1 · CHRONIC KIDNEY DISEASE - USRDS Home … 3 48 CHRONIC KIDNEY DISEASE 1 Introduction Chronic...

Click here to load reader

  • date post

    25-May-2019
  • Category

    Documents

  • view

    230
  • download

    0

Embed Size (px)

Transcript of 1 · CHRONIC KIDNEY DISEASE - USRDS Home … 3 48 CHRONIC KIDNEY DISEASE 1 Introduction Chronic...

1chronic kidney disease

The wrong of unshapely things is a wrong too great to be told;

I hunger to build them anew and sit on a green knoll apart,

With the earth and the sky and the water, remade, like a casket of gold

For my dreams of your image that blossoms a rose in the deeps of my heart.

william butler yeatsthe lover tells of the rose in his heart

CCHRONIC KIDNEY DISEASE3

48

1 Introduction

Chronic kidney disease (CKD), the precursor to ESRD, received little attention prior to the 2002 publication of the National Kidney Foundations Chronic Kidney Disease Guidelines. This docu-ment created a CKD classification system, using five stages defined by increasing evidence of kid-ney damage, as shown by microalbuminuria and estimated glomerular filtration rates (eGFRs).

Applying this new classification system to NHANES III data, Coresh and Levey et al. esti-mated that eight million people in the U.S. had an eGFR less than 60 ml/min/.73 m2, and another 2 million had evidence of microalbuminuria.

The USRDS uses diagnosis codes from hos-pital and outpatient encounters to determine the size of the CKD populationa method which utilizes only those codes identified by providers. In the Prcis we define the CKD population over a two-year period, looking at those who survive until the end of the year, and considering them point prevalent at the start of the next year. To approximate a period prevalent CKD population, we also include new patients identified with CKD in the next year; expenditures are determined in the second year.

In this chapter, however, we use a slightly dif-ferent definition for CKD patients, requiring sur-vival for the entire second year in order to give each patient equal opportunity for access to care. This definition may exclude the sickest CKD pa-tients, since they may have early mortality.

Figure . shows that in almost two-thirds of Medicare patients with CKDidentified by di-

agnosis codes over the two-year periodCKD is accompanied by diabetes, CHF, or both diseases combined. In employed CKD patients under 65, a similar proportion has diabetes, but far fewer patients have CHF alone or with diabetes. Over the past decade the CKD population has become more complex, with rates of diabetes alone and in combination with CHF growing 38 and 26 per-cent, respectively, in the Medicare population, while in the employed population rates of CKD accompanied by diabetes, CHF, or both diseases have grown 93 percent.

Figure .4 illustrates these changes in further detail. The percent of CKD patients with diabetes increased in the Medicare population from 6.7 in 992993 to 23.7 in 20022003, and in the EGHP population from 23.8 in 9992000 to 26.6 in 20022003. While the percent whose CKD is ac-companied by heart failure has remained steady or fallen, the burden of CKD with both diabetes and heart failure has grown.

Few patients who are at risk for CKD receive optimal assessment. Testing for microalbumin-uria or proteinuria is done in only 20 percent of diabetic patients, and testing rates are even lower in diabetic patients who also have CHF.

Preventive care in CKD patients is similarly infrequent. Measurement of at least two glyco-sylated hemoglobins each year, as recommended by the American Diabetes Association, occurs in only 56 percent of Medicare patients with both diabetes and CKD, and 35 percent of the EGHP population with the same diagnoses.

50 prevalence of CKD

52 assessment of populations at risk for CKD

54 preventive

healthcare monitoring in CKD patients

56 prescription drug

therapy in CKD patients

geographic variations in the prevalence of CKD interactions of diabetes, CHF, & CKD

creatinine proteinuria renal ultrasound calcium & phosphorus parathyroid hormone glycosylated hemoglobin & lipid testing use of diabetic testing supplies vaccinations ACE-Is/ARBs beta blockers calcium channel blockers diuretics lipid-lowering agents diabetes drugs

58 racial disparities in the assessment

of CKD

60 hospitalization rates in CKD & non-

CKD patients

62 acute kidney failure:

hospitalization & outcomes

64 summary

assessments for renal function drug therapy geographic variations in assessment rates all-cause & cause specific hospitalization rates geographic variations in hospitalization rates

probability of hospitalization for AKF ESRD & mortality rates following hospitalization for AKF

Figure 1.6 The overall probability of microalbuminuria or proteinuria testing reaches only

