2008 USRDS Annual Data Report
This Chapter
Download
Download PDF: Win*   Mac*
Download all slides: Win*   Mac*
*corresponding data in Excel included
Search This Page
Search All

Precis—an introduction to CKD in the U.S.

Prior to 2002, the disease now defined as chronic kidney disease (CKD) had multiple designations, including chronic renal failure, pre-ESRD, renal failure, renal damage, and kidney disease. Since the introduction of the new NKF classification system for CKD in February, 2002, however, the disease has become more clearly defined, and has emerged as a major public health issue.

The five-stage classification was developed through analysis of data from the National Health and Nutrition Examination Survey (NHANES), and is based on an estimated glomerular filtration rate (eGFR, calculated from the serum creatinine level) and level of proteinuria. Stage 1 is defined by an eGFR of >90 ml/min/1.73 m2 and either >20 mg/dl of albumin in the urine, an albumin/creatinine ratio ≥30 mg/g, or other evidence of structural damage to the kidney. Stage 2 has a similar definition, but an eGFR of 60–89. And Stages 3, 4, and 5 are defined solely by an eGFR of 30–59, 15–29, and less than 15, respectively. As identified through these definitions, the burden of CKD in the U.S. has been approximately 13–16 percent. In any estimate, certainly, there are assumptions that need to be addressed, in this case assumptions based on the definitions of proteinuria and albuminuria, and on the level of eGFR in elderly and minority populations that truly defines disease.

The impact of this classification has been considerable, with over 3,000 articles on CKD referenced in Medline since publication of the guidelines — compared to just 200 in the prior 20 years. And the association between CKD stage and morbidity and mortality is now well established across the world. Patients with CKD have a high burden of cardiovascular disease, diabetes, and laboratory abnormalities consistent with metabolic syndrome.

This year the USRDS introduces a full volume on CKD, its relationship to and interaction with other diseases, and its adverse outcomes. On the next page we present estimates of the number of people in the U.S. with CKD — using data from the NHANES surveys — as well as the number of recognized CKD patients in the Medicare data, defined from reported diagnosis codes. These data show that the prevalence of estimated CKD has increased 20–25 percent over the past decade, an increase similar to the 30 percent reported in JAMA in November, 2007 (Coresh et al.). Recognized CKD in the Medicare population, however, shows a different trend, doubling between 1992 and 2002, and increasing another 65 percent between 2002 and 2006. (The large increase in 2006 may reflect a change in ICD-9-CM diagnosis codes, which now reflect CKD stages, but may also indicate increased availability of information in both the lay and scientific press on the importance of CKD.) This growth in the recognized Medicare CKD population demonstrates the complex issues of analyses based on recognition versus those using estimates from laboratory data, as from the NHANES sample.

The recognized CKD population covered by private employer group health plans is considerably smaller, and estimates are vulnerable both to the number of available diagnosis code fields and to private payor requirements regarding the use of the new ICD-9-CM codes (introduced in October, 2005) with the same rules noted in Medicare. Claims-based CKD diagnosis definitions have very high specificity but low sensitivity, demonstrating a lack of disease recognition in these populations compared to the real prevalence in the general population. As a result, prevalence estimates from billing data are only about one-sixth as great as those from NHANES. Even in the Medicare system, in fact, CKD as identified solely through the new diagnosis codes — without the use of prior criteria such as diabetes or hypertension with renal complications — has been underreported. It will likely take a number of years to assess the impact of the new diagnosis codes, and to determine the public health knowledge of CKD as a major comorbidity and as a cost multiplier in relationship with diabetes and cardiovascular disease.

figure p.1 NHANES: point prevalent NHANES 1988–1994 & 1999–2006 participants, age 20 & older. Patient counts from 1988–1994 survey adjusted to 1990 U.S. population; counts from 1999–2006 survey adjusted to 2000 U.S. population; CKD defined by biochemical data. Medicare: point prevalent general Medicare patients age 65 & older, surviving entire cohort year with Medicare as primary payor (& not enrolled in an HMO). ESRD patients excluded for CKD estimation; CKD defined by diagnosis codes.

Morbidity from hospitalization can be difficult to assess in the national sample, but can be examined in the Medicare and EGHP datasets. In the administrative data CKD is, by its nature, severe enough to be reported, and therefore reflects sicker populations. Once again, the hazard ratio of hospitalization shows the importance of CKD and its interactions with other major chronic diseases such as congestive heart failure and diabetes. Associations of mortality and CKD are also clear across datasets. Using current clinical practice guidelines, the treatment of CKD patients centers on blood pressure control through angiotensin converting enzyme inhibitors and angiotensin receptor blockers (ACE-Is/ARBs). We assess this aspect of care by using Medicare Current Beneficiary Survey (MCBS) data, a collection of information on prescription drug treatment in the Medicare population and of medication billing records from health plan pharmacy claims. With this approach, the high specificity provides a high likelihood that an individual has CKD, and the prescription drug therapy provides some assessment of treatment. In the Medicare population carrying a diagnosis of diabetes, for example, it appears that two in three patients use an ACE-I/ARB. Of CKD patients without diabetes, 57 percent appear to use either drug. Use is slightly higher in the non-Medicare population with CKD and diabetes, but considerably lower in those without diabetes.

