Appendices

In this appendix we describe the datasets and methods used for the CKD analyses in this volume. Appendix B includes information on all USRDS products and services. Data management and preparation, database definitions, and the data sources used for ESRD analyses are described in the appendix of Volume Two.

Data sources

The USRDS maintains a stand-alone database with data on diagnoses and demographic characteristics of CKD and ESRD patients, along with biochemical data, dialysis claims, and information on treatment and payor histories, hospitalization events, deaths, physician/supplier services, and providers.

CMS MEDICARE ENROLLMENT DATABASE

The Enrollment Database (EDB) of the Centers for Medicare and Medicaid Services (CMS) is the designated repository of all Medicare beneficiary enrollment and entitlement data, and provides current and historical information on residence, Medicare as secondary payor (MSP) and employer group health plan (EGHP) status, and Health Insurance Claim/Beneficiary Identification Code (HIC/BIC) cross-referencing.

ESRD MEDICAL EVIDENCE FORM (CMS 2728)

The ESRD Medical Evidence (ME) form is used as the official form for registering individual patients at the onset of ESRD. This form must be submitted by dialysis or transplant providers within 45 days of ESRD initiation. The CMS, USRDS, and renal research communities rely on this form to ascertain basic patient demographic attributes, the primary cause of renal failure, major comorbidities, and biochemical test results at the time of ESRD initiation.

The third major revision of the ME form, released in May, 2005, was intended to remedy several shortcomings found in the 1995 form and its earlier version. Key additions target pre-ESRD care and vascular access use, and additional new fields collect information on glycosylated hemoglobin and lipid testing, on the frequency of hemodialysis sessions, and on whether patients are informed of transplant options. This new form will help federal and private researchers gain better insights into the health and care of ESRD patients prior to their entry into the program.

ESRD DEATH NOTIFICATION FORM (CMS 2746)

The ESRD Death Notification form is used as the official form for reporting the death of individual patients with ESRD. According to CMS policy, this form must be submitted by dialysis or transplant providers within 30 days of a patient’s death, and provides the date and causes of death (primary and secondary), reasons for discontinuation of renal replacement therapy, if applicable, and evidence of hospice care prior to death. It is the primary source of death information for CMS and the USRDS, identifying more than 99 percent of deaths. The USRDS also utilizes the Social Security Administration’s (SSA) Death Master File as a supplemental data source for ascertaining death in a small group of lost-to-follow-up ESRD patients; this file, however, identifies only all-cause deaths.

CMS 5 percent STANDARD ANALYTICAL FILES (SAFs)

These files contain billing data from final action claims, submitted by Medicare beneficiaries, in which all adjustments have been resolved. The claims data are selected randomly from general Medicare claims (i.e. final action claims) using five combinations of the last two digits of the CMS Health Insurance Claims (HIC) number: 05, 20, 45, 70, and 95. Since the same two-digit numbers are used each year to create the 5 percent general Medicare SAFs, one should expect to see the same beneficiaries in these annual datasets. These claims are categorized into the inpatient (IP), outpatient (OP), home health agency (HHA), hospice (HS), skilled nursing facility (SNF), physician/supplier (PB), and durable medical equipment (DME) SAFs.

The files are updated each quarter through June of the next year, when the annual files are finalized. Datasets for the current year are created six months into the year and updated quarterly until finalized at 18 months, after which they are not updated to include late arriving claims. Annual files are thus approximately 98 percent complete. The USRDS 2009 ADR includes all claims up to December 31, 2007.

MEDICARE CURRENT BENEFICIARY SURVEY (MCBS)

The Medicare Current Beneficiary Survey is a longitudinal survey of a nationally representative sample of aged, disabled, and institutionalized Medicare beneficiaries. The MCBS contains information on the health status, health care use and expenditures, drug prescriptions, health insurance coverage, and socioeconomic and demographic characteristics of the entire spectrum of Medicare beneficiaries. Data are made available by CMS in two datasets: Access to Care (1992–2006), and Cost and Use (1992–2005), with the 2006 and 2005 files, respectively, the latest updates for the 2009 ADR.

In the fall of 1991, the MCBS began to be conducted three times per calendar year (winter, summer, and fall), and in 1994 a sample rotation scheme was introduced. Survey participants are kept in the sample for four years, with approximately one-third rolling off, and with new participants added each fall to keep the overall sample size at approximately 12,000 each calendar year.