0.04 in the Medicare population, and 0.01 for EGHP patients. Figure 1.14 In the Medicare

CKD population, the probability of receiving at least two HbA1c tests within a year is 0.56,

compared to 0.36 for EGHP patients. Figure 1.19 The use of ACE-Is and ARBs for renal pro-

tective treatment has almost doubled in the CKD population. Figure 1.32 In the Medicare

population, the overall hospitalization rate for CKD patients reached 1,068 per 1,000 pa-

tient years in 2003nearly three times greater than that in patients without CKD. Figure

1.39 For patients with a prior diagnosis of CKD, the rate of death in the first three months

after hospitalization is 2.74 times greater than the rate of an ESRD diagnosis.

Medicare

CKD only: 423,200 36%

CKD+CHF: 200,900

17%

CKD+DM: 299,600

25%

EGHP (age

25.2 + (29.4)22.3 to

Medicare, 19921993

CKD only: 257,98040.7%

CKD+CHF: 149,900

23.6%

CKD+DM+CHF:

119,960 18.9%

CKD+DM: 106,000

16.7%

CKD total: 633,840(2.34% of total Medicare)

Medicare, 19971998

CKD only: 307,14035.4%

CKD+CHF: 192,480

22.2%

CKD+DM+CHF:

193,22022.2%

CKD+DM: 175,840

20.2%

CKD total: 868,680(3.38% of total Medicare)

EGHP, 19992000

CKD only: 4,23363.6%

CKD+CHF: 3525.3%

CKD+DM+CHF:

4807.2%

CKD+DM: 1,58623.8%

CKD total: 6,651(0.47% of total EGHP)

EGHP, 20022003

CKD only: 9,78959.4%

CKD+CHF: 9115.5%

CKD+DM+CHF: 1,4008.5%

CKD+DM: 4,38226.6%

CKD total: 16,482(0.66% of total EGHP)

Medicare, 20022003

CKD only: 520,52033.4%

CKD+CHF: 289,380

18.6%

CKD+DM+CHF:

380,30024.4%

CKD+DM: 368,780

23.7%

CKD total: 1,558,980(5.62% of total Medicare)

eg

nah c t

ne creP

0

50

100

150

200

250

300

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

Percent change in population size

Medicare: 1992-93 to 2002-03

EGHP: 1999-2000 to 2002-03

51

2005 Annual Data Report 31CHRONIC KIDNEY DISEASE

1.4 Medicare CKD patients, & EGHP CKD patients younger than 65Trends in the interactions of diabetes,

congestive heart failure, & CKD

{Figure .3} per ,000 patients, by HSA, unadjusted. General Medicare patients continuously enrolled in Medicare Parts A & B during two consecutive years; patients enrolled in an HMO or diagnosed with ESRD during the period are excluded. {Figure .4} Medicare: general Medicare CKD patients continuously enrolled in Medicare Parts A & B for two consecutive years (numbers estimated from 5 percent sample); patients enrolled in an HMO or diagnosed with ESRD during the period are excluded. EGHP: CKD patients younger than 65 & continu-ously enrolled in a fee-for-service-plan for two consecutive years.

