned by Dialysis Clinics Incorporated,

mean age varies little by unit affiliation (Fig-

ure 11.11). Differences in racial distribu-

tion among units are also unremarkable, as

are differences by diabetic status, although

there are fewer diabetics in hospital-based

units. The majority of patients in all units

have diabetes as their primary cause of renal

failure, and over 90 percent are on hemodi-

alysis. Mean hemoglobins at initiation av-

erage close to 10 g/dl, and on a unit level do

not appear related to whether or not pa-

tients receive EPO prior to starting therapy.

The percentage of patients with serum al-

bumins below the test’s lower limit is high-

est for hospital-based units, and mean BMIs

and eGFRs are similar for all unit types.

The number of chain-affiliated dialysis

units, and their patient populations, have

increased quite dramatically since 1998 (Fig-

ure 11.12). The number of patients treated

by Renal Care Group, for example, has

doubled, while the population in units

owned by Fresenius has grown 42 percent.

In units not affiliated with chains, in con-

trast, the number of patients has increased

only 2 percent (growth in unit counts is

slightly higher, at 7 percent), and it has fallen

more than 12 percent in units that are hos-

pital-based.

The prevalent ESRD population tends to

be younger than the incident population,

with a mean age of 56.9 across all units (Fig-

ure 11.13). The distribution of females by

affiliation is similar. Across units the per-

cent distribution of prevalent diabetics is simi-

lar, and is slightly lower than that found in

the incident population.

{Figures 11.11–12} incident ESRD patients, 2002.Facility data obtained from the CMS annual End-Stage Renal Disease Facility Survey, the CMS Inde-pendent Renal Facility Cost Reports, & the CMS“Dialysis Facility Compare” website. {Figure 11.13}December 31 point prevalent ESRD patients, 2002.Facility data obtained from the CMS annual End-Stage Renal Disease Facility Survey, the CMS Inde-pendent Renal Facility Cost Reports, & the CMS“Dialysis Facility Compare” website. x {Figure11.11} The lower limit of albumins measured by bro-mcresol purple is 3.2 g/dl, & by bromcresol green is 3.5g/dl. {Figure 11.12} Chain 6 did not exist in 1998.

{11.12} Unit & patient counts, by unit affiliation

{11.13} Characteristics of prevalent dialysis patients, by unit affiliation, 2002

All · All unitsChain 1 · FreseniusChain 2 · GambroChain 3 · DaVita

Chain 4 · Renal Care GroupChain 5 · Dialysis Clinics, Inc.Chain 6 · Nat’l Nephrology Assoc.NC · Non-chain units

HB · Hospital-based unitsU · Unknown affiliation

198 ESRD providersh

stinuf

otnecreP 0

20

40

60

80

100Hemoglobin (g/dl) EPO dose (units)

Unit affiliation (see table at right for codes)

All 1 2 3 4 5 6 NC HB U All 1 2 3 4 5 6 NC HB U0

20

40

60

80

100

<9

9-<10

10-<11

11-<12

12+

<7,000

7,000- <11,000

11,000- <16,000

16,000- <23,000

23,000+

<60

60-<65

65-<70

70-<75

75+

<22-<2.6

2.6-<3.03.0-<3.6

3.6+Median URR (%): hemodialysis Mean weekly Kt/V: peritoneal dialysis

Unit affiliation (see table at right)

All 1 2 3 4 5 6 NC HB

stneita

pfo

tnecreP

0

20

40

60

80

100

11.85+ (11.93)11.75 to <11.8511.68 to <11.75

11.59 to <11.68below 11.59 (10.46)

17,484+ (18,546) 16,593 to <17,48415,433 to <16,593

14,146 to <15,433below 14,146 (11,007)

11.85+ (12.00)11.75 to <11.8511.68 to <11.75

11.59 to <11.68below 11.59 (10.74)

17,484+ (18,564)16,593 to <17,48415,433 to <16,593

14,146 to <15,433below 14,146 (11,393)

80.5+ (85.0)76.6 to <80.574.4 to <76.6

69.3 to <74.4below 69.3 (55.2)

80.5+ (87.4)76.6 to <80.574.4 to <76.6

69.3 to <74.4below 69.3 (64.9)Insuff. data

Cuffed catheter Non-cuffed catheter AV fistula AV graft

cross all unit affiliations, mean

hemoglobin levels are at or above

11 g/dl—the K/DOQI target—in

77–89 percent of prevalent dialysis patients

(Figure 11.14). In units owned by DaVita

and Renal Care Group, 44–51 percent of

patients have a mean hemoglobin of 12 g/dl

or greater.

The percent of dialysis patients receiv-

ing mean weekly EPO doses of 16,000 units

and above ranges from 34 in units owned

by Dialysis Clinics, Inc. to 43 in those owned

by DaVita.