Expenditures associated with CKD are considerable in both the Medicare and EGHP populations. Private health plan expenditures are almost double those of the Medicare population, which may reflect cost shifting or payment practices by the health plans, or relate to provider pricing. In almost all categories, private health plan expenditures are higher than those of Medicare, which covers an older and sicker population. Overall, the CKD population presents an important public health issue for the United States and other countries. As demonstrated in Volume Two of the ADR, the ESRD population is increasing in both size and cost. Those patients transitioning from CKD to ESRD merit more attention, since the delivery of care is less than optimal and has been shown to influence morbidity and mortality on dialysis. More importantly, the CKD population is more likely to die from cardiovascular disease than to reach ESRD, a finding central to the public health issue. Patients with CKD have poor blood pressure control, poor control of cardiovascular risk factors, and high event rates, all amenable to detection and treatment. This new volume on CKD will increase the scope of USRDS work, as we define the burden of disease, access to care, and the impact of treatment on overall morbidity and mortality.

Top

figure p.2 The prevalence of CKD as identified through claims in the Medicare and Medstat cohorts is considerably lower than that identified through estimated glomerular filtration rates in the Ingenix i3 cohort, and also lower than that found in the NHANES cohort, suggesting that mild CKD is not picked up from diagnosis codes. table p.b Comorbidity generally increases with CKD stage, increasing age, and African American race. table p.e Detail on PPPM expenditures shows that patients with CKD incur greater expenditures than those without. Most outpatient costs are greater in the Medstat population; costs for such services as skilled nursing facilities, home health care, and hospice, in contrast, are greater for patients in the Medicare population.

table p.a This table looks at patients in the Medicare, Medstat, and Ingenix i3 datasets, defining CKD through the standard methodology of one or more inpatient diagnosis codes or two or more outpatient codes (“All codes”), and through the new ICD-9-CM codes; it also presents NHANES data by CKD stage. Patterns of increasing age and comorbidity by higher CKD stage are more apparent in the NHANES data than in the new ICD-9-CM codes, suggesting some inaccuracy in the use of these codes.

table p.b Comorbidity generally increases with CKD stage, increasing age, and African American race. The prevalence of hyperten-sion is considerably higher than that of other comorbidities, approaching 80–90 percent in patients with CKD of Stages 4–5.

prevalence of chronic kidney disease; comorbidities

figure p.3 The employed populations in the Medstat and Ingenix i3 datasets show similar patterns of comorbidities. Not surprisingly, Medicare patients age 65 and older have con-siderably more hypertension, which does not vary by the presence or absence of CKD.

figures p.4, p.5 & table p.c Hazard ratios for hospitalization based on combinations of comor-bid conditions (CKD, diabetes, and congestive heart failure) generally show an increasing hazard with more conditions. Although direct comparison of hazard ratios across datasets is not possible, qualitatively, ratios tend to be higher for Medstat patients than for those with Medicare coverage. Ratios are similar, within comorbidity burden, in the Medstat and Ingenix i3 cohorts. And ratios for dually-enrolled Medicare patients are slightly lower than those for all Medicare patients. These patterns are inversely correlated with the prevalence of comorbidity: the lower the prevalence (Medstat and Ingenix have the lowest), the higher the hazard ratio for hospitalization; while the highest prevalence (dually-enrolled patients) shows the lowest hazard ratios.

morbidity & mortality

figures p.6, p.7 & table p.d Hazard ratios for mor-tality show similar patterns to those seen for hospitalization on the previous page. Com-pared to patients in other cohorts, patients with dual Medicare/Medicaid coverage have the lowest hazard ratios associated with increasing comorbidity burden

figures p.8, p.9, & p.10 Among CKD patients in the Medicare Current Beneficiary Survey (MCBS), use of ACE-Is/ARBs has changed little since 1999 for those with diabetes, remaining at 69–73 percent. Use in the non-diabetic CKD population, in contrast, has grown from 31 percent in 1999 to 57 percent in 2004. Use of lipid lowering agents has increased for CKD patients with and without diabetes, reaching 69 and 46 percent, respectively, in 2004. Among Medstat CKD patients, the use of ACE-Is/ARBs is higher in those with diabetes than in those without. In 2006, for example, 79.4 and 75.8 percent of diabetic patients age 20–64 and 65 and older, respectively, used these medications, compared to 47.8 and 60.0 percent of non-diabetic patients. These pat-terns are similar in patients using lipid lower-ing agents. In 2006, for instance, 70.1 and 75.2 percent of diabetic patients age 20–64 and 65 and older used some type of lipid lowering medication, compared to 35.9 and 56.2 per-cent of non-diabetics. Among Ingenix i3 patients with CKD, use of ACE-Is/ARBs is higher for diabetics as well, at 78.5 percent in 2006 for patients age 20–64, and 70.5 percent for those 65 and older, compared to 41.5 and 53.6 percent in non-diabetic patients. In the younger popu-lation, use of lipid lowering medications is twice as high among those with diabetes, and reaches 69.4 percent in diabetics age 65 and older, compared to 49.2 percent in their non-diabetic counterparts