1_4.zip

All NCKD Diabetes CHF DM+CHF Other

All NCKD Diabetes CHF DM+CHF Other

All NCKD Diabetes CHF DM+CHF Other

ytiliba

borP

0.0

0.1

0.2

0.3

Medicare

EGHP

Medicare

EGHP

Medicare

EGHP

ytiliba

borP

0.0

0.1

0.2

0.3

ytiliba

bor P

0.00

0.02

0.04

0.06

0.08

0.10

0.286 + (0.404)0.187 to

yti liba

borP

0.00

0.02

0.04

0.06

0.08

0.10

0.12Medicare

EGHP

Medicare

EGHP

Medicare

EGHP

ytiliba

bor P

0.00

0.01

0.02

ytiliba

bo rP

0.0

0.1

0.2

0.3

0.4

All NCKD Diabetes CHF DM+CHF Other

All NCKD Diabetes CHF DM+CHF Other

All NCKD Diabetes CHF DM+CHF Other0.094 + (0.121)0.073 to

All CKD CKD+CHF CKD+other

ytiliba

borP

0.0

0.1

0.2

0.3

Medicare

EGHP

All CKD CKD+CHF CKD+other

ytiliba

borP

0.0

0.1

0.2

0.3

0.4

0.5

0.6

All CKD CKD+CHF CKD+other

ytiliba

borP

0.0

0.1

0.2

0.3

0.4

0.5

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

ytiliba

borP

0.0

0.1

0.2

0.3

0.4Medicare

EGHP

ytiliba

bo rP

0.00

0.02

0.04

0.06

0.08

0.10

0.285 + (0.318)0.257 to

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

All CKDCKD+DM

CKD+CHFCKD+DM+CHF

CKD other

ytiliba

borP

0.0

0.1

0.2

0.3

0.4

0.5

0.6

ytiliba

bo r

P

0.00

0.05

0.10

0.15

ytiliba

borP

0.0

0.2

0.4

0.6

0.8

0.57 + (0.61)0.55 to

All CKD CKD+DM CKD+HTN CKD+CHF CKD other

stneita

p fo t

n ecreP

0

20

40

60

80NHANES III

NHANES 99-02

NHANES III

NHANES 99-02

*

All CKD CKD+DM CKD+HTN CKD+CHF CKD other

stneita

p fo t

necreP

0

20

40

All CKD CKD+DM CKD+HTN CKD+other

stneita

p fo t

n ecreP

0

5

10

15

20

25NHANES III

NHANES 99-02

Dihydropyridine calcium channel blocker Non-dihydrorpyridine calcium channel blocker

All CKD CKD+DM CKD+HTN CKD+other

**

stneita

p fo t

nec reP

0

10

20

30

40

50

60 Loop diuretic Thiazide-like diuretic

*

All CKDCKD+DM

CKD+HTN CKD otherCKD+CHF

All CKDCKD+DM

CKD+HTN CKD otherCKD+CHF

NHANES III

NHANES 99-02

CHRONIC KIDNEY DISEASE3

56

1 Prescription drug therapy in CKD patientsNHANES III & NHANES 19992002 participants, age 60 & older

CKD patients receiving ACE-Is/ARBs 1.19

NHANES III & NHANES 19992002 participants, age 60 & older

CKD patients receiving beta blockers1.20

NHANES III & NHANES 19992002 participants, age 60 & older

CKD patients receiving calcium channel blockers1.21

NHANES III & NHANES 19992002 participants, age 60 & older

CKD patients receiving diuretics1.22

iven the high percentage of CKD patients with diabetes and/or congestive heart failure,

there are several preventive and mainte-nance therapies that one would expect to see routinely prescribed. We use the NHANES III (988994) and NHANES 9992002 datasets to evaluate prescription drug use in CKD patients. These data rep-resent a cross-sectional evaluation of med-ication use in the past month, in contrast to the evaluation we performed on the Med-stat database for the 2004 ADR, in which cumulative use of medications in CKD Stages 35 over a one-year period of time was evaluated. It is important to note that CKD Stage 5 patients are excluded from these analyses because it is not possible to evaluate dialysis status. These data thus represent CKD Stages 4, with Stages and 2 determined by a calculated GFR plus a positive urine microalbumin. In addi-tion, because inclusion of all age groups leads to erratic results and there are too few patients younger than 60 to evaluate, this spread represents patients age 60 and older with Stage 4 CKD.

Overall, older CKD patients in 9992002 were treated more aggressively with medications than in previous years. The use of angiotensin-converting enzyme inhibitors (ACE-Is) and angiotensin recep-tor blockers (ARBs) has more than dou-bled in CKD patients with diabetes, from 22 to 46 percent, and has almost doubled in CKD patients with hypertension (Figure .9). Further growth in use of these agents should be fueled by National Kidney Foun-dation (NKF) CKD practice guidelines, released in 2002, which support ACE-I and ARB use in CKD patients with diabetes, by 2004 guidelines advocating ACE-I and ARB use in CKD patients with hypertension, and by congestive heart failure guidelines pub-lished in 200 advocating ACE-Is and ARBs in CHF patients.