K/DOQI also set a target urea reduction

ratio (URR) guideline of ≥ 65 percent, which

in the CPM dataset is met by 86 percent of

hemodialysis patients—from 84 percent of

patients in non-chain and hospital-based

units to 91 percent of those treated in units

owned by Gambro. Sixty percent of perito-

neal dialysis patients in the CPM dataset

have a mean weekly Kt/V of 2.6 or above.

Chain-affiliated units with the highest

mean hemoglobin levels are more highly

concentrated west of the Mississippi River;

non-chain units with the lowest hemoglo-

bins are located primarily in the southeast

and southwestern states (Figure 11.15).

Variations in mean EPO dose track accord-

ing to hemoglobin levels, demonstrating a

general tendency within providers to main-tain established hemoglobin targets.

The percent of patients with internal

accesses is slightly higher in chain-affiliated

units across the country (Figure 11.16).

K/DOQI guidelines recommend in-

creased use of arteriovenous fistulas. In

2001, 31 percent of hemodialysis patients

had this type of vascular access, and use var-

ied little by unit affiliation (Figure 11.17).

Overall, less than a third of incident di-

alysis patients initiate therapy with a serum

albumin above the test’s lower limit (Figure

{11.14} Anemia treatment & dialysis adequacy, by unit affiliation, 2002

Provider compliance with K/DOQI guidelines

{11.15} Geographic variations in anemia treatment, by unit affiliation, 2002

{11.16} Geographic variations in the % of pts with an internal access, by unit affiliation, 2002 CPM data

{11.17} Vascular access use, by unit aff., 2001

Mean hemoglobin: chain Non-chain

Mean EPO dose: chain Non-chain

Chain Non-chain

199H2004 USRDS Annual Data Report

stneita

pfo

tnecreP

0

20

40

60

80

100

Unit affiliation (see table below for codes)

All 1 2 3 4 5 6 NC HB U All 1 2 3 4 5 6 NC HB UAll 1 2 3 4 5 6 NC HB U

Albumin > test’s lower limit Serum creatinine <10 mg/dl Hemoglobin 11 g/dl

All 1 2 3 4 5 6 NC HB

etarn

oitaniccav

hcaeg

ni tro

pe rsti

nu f

o%

0

20

40

60

80

100

Unit affiliation (see table below for codes)

All 1 2 3 4 5 6 NC HB

Influenza Pneumococcal pneumonia

Unit affiliation (see table below)

All 1 2 3 4 5 6 NC HB

detaniccav

tnecreP

80

82

84

86

88

90

92

94

<25%

25-49%

50-74%

75-100%

Unknown

Not offered

>_

{11.18} Nutritional & hemopoietic parameters in incident dialysis patients, by unit affiliation, 2002

{11.19} Dialysis patient vaccinations, by unit affiliation, 2001 {11.20} Hepatitis B: staff vaccination rates, 2001

All · All unitsChain 1 · FreseniusChain 2 · GambroChain 3 · DaVitaChain 4 · Renal Care GroupChain 5 · Dialysis Clinics, Inc.Chain 6 · Nat’l Nephrology Assoc.NC · Non-chain unitsHB · Hospital-based unitsU · Unknown affiliation

11.18), and over 80 percent be-

gin treatment with a serum

creatinine of less than 10 mg/

dl. Slightly more than one-

quarter of new patients begin

dialysis with a hemoglobin at

or above the recommended

K/DOQI guideline of 11 g/dl.

In 2001, nearly half of the

dialysis units across the country reported giving influenza vaccina-

tions to 75–100 percent of their patients (Figure 11.19). Sixty-two

percent of units owned by Renal Treatment Centers achieved this

rate, compared to fewer than one-third of those owned by Gambro.

Vaccinations for pneumococcal pneumonia are not offered in

41 percent of units nationwide. Units owned by Renal Treatment

Centers are again most likely to vaccinate their patients, with al-

most one-third of these units reporting a pneumococcal pneumo-

nia vaccination rate of 75–100 percent in 2001 (Figure 11.19).

Hepatitis B vaccination rates for dialysis unit staff range from

85 to 92 percent, with an overall rate of 87 percent nationwide

(Figure 11.20).