prescription drug therapy; economics of chronic kidney disease

figure p.11 Since 1993, per person per month (PPPM) inpatient/outpatient costs have increased 28 percent for Medicare patients with CKD, rising to $1,089 in 2006. Costs for patients with no CKD nearly doubled dur-ing the same period, reaching $267. Similar increases are evident in the Medstat (EGHP) CKD and non-CKD populations, at 33.4 and 56.6 percent, respectively. Costs reached $2,029 in 2006 for those with CKD — nearly double that found in the Medicare popula-tion. Physician/supplier costs in 2006 for Medstat patients with CKD were $540 per person, compared to $456 in Medicare patients. Costs for those with no CKD were $118 and $185, respectively.

table p.e Detail on PPPM expenditures shows, not surprisingly, that patients with CKD incur greater expenditures than those without. Most outpatient costs are greater in the Medstat population; costs for such ser-vices as skilled nursing facilities, home health care, and hospice, in contrast, are greater for patients in the Medicare population.

Top


Captions

figures p.2–3 & tables p.a–b general Medicare: point prevalent patients age 65 & older, surviving all of 2006 with Medicare as primary payor & not enrolled in an HMO. ESRD patients excluded. CKD & other comorbidities defined by diagnosis codes in 2006. Medstat: point prevalent patients age 20–64, surviving all of 2006 & enrolled in a fee-for-service plan. ESRD patients excluded. CKD & other comorbidities defined by diagnosis codes in 2006. Ingenix i3: patients age 20–64, surviving all of 2006, & enrolled in a fee-for-service plan. ESRD patients excluded. Comorbidities defined by diagnosis codes. NHANES: 1999–2006, participants age 20 & older. • *All codes: CKD identified through one or more inpatient/outpatient institutional claims (inpatient hospi-talization, skilled nursing facility, or home health agency), or two or more institutional claims (outpatient) or physician/supplier claims — the method used in other USRDS studies. **CKD iden-tified through the 585.x ICD-9-CM codes. ^In USRDS analyses, patients with ICD-9-CM code 585.6 are considered to have code 585.5; see Appendix A for details.

figures p.4–5 & table p.c Medicare-only: includes gen-eral Medicare patients, age 66 & older on January 1, 2005, continuously enrolled in Medicare inpatient/outpatient & physician/supplier program during 2004. Patients enrolled in an HMO or Medicaid during 2004 are excluded. Dually-enrolled: same as Medicare cohort, with patients also enrolled in Medicaid at any time during 2004. Medstat & Ingenix i3: patients age 50–64 on December 31, 2004, continuously enrolled in a fee-for-service commer-cial health plan during 2004. All datasets exclude patients diagnosed with ESRD before January 1, 2005, & patients with a bridge hospitalization spanning January 1, 2005. Comorbidity groups are mutually exclusive, & CKD, diabetes, & CHF are defined from claims during 2004. Patients are followed for up to two years from January 1, 2005, to December 31, 2006. Figure p.4 adjusted for age, gender, & race; Figure p.5 adjusted for age & gender. figures p.6–7 & table p.d Medicare-only: general Medicare patients entering Medicare before January 1, 2004, alive & age 65 or older on December 31. Patients enrolled in an HMO, with Medicare as secondary payor, diagnosed with ESRD during the year, or enrolled in Medicaid in 2004 are excluded. Dually-enrolled: general Medicare patients entering Medicare before January 1, 2004, dually-enrolled in Medicare & Medicaid in 2004, & alive & age 65 or older on December 31. Patients enrolled in an HMO, with Medicare as secondary payor, or diagnosed with ESRD during the year are excluded. Comorbidity groups are mutually exclusive, & CKD, diabetes, & CHF are defined during 2004. Patients are followed for two years from January 1, 2005, to December 31, 2006.

figure p.8 MCBS patients with CKD, age 65 & older; ESRD patients excluded. figure p.9 Med-stat patients with CKD, age 20 & older; ESRD patients excluded. figure p.10 Ingenix i3 patients with CKD, age 20 & older; ESRD patients excluded. figure p.11 Medicare: period prevalent general Medicare patients age 65 & older, with Medicare as primary payor, & not enrolled in Medicare Advantage. Medstat; period prevalent patients age 50–64, enrolled in a fee-for-service plan. CHF, diabetes, & CKD determined from claims, & costs are PPPM inpatient plus outpa-tient cost (does not include skilled nursing facil-ity, home health, or hospice) & physician/sup-plier costs. table p.e Medicare: period prevalent general Medicare patients age 65 & older, with Medicare as primary payor, & not enrolled in Medicare Advantage. Medstat; period prevalent patients age 50–64, enrolled in a fee-for-service plan. CHF, diabetes, & CKD determined from claims, and costs are PPPM for calendar year 2006, with actuarial categories determined by place of service, CPT, revenue, & HCPCS codes, & provider specialty. Oral prescription drugs (not present for Medicare) are included in the outpa-tient pharmacy category for Medstat.