CHF and hypertension practice guide-lines advocate the use of beta blockers in the treatment of these diseases. Use of these agents has increased from 9 to 26 percent and 5 to 28 percent in older CKD patients with hypertension and congestive heart failure, respectively (Figure .20). Interest-ingly, beta blocker use has also increased tremendously in CKD patients with diabe-tes. Increased use overall, in hypertensive

G

1_19.zip1_20.zip1_21.zip1_22.zip

All CKD CKD+DM CKD+CVD CKD otherAll CKD CKD+DM CKD+CVD CKD other

stneita

p fo t

ne creP

0

10

20

30

40

NHANES III

NHANES 99-02

Statins Non-statins

* **

* ** *

*

Insulin TZDs Secretagogues MetforminInsulin TZDs Secretagogues Metformin

stneita

p fo t

necr eP

0

10

20

30

40

50NHANES III

NHANES 99-02

CKD Non-CKD

*

57

2005 Annual Data Report 31CHRONIC KIDNEY DISEASE

NHANES III & NHANES 19992002 participants, age 60 & older

CKD patients receiving lipid-lowering agents

patients, and in patients with diabetes may reflect lower blood pressure goals and the fact that beta blockers and diuretics were advocated by Joint National Committee on Prevention, Detection, Evaluation and Treatment 6 (JNC 6) guidelines in 997 as first-line therapy to reach new blood pres-sure goals.

Use of dihydropyridine calcium-chan-nel blockers (CCBsamlodipine, felo-dipine, isradipine, nicardipine, nifedipine, nimodipine, nisoldipine) has risen substan-tially in older CKD patients with either dia-betes or hypertension (Figure .2). This is in stark contrast to nondihydropyri-dine CCB (diltiazem and verapamil) use, which has declined in the same timeframe. These results are not surprising, given the superior antihypertensive effect of the dihy-dropyridine CCB group.

This year we evaluate the use of both loop and thiazide (including metolazone) diuretics in CKD patients. Use of loop diuretics (bumetanide, furosemide or torse-

1.23

1.24 NHANES III & NHANES 19992002 participants, age 60 & olderPatients with diabetes who

receive glucose-lowering agents

mide) has increased in all groups of older CKD patients (Figure .22). This may indi-cate that more CKD patients are being iden-tified and treated. In contrast, except for use in CKD patients with congestive heart fail-ure, use of thiazide diuretics has declined. The use of diureticsparticularly thiazide diureticsmay increase, since JNC 7 guide-lines published in 2003 strongly advocate the use of thiazide diuretics as a first-line agent, and new NKF clinical practice guide-lines on hypertension in CKD patients advo-cate diuretic use following institution of an ACE-I or ARB.

The use of lipid-lowering agents has dramatically increased in all older CKD patients, from virtually no use in the NHANES III study to over 28 percent in the NHANES 9992002 population (Fig-ure .23). This most likely represents the influence of the National Cholesterol Edu-cation Program: Second Report of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in

Adults (Adult Treatment Panel IINCEP II), published in 994 and advocating more aggressive treatment of high cholesterol levels in the general population. In partic-ular, Figure .23 shows that statin use has increased up to ten-fold, while use of non-statins (bile acid sequestrants, fibrate acid derivatives, and niacin) actually fell slightly over the same period. In CKD patients with cardiovascular disease, statin use increased from 4 to 39 percent. Presumably, the use of statins will be even higher in the next NHANES dataset due to even more aggres-sive lipid goals as advocated by NCEP III, new NKF CKD guidelines on dyslipidemia, and increased safety data on statins.

We evaluated the use of glucose-low-ering medications in older CKD and non-CKD patients in the two NHANES pop-ulations (Figure .24). The use of insulin was substantially higher (30 percent and 28 percent versus 2 percent and 5 percent) in CKD Stage 4 patients than in non-CKD patients, most likely reflecting more severe diabetes in CKD patients. Accord-ingly, secretagogue use (sulfonylureas) was lower in CKD patients. Use of thiazolidine-dione (pioglitazone, rosiglitazone, and tro-glitazone, which was taken off the mar-ket in 2000) is approximately the same in CKD and non-CKD patients in the 9992002 NHANES population. As expected, metformin use is substantially less in CKD patients than non-CKD patients (2 ver-sus 34 percent), as it is contraindicated in patients with kidney disease. The manufac-turer has recently tightened its recommen-dation for use of metformin in patients with kidney disease, and currently states that patients whose serum creatinine concen-trations exceed the upper limit of normal for their age should not receive metformin due to increased risk of lactic acidosis.

{Figures .924} NHANES III 988994 & NHANES 9992002 patients age 60 & older; patients with eGFRs of less than 5 ml/min/.73 m2 are excluded. *Sample size less than 30, or coefficient of variation is not less than 30 percent.