{Figure 11.14} period prevalent dialysis patients, 2002. Hemoglobin graph in-cludes only patients treated with EPO, & the mean hemoglobin represents the av-erage hemoglobin value for the year across all patients. EPO dose adjusted forinpatient days. URR & Kt/V data obtained from 2002 CPM data, & include onlypatients who are in both the USRDS & CPM databases. {Figure 11.15} periodprevalent dialysis patients, 2002. Hemoglobin maps include only patients treatedwith EPO, & the mean hemoglobin represents the average hemoglobin value forthe year across all patients. EPO dose adjusted for inpatient days. {Figure 11.16}prevalent hemodialysis patients from the 2002 CPM data who are also in theUSRDS database; current access determined from CPM data. {Figure 11.17}prevalent hemodialysis patients. {Figure 11.18} incident dialysis patients, 2002.{Figures 11.17 & 11.19–20} data obtained from the CDC’s National Surveillanceof Dialysis-Associated Diseases in the United States. x {Figure 11.18} Thelower limit of albumins measured by bromcresol purple is 3.2 g/dl, & by bromcresolgreen is 3.5 g/dl.

200 ESRD providersh

90 91 92 93 94 95 96 97 99 00 01

stin

ufot

necreP

0

20

40

60

80

100

90 91 92 93 94 95 96 97 99 00 01

stin

ufo

tnecre P

0

20

40

60

80

100

ESRD network

17 18 9 15 13 7 11 6 14 All 16 10 8 5 3 12 1 4 2

esuer

taht

stin

ufo

tne cre P

0

20

40

60

80

100

Reuse No reuse

Reuse, chain

No reuse, chain

Reuse, non-chain

No reuse, non-chain

Units addedUnits unchanged

Units dropped

fter peaking at 82 percent in 1997, the percent of units

that reuse their dialyzers has dropped slightly, to 76 per-

cent in 2001 (Figure 11.21). With Fresenius’ recent deci-

sion to discontinue reuse in its dialysis units, this decrease is likely

to become more significant.

Changes in reuse practices are occurring primarily in chain-

affiliated units (Figure 11.22). Since 1990, the proportion of non-

chain units reusing dialyzers has fluctuated only between 65–70

percent, while in chain-owned units the proportion fell from 96–

98 percent in the

early 1990s to 82

percent in 2001.

Only 54 percent

of Network 2 units

reuse dialyzers (Fig-

ure 11.23); this net-

work also contains

a large proportion of

non-profit and hos-

pital-based units. In

Networks 17, 18,

and 9, in contrast,

88–91 percent of

{11.21} Reuse practices, overall

Reuse pracices

{11.22} Reuse practices, by chain status {11.23} Reuse practices, by ESRD network, 2001

{11.24} Changes in the location of reuse & non-reuse units between 1997 & 2001

units practice reuse.

The map of units that reuse hemodialyzers parallels that of

chain-affiliated and non-chain units. There is a high concentra-

tion of reuse units in the eastern half of the country, with densities

highest in the Mid-Atlantic and East Coast states (Figure 11.24).

The map of non-reuse units shows no discernable pattern other

than higher concentrations in the eastern United States, an area

that contains the majority of dialysis units.

Due to its lack of carcinogenic properties and its efficacy as a

disinfectant, the use of peracetic acid in the reuse process has grown

steadily, from 49 percent in 1990 to 62 percent in 2002 (Figure

11.25). Rates of peracetic acid use do, however, vary quite widely

across the country (Figure 11.26). In Networks 12, 16, and 11, the

germicide is used in 78–80 percent of units; in Networks 3, 13, and

14, however, 41–48 percent of units continue to use formalde-

hyde. The wide use of peracetic acid is reflected geographically in

Figure 11.27. Formaldehyde use is concentrated in units located in

the eastern third of the country, and should decrease even more

with Fresenius’ decision to discontinue reuse.

The percentage of units using bleach is highest in the southern

and eastern states; this is most likely a direct correlation with the

location of units using formaldehyde as a disinfectant (Figure 11.28).

Differences in hemoglobin levels, EPO use, serum albumin

Reuse units Non-reuse units

201H2004 USRDS Annual Data Report

0

20

40

60

80

100 Hemoglobin (g/dl) EPO dose (units)

Germicide (see table at right for codes)

PA F G B B+G NR

stneita

pfot

n ecre P

0

20

40

60

80

100

PA F G B B+G NR

Serum albumin (g/dl) Urea reduction ratio (%)

ESRD network

12 16 11 18 15 5 6 9 17 2 All 4 1 7 3 13 10 8 14

stneita

pfo

tn ecreP

0

20

40

60

80

100

90 91 92 93 94 95 96 97 99 00 01

stin

ufo

tnecreP

0

20

40

60

80

100

38.8+ (61.1)29.6 to <38.818.8 to <29.6

2.9 to <18.8below 2.9 (0.0)

Formaldehyde

Glutaraldehyde

Peracetic acid

Other (includes heat & Amuchina)

Peracetic acid Formaldehyde Glutaraldehyde Other

<99-<10

10-<1111-<1212+

<33-<44-<55+

<60

60-<65

65-<70 70-<75 75+

<7,0007,000-<11,000

11,000-<16,000

16,000-<23,000

23,000+

Other

Peracetic acidGlutaraldehyde

Formaldehyde

{11.25} Germicide use, overall {11.26} Germicide use, by ESRD network, 2001

{11.27} Location of units, by germicide type, 2001

{11.29} Clinical parameters, by germicide type, 2001

{11.28} Percent of units using bleach, 2001

PA · Peracetic acidF · FormaldehydeG · GlutaraldehydeB · BleachB + G · Bleach + germicideNR · No reuse

levels, and dialysis adequacy are unremark-

able when based on germicide type (Figure

11.29).