THOMSON reuters MARKETSCAN DATA

The Thomson Reuters MarketScan Commercial Claims and Encounters Database includes specific health services records for employees and their dependents in a selection of large employers, health plans, and government and public organizations. The database includes nine files: Annual Enrollment Summary Table, Enrollment Detail Table, Inpatient Admissions Table, Inpatient Services Table, Outpatient Services Table, Outpatient Pharmaceutical Claims Table, Facility (Inpatient and Outpatient) Header Table, Aggregated Populations Table, and the RED BOOK (prescription drug information by National Drug Code). The strength of this database lies in the quality of its cost information, where claims data include actual paid dollars and net payments by the insurer.

The MarketScan database links billing and encounter data to detailed patient demographic and enrollment information across sites and types of providers, and over time from 1999 to 2007, and includes commercial health data from approximately 100 payors. About 80 percent of those covered are self-insured. Each year the database contains health data for about 10.5 million people. For details about the MarketScan data, please visit www.usrds.org.

INGENIX i3 DATA

The Ingenix i3 database is a commercial and non-capitated health plan database covering employees from multiple employers within a single insurer. In addition to the usual service encounter and drug data, similar to that of the MarketScan database, this database also includes laboratory data, allowing for comparisons between claims-based and lab-based definitions of diseases. In order to protect the discount structure of its business, the billing data of this single insurer discloses only charged dollars without actual paid amounts or the portion paid by the insurer.

The Ingenix i3 database links billing and encounter data to detailed demographic and enrollment information of individual employees from 2000 to 2007, and contains health data for about 14 million people annually. For details about what is contained in the Ingenix i3 data, please visit our website at www.usrds.org.

NATIONAL HEALTH & NUTRITION EXAMINATION SURVEY (NHANES)

NHANES is a series of health examination surveys conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC). Begun in 1959, NHANES is designed to monitor the health and nutritional status of the non-institutionalized civilian population in the United States. NHANES III was conducted in two phases between 1988 and 1994. In 1999, NHANES became a continuous annual survey to allow annual estimates, with release of public-use data files every two years. Both NHANES III and NHANES 1999–2006 were nationally representative cross-sectional surveys and used a complex, stratified, multistage probability cluster sampling design that included selection of primary sampling units (counties), household segments within the counties, and sample persons from selected households. Survey participants were interviewed in their homes and/or received standardized medical examinations in mobile examination centers. Both surveys over-sampled African Americans, Mexican Americans, and individuals age 60 or older to improve the estimates for these subgroups.

PAYORS

Information on payors is obtained from the CMS EDB. We also examine Medicare outpatient claims to identify patients for whom the EDB does not indicate Medicare as primary payor (MPP), but who have at least three consecutive months of dialysis treatment covered by Medicare; these patients are also designated as having MPP coverage. From these two data sources we construct a payor sequence file to define payor history, and, starting with the 2003 ADR, we use this file to identify Medicare eligibility status and other payors.

The construction of this file is similar to that of the treatment history file. Payor status is maintained for each ESRD patient from the first ESRD service date until death or the end of the study period. Payor data are used to categorize a patient as MPP, Medicare as secondary payer (MSP) with EGHP, MSP non-EGHP, Medicare Advantage (Medicare + Choice), Medicaid, or a combination of payers. With this approach, the USRDS is now able to apply payor status information in all outcome analyses using the “as-treated” model (see the discussion of Chapter Eleven in Volume Two).

UNITED STATES CENSUS

In rate calculations throughout this year’s ADR we use data from the 2000 U.S. Census, and incorporate CDC population estimates by race.

Database definitions

EGHP DATA

To examine the demographic segment represented by the EGHP data, we use enrollment information to construct yearly cohorts of enrollees younger than 65. To ensure that we select enrollees with the potential to generate claims evidence appropriate to the demands of analytical methods, rules for inclusion also include 12 months of continuous coverage in a commercial fee-for-service plan, and, for medication analyses, continuous prescription drug coverage. Comorbidities are identified using claims. Patients with at least one inpatient claim or at least two outpatient claims during the period of interest and with a diagnosis code of a particular comorbidity are identified as having that comorbidity.