1_23.zip1_24.zip

All NCKD DM CHF DM+CHF Other

ytiliba

borP

0.0

0.1

0.2

0.3

White

Black

All NCKD DM CHF DM+CHF Other

ytiliba

bor

P

0.0

0.1

0.2

0.3

White

Black

All CHF Other

ytiliba

bor

P

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7CKD

Non-CKD

All CHF Other All CHF Other

White

Black

All patients CKD Non-CKD

Insulin TZDs Secretagogues MetforminInsulin TZDs Secretagogues Metformin

stneita

p fo t

ne cr eP

0

10

20

30

40

50

NHANES III

NHANES 99-02

White Other race

*

White

Black

0.274 + (0.394)0.192 to

White

Black

OtherAllDM

CHFDM+CHF

OtherAllDM

CHFDM+CHF

OtherAllDM

CHFDM+CHF

ytiliba

borP

0.0

0.2

0.4

0.6

0.8CKD

Non-CKD

White

Black

All patients CKD Non-CKD

CKD+DM CKD+HTN CKD+CHF CKD+CVD

stneita

p fo t

necr e P

0

10

20

30

40

50

CKD+DM CKD+HTN CKD+CHF CKD+CVD

White Other race

NHANES III

NHANES 99-02

* **

* *

CKD+DM CKD+HTN CKD+CHF CKD+CVD

stneita

p fo t

n ecre P

0

20

40

60

80

NHANES III

NHANES 99-02

CKD+DM CKD+HTN CKD+CHF CKD+CVD

White Other race

59

2005 Annual Data Report 31CHRONIC KIDNEY DISEASEprevalentMedicare patients

Probability of lipid assessments within a year, by race1.29

1.30

1.31 NHANES III & NHANES 19992002 participants age 60 & olderUse of ACE-Is/ARBs in

patients with CKD, by race

NHANES III & NHANES 19992002 participants

Use of lipid lowering agents in patients with CKD, by race

In both the CKD and non-CKD dia-betic populations, and regardless of accom-panying diagnoses, white patients are more likely than black patients to receive at least two glycosylated hemoglobin tests within a yearthe minimum recommended by the American Diabetes Association (Fig-ure .27). Among CKD patients, for exam-ple, the probability of testing in whites is 0.5 percent greater overall, and 5 and 8 percent higher, respectively, for those with CHF or with other diagnoses.

Non-white patients are slightly more likely than whites to receive TZDs, secreta-gogues, or metformin (Figure .28).

The probability of lipid assessment is 5 percent higher in CKD patients overall than in those without the disease (Figure .29). Among those with diabetes, CHF, or both diagnoses, however, the probability of testing is nearly equal. By race, and regard-less of diagnoses, the probability of testing is greater in whites than in blacks0 per-cent higher in the CKD population overall, and 4 percent higher for those with diabe-tes or CHF.

In the NHANES 9992002 popula-tion, use of lipid lowering agents is much more common in whites compared to peo-ple of other races. In contrast, 62.8 percent of non-white patients with congestive heart failure receive ACE inhibitors compared to 42.5 percent in the white population (Fig-ures .303).

{Figures .2526} general Medicare: patients enter-ing Medicare before January , 2002, alive & remain-ing in the program through December 3, & with-out CKD diagnosed during 2002. Patients enrolled in an HMO, with Medicare as secondary payor, or with ESRD diagnosed during the year are excluded. Diabetes, CHF, & other comorbidities are defined in 2002. Patients censored at end of plan, death, & end of 2003. All testing tracked in 2003. In Figure .25, data on serum creatinine testing obtained from individual tests. CPT codes used for assessment of serum creat-inine include 80069, & 82565. Maps by HSA, unad-justed. {Figure .27} general Medicare patients enter-ing Medicare before January , 2002, alive & remaining in the program through December 3, & with diabetes diagnosed during 2002. Patients enrolled in an HMO, with Medicare as secondary payor, or with ESRD diagnosed during the year are excluded. CKD, CHF, & other comorbidities are defined in 2002. Patients censored at end of plan, death, & end of 2003. First glycosylated hemoglobin testing tracked in 2003. {Fig-ure .28} NHANES III 988994 & NHANES 999

lipid monitoring

2002 patients age 60 & older; patients with eGFRs of less than 5 ml/min/.73 m2 are excluded. *Sam-ple size less than 30 or coefficient of variation is not less than 30 percent. {Figure .29} general Medicare patients entering Medicare before January , 2002, alive & remaining in the program through December 3. Patients enrolled in an HMO, with Medicare as secondary payor, or with ESRD diagnosed during the year are excluded. CKD, CHF, & other comorbidities

are defined in 2002. Patients censored at end of plan, death, & end of 2003. First lipid monitoring tracked in 2003. {Figures .303} NHANES III 988994 & NHANES 9992002 patients age 60 & older; patients with eGFRs of less than 5 ml/min/.73 m2 are excluded. *Sample size less than 30 or coefficient of variation is not less than 30 percent.