{Figures 11.21–28} data obtained from the CDC’sNational Surveillance of Dialysis-Associated Diseasesin the United States, 2001. Units in Figures 11.24 &11.27 are mapped by zip code. {Figure 11.29} preva-lent hemodialysis patients treated in a known pro-vider, 2001. Data on hemoglobin & EPO dose includeonly EPO-treated patients. Albumin & URR data ob-tained from CPM data, include only patients who arein both the CPM & USRDS database, & reflect themedian value for each patient. EPO doses adjustedfor IP days. x {Figure 11.25} “Other” includes heatand Amuchina.

The CDC did not conduct a survey in 1998, & data arenot available for 2002. Figure 2.38, in Chapter Two,contains a map of the ESRD networks; a list of networkcontacts can be found on page 230 of the Appendix.

Chap

terJo

summary

Chapter summary202 ESRD providersh

Provider compliance with K/DOQI guidelines{Figure 11.14 In the prevalent population, average hemoglobin levels vary by chain, with units owned by DaVita and Renal Care Grouphaving the greatest percentage of patients with hemoglobins of 12 g/dl or higher for a year. EPO doses are highest in units owned by DaVita.The percent of dialysis patients receiving mean weekly EPO doses of 16,000 units and above ranges from 34 in units owned by DialysisClinics, Inc. to 43 in units owned by DaVita.} {Figure 11.19 Patient vaccination rates differ across the dialysis providers. Renal Care Groupappears to have the highest rates of vaccination for influenza and pneumococcal pneumonia, and Gambro among the lowest rates.}

Reuse practices{Figure 11.21 Dialyzer reuse grew during the 1990s, but since 1997 has fallen slightly.} {Figure11.22 Since 1990, the proportion of non-chain units reusing dialyzers has fluctuated onlybetween 65–70 percent, while in chain-owned units the proportion fell from 96–98 percent inthe early 1990s to 82 percent in 2001.} {Figure 11.25 The use of peracetic acid in the reuseprocess has grown steadily since 1990, from 49 to 62 percent.} {Figure 11.29 Hemoglobinlevels and erythopoietin doses vary slightly in relation to the type of germicide used by a dialysisunit, while serum albumin levels and urea reduction ratios are relatively consistent. Since reusepractices and germicide use are highly dependent on providers, these results need to be factoredinto analyses of provider-specific outcomes.}

Maps: National means & patient populationsFigure number 11.15 11.15 11.15 11.15 11.16 11.16

Hgb ch Hgb nch EPO ch EPO nch chain n-chainOverall value for all pts 11.7 11.7 16,894 15,770 75.7 74.0Total patients 130,357 68,183 130,357 68,183 4,677 2,657Overall value for pts mapped 11.7 11.7 16,865 15,762 75.8 74.1Missing HSA/state: pts dropped 2,798 1,387 2,798 1,387 102 1,040

Patient characteristics by unit afiliation{Figure 11.11 The distribution of incident patients by age is generally comparable across providers. There isconsiderable variation by provider in the pre-dialysis treatment of anemia, with the percent of patientsreceiving EPO prior to initiation ranging from 29 to 38. The percentage of patients with serum albuminsbelow the test’s lower limit is highest in hospital-based units, and mean BMIs and eGFRs are similar for allunit types.} {Figure 11.12 While the dialysis chains, particularly Fresenius, have grown significantly, there hasbeen little change in the number of non-chain and hospital-based units.}

Diferences in provider growth{Figure 11.2 Since 1998 the greatest growth in the number of in-center hemodialysis treatments has occurred in Network 17,and the smallest growth in Network 5. In most networks the number of peritoneal dialysis treatments has fallen or remainedrelatively stable, though significant growth has occurred in Networks 9, 10, and 18.} {Figure 11.5 The number of patientsreceiving treatment in chain-affiliated units has increased six-fold since 1993, while the number treated in non-chain unitsfell after peaking in the early 1990s, and has remained stable during the past four years.}

Introduction{Figure 11.1 The number of freestanding units continues to grow, and has more than tripledover the last twelve years.}