1_29.zip1_30.zip1_31.zip

0

600

1,200

1,800

2,400

AllDiabetesCHFDiabetes+CHF

CKD: Medicare EGHP

93 94 95 96 97 98 99 00 01 02 03 0

600

1,200

1,800

2,400

00 01 02 03

Non-CKD

ksir ta sr ae y tneit a

p 0 00 ,1 rep et a

R

130 + (147)114 to

0

200

400

600

AllDiabetesCHFDiabetes+CHF

CKD: Medicare EGHP

93 95 97 99 01 030

200

400

600

00 01 02 03

Non-CKD

ksir ta sra ey tn eit a

p 000,1 rep et aR

0

50

100

150

200

AllDiabetesCHFDiabetes+CHF

CKD: Medicare EGHP

93 95 97 99 01 030

50

100

150

200

00 01 02 03

Non-CKD

ksir ta sraey tneita

p 000,1 rep etaR

0

25

50

75

100

125

AllDiabetesCHFDiabetes+CHF

CKD: Medicare EGHP

93 95 97 99 01 030

25

50

75

100

125

00 01 02 03

Non-CKD

ks ir ta s raey tne ita

p 00 0, 1 rep e ta R

00 01 02 03

0

20

40

60

AllDiabetesCHFDiabetes+CHF

CKD: Medicare EGHP

93 95 97 99 01 030

20

40

60 Non-CKD

ksir ta sr aey tne ita

p 00 0, 1 rep e taR

61

2005 Annual Data Report 31CHRONIC KIDNEY DISEASE1.34 1.35 prevalent patients

for ASHD

1.37 prevalent patients for bacteremia/

septicemia

prevalent patients

for congestive heart failure

1.36 prevalent patientsfor

pneumonia

cause-specific hospitalizations in patients with & without CKD, by prior comorbidity

Differences between the CKD and non-CKD populations are even greater for bacteremia/septicemia hospitalizations (Figure .37). For all patients and for those with diabetes, for example, rates are 4.6 and 3.0 times higher, respectively, in the Medicare popula-tion, and 27.5 and 7.5 times greater in those with EGHP coverage.

For both all-cause hospitalizations and the cause-specific rates examined here, the greatest differences between CKD and non-CKD patients consistently occur in the EGHP populationin

overall rates and in rates for patients diagnosed with diabetes dur-ing the entry period.

{Figures .3237} Medicare: prevalent patients continuously enrolled in Medicare Parts A & B, with no HMO coverage, & alive during the one-year entry period; adjusted for age, gender, & race. EGHP: prevalent patients age 2065 with fee-for-service coverage during the entire calendar year, & alive on the last day of the entry period; adjusted for age & gender. Patients diagnosed with ESRD before or during the entry period are omitted; 2003 patients used as reference cohort. Because of different age distributions in the two cohorts, rates are comparable only within the Medicare & EGHP cohorts, not between them. In Figures .3233, rates exclude hospitaliza-tions related to pregnancy & childbirth. In Figure .33, maps by HSA, unadjusted.

1_34.zip1_35.zip1_36.zip1_37.zip

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40 All

Acute kidney failure as the principal diagnosis

Acute kidney failure as the secondary diagnosis

ytil iba

bor

p e vi talu

mu

C

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

All Diabetes CHF DM+CHF

1-year

2-year

3-year

1-year

2-year

3-year

1-year

2-year

3-year

With prior CKDWithout prior CKD

CHRONIC KIDNEY DISEASE3

62

1

1.38 Medicare Parts A & B patientsProbability of hospitalization

for acute kidney failuren this spread we examine hospitalizations for acute kid-ney failure, along with their subsequent outcomes. In the Medicare population without CKD, the probability of a

hospitalization for acute kidney failure is 0.034 at three years (Fig-ure .38). As we look at complicating diagnoses of diabetes, con-gestive heart failure, and diabetes combined with CHF, this three-year probability rises steadilyto 0.07, 0., and 0.6, respectively. Among patients with a diagnosis of CKD, the probability of an acute kidney failure hospitalization is, not surprisingly, far higher, from 0.22 overall to 0.39 for those with diabetes and CHF com-bined. In this CKD population, the probability of hospitalization in patients with both diabetes and CHF is 49 percent greater at one year, and 43 percent greater at three years, than that of patients with diabetes alone.

Similar patterns occur when we categorize hospitalizations by whether acute kidney failure is the primary or secondary diagno-sis at admission. For acute kidney failure as a primary diagnosis, the one-year probability of hospitalization in CKD patients ranges from 0.03 overall to 0.05 in those with diabetes and CHF, while the three-year probability ranges from 0.07 to 0.3. Probabilities are greater for hospitalizations in which acute kidney failure is the secondary admitting diagnosis, at three years reaching 0.8 overall and 0.32 for those with combined diabetes and CHF.

In Figures .3942 we present data on ESRD and death follow-ing a hospitalization for acute kidney failure. Across outcomes and diagnoses, the highest event rates occur within the first three months after an acute kidney failure hospitalization, with rates then leveling out or decreasing slightly in the following 33 months.

For patients who have a prior diagnosis of CKD, the rate of death in the first three months after hospitalization is 2.74 times greater than the rate of an ESRD diagnosis. Patients without a CKD diagnosis prior to that hospitalization, however, face a death rate that is 320 percent greater than the risk of developing ESRD.

In the prevalent Medicare population as a whole, 2.6 percent of patients without prior CKD are diagnosed with ESRD in the first three months, compared to . percent of patients with an earlier CKD diagnosis. The three-month rates do not vary widely by diag-

Acute kidney failure: hospitalization & outcomes

O

1_38.zip

Months after AKF hospitalization

3 6 9 12 15 18 21 24 27 30 33 360

2

4

6

8

10

12

14

With prior CKDWithout prior CKD

3 6 9 12 15 18 21 24 27 30 33 360

10

20

30

40

50All: ESRD

stneita

p 001 rep etaR

Death

Months after AKF hospitalization

3 6 9 12 15 18 21 24 27 30 33 360

2

4

6

8

10

12

14

With prior CKDWithout prior CKD

3 6 9 12 15 18 21 24 27 30 33 360

10

20

30

40

50Diabetes: ESRD

stneita

p 001 rep etaR

Death

Months after AKF hospitalization

3 6 9 12 15 18 21 24 27 30 33 360

2

4

6

8

10

12

14

With prior CKDWithout prior CKD

3 6 9 12 15 18 21 24 27 30 33 360

10

20

30

40

50CHF: ESRD

stnei ta

p 001 r ep et aR

Death

Months after AKF hospitalization3 6 9 12 15 18 21 24 27 30 33 36

0

2

4

6

8

10

12

14

With prior CKDWithout prior CKD

3 6 9 12 15 18 21 24 27 30 33 360

10

20

30

40

50Diabetes+CHF: ESRD

stneita

p 001 rep etaR

Death

63

2005 Annual Data Report 31CHRONIC KIDNEY DISEASE

1.39 prevalent Medicare patientsAll

patients 1.40prevalent Medicare patients

Patients with diabetes

1.41 prevalent Medicare patientsPatients with

congestive heart failure 1.42prevalent Medicare patients

Patients with diabetes & congestive heart failure

outcomes following first hospitalization for acute kidney failure

nosis: 3.2 and 2.5, respectively, for patients with diabetes, 2.4 and 0.6 for those with CHF, and 3. and 2.0 for those with diabetes and CHF combined. Patients with CKD are, then, 3.94.4 times more likely than patients without prior CKD to be diagnosed with ESRD during the three months following an acute kidney failure hospitalization.

Results differ significantly for death after an acute kidney fail-ure hospitalization. Across diagnoses, three-month death rates are higher in patients without prior CKD than in those who have the

diagnosis2.4 percent higher overall, and 9.4, 3.5, and 5.9 per-cent higher in those with diabetes, CHF, and the two combined, respectively. After this three-month period, however, rates are con-sistently greater in the CKD population.

{Figure .38} patients continuously enrolled in Medicare Parts A & B, with no HMO coverage during 2000, & alive on December 3, 2000. Patients diagnosed with ESRD before December 3, 2000 are excluded. {Figures .3942} prevalent Medicare patients with hospitalization for acute kidney failure during 9992000; adjusted for age, race, & gender. A one-year entry period prior to the first hospital-ization is used to define cardiovascular disease & diabetes.

1_39.zip1_40.zip1_41.zip1_42.zip

Figure number 1.3 1.3 1.3 1.5 1.6 1.7 1.8 1.9 1.10 1.11 MC 92-3 MC 97-8 MC 02-3 Overall value for all pts 23.4 33.8 56.25 0 .20 0.20 0.06 0.09 0.01 0.38 0.27Total patients 1,354,207 1,283,428 1,385,876 235,798 235,798 235,798 235,798 235,798 235,798 52,121 Overall value for pts mapped 23.4 33.8 56.2 0.20 0.20 0.06 0.09 0.01 0.38 0.27Missing HSA/state: pts dropped 17,007 19,063 20,706 5,782 5,782 5,782 5,782 5,782 5,782 832

Figure number 1.12 1.13 1.14 1.15 1.16 1.17 1.18 1.25 1.25 1.26 Wh Bl WhOverall value for all pts 0.08 19.3 56.2 41.1 58.5 53.1 12.9 0.20 .018 0.21Total patients 52,121 24,531 24,531 24,531 52,121 45,414 44,942 191,934 30,744 191,934 Overall value for pts mapped 0.08 19.3 56.2 41.1 58.5 53.1 12.9 0.20 0.18 .21Missing HSA/state: pts dropped 832 547 547 547 832 756 750 3,483 363 3,483

Figure number 1.26 1.33 1.33 Bl w/CKD w/o CKDOverall value for all pts 0.18 107.7 36.3 Total patients 30,744 52.121 1,421,218Overall value for pts mapped 0.19 107.9 36.4Missing HSA/state: pts dropped 363 832 19,673

maps: national means & patient popUlations

CHRONIC KIDNEY DISEASE3

64

1 Chapter summaryFigure . In almost two-thirds of Medi-care patients age 65 and older with CKD, CKD is accompanied by diabetes, CHF, or both diseases combined.

Figure .4 The percent of CKD patients with just diabetes increased in the Medi-care population from 6.7 in 992993 to 23.7 in 20022003, and in the EGHP population from 23.8 in 9992000 to 26.6 in 20022003.

Figure .6 The overall probability of microalbuminuria or proteinuria testing reaches only 0.04 in the Medicare popu-lation, and 0.0 for EGHP patients. Fig-ure .8 The probability of calcium and phosphorus assessment is 0.06 in Medi-care non-CKD patients overall, and 0.02 in the EGHP population.

Figure . The probability of a CKD patient receiving calcium phospho-rus testing is 0.27 for those with Medi-care coverage, and 0.7 for those under EGHP. Figure .4 In the Medicare CKD population, the probability of diabetic patients receiving at least two HbAc tests within a year is 0.56, compared to 0.36 for EGHP patients.

Figure .9 The use of ACE-Is and ARBs for renal protective treatment has almost doubled in the CKD population. Figures .2223 Use of diuretic therapy in CKD

introduction

prevalence of CKD

assessment of CKD in risk populations

preventive healthcare monitoring in CKD patients

prescription drug therapy in CKD patients

patients with CHF has grown, as has the use of lipid-lowering agents in both dia-betics and those with CVD.

Figure .27 In both the CKD and non-CKD diabetic populations, and regard-less of accompanying diagnoses, white patients are more likely than black patients to receive at least two glycosyl-ated hemoglobin tests within a year.

Figure .32 The overall hospitalization rate for Medicare CKD patients reached ,068 per ,000 patient years in 2003nearly three times greater than that in non-CKD patients. The rate for diabetic patients is 2.3 times higher in those with CKD, reaching ,276.

Figure .38 In non-CKD Medicare patients, the probability of a hospital-ization for acute kidney failure is 0.034 at three years. Among patients with a diagnosis of CKD, the probability is far higher, from 0.22 overall to 0.39 for those with diabetes and CHF combined. Fig-ure .39 For patients with a prior diagno-sis of CKD, the rate of death in the first three months after hospitalization for AKF is 2.74 times greater than the rate of an ESRD diagnosis. Patients without a CKD diagnosis prior to that hospital-ization, however, face a death rate that is 320 percent greater than the risk of developing ESRD.

racial disparities in the assessment of CKD

hospitalization rates in CKD & non-CKD patients

acute kidney failure: hospitalization & outcomes

Chronic Kidney DiseaseIntroductionPrevalence of chronic kidney diseaseAssessment of populations at risk for CKDPreventive healthcare monitoring in CKD patientsPrescription drug therapy in CKD patientsRacial disparities in the assessment of CKDAdjusted hospitalization rates in CKD & non-CKD patientsAcute kidney failure: hospitalization & outcomesChapter summary