ESRD COHORT IN THE EGHP POPULATION

Because the MarketScan and Ingenix i3 databases do not provide identifiable data elements, we are unable to link them directly to the USRDS ESRD registry. To identify ESRD patients, we therefore use a process similar to that used in the registry. Transplant patients are identified by evidence of a kidney transplant procedure or an adverse graft event, and chronic dialysis patients by evidence of continuous history of dialysis therapy, with at least three consecutive months of dialysis service and with dialysis service claims in at least 70 percent of treatment months. Treatment months are defined by the period from the first dialysis claim to the earliest of kidney transplant, death, or end of enrollment. Both inpatient and outpatient claims are evaluated for evidence of dialysis service history.

The first ESRD service date is set to the earliest of the first dialysis service date or the transplant date. If neither is available, the start of enrollment is used. Incidence is defined by a first ESRD service date at least 60 days after the start of enrollment.

Précis

For a description of analytical methods related to age, gender, race, ethnicity, comorbidities, and CKD stages in Table p.a, please see the discussion for Chapter One, below. The glomerular filtration rate (ml/min/1.73 m2) is estimated by the MDRD method, based on the standardized creatinine value for NHANES III and NHANES 1999–2000, 2001–2002, 2003–2004, and 2005–2006, separately, as based on NCHS recommendations.

Additional figures and tables in the Précis are taken directly from the chapters; methods can be found in the chapter discussions.

CKD in adult NHANES participants
Chapter One

DATABASE DESIGN, SETTING, & STUDY PARTICIPANTS

The surveys used in this chapter include NHANES III (1988–1994), NHANES 1999–2000, NHANES 2001–2002, NHANES 2003–2004, and NHANES 2005–2006, and populations are limited to participants age 20 and older. The public use NHANES III Linked Mortality File provides mortality follow-up data from the date of NHANES III survey participation (1988–1994) through December 31, 2000, based upon the results from a probabilistic match between NHANES III and the National Death Index death certificate records. Study populations using this mortality data are limited to participants age 20 and older, and the mortality follow-up month is greater than zero.

MEASUREMENTS

In this chapter age is defined as the participant’s age at the time of the household interview, and grouped into ages 20–39, 40–59, and 60 and older. Race/ethnicity is defined as non-Hispanic white, non-Hispanic black, and other, and ethnicity as Hispanic (including Mexican-American and other Hispanic) and non-Hispanic only.

Obesity is defined as a BMI of 30 kg/m2 or above.

Participants with self-reported diabetes are those ever told by a doctor that they have diabetes or sugar diabetes (other than during pregnancy). In NHANES 1999–2006, participants answering “borderline” are classified as non-diabetic. Participants with self-reported congestive heart failure are those ever told by a doctor that they have congestive heart failure. And participants with self-reported cardiovascular disease are those with at least one of the following self-reported diseases: coronary heart disease, angina/angina pectoris, heart attack, congestive heart failure, or stroke.

Smokers are identified by an affirmative answer to the question: “Have you smoked at least 100 cigarettes during your entire life?” then further classified by their answer to the question: “Do you smoke cigarettes now?” Those with affirmative answers are classified as smokers; others are defined as non-smokers.

WHO anemia is defined as a hemoglobin less than 13 g/dl in males and less than 12 g/dl in females.

Self-reported hypertension is identified by an affirmative answer to the question: “Have you ever been told by a doctor that you had hypertension, also called high blood pressure?”

Connected tissue disorder is identified by an affirmative answer to the question: “Has a doctor ever told you that you had arthritis?”

Hip fracture is identified by an affirmative answer to the question: “Has a doctor ever told you that you had a broken or fractured hip?”

Cancer is identified by an affirmative answer to the question: “Has a doctor ever told you that you had cancer or malignancy?”

Participants with self-reported COPD are those with at least one of the following self-reported diseases: asthma, chronic bronchitis, or emphysema.

Hepatitis C disease is defined as a confirmed hepatitis C antibody.

In NHANES 1999–2006, systolic blood pressure (SBP) / diastolic blood pressure (DBP) for each participant is calculated as the mean of all measured SBP / DBP.

Microalbuminuria is defined by the ratio of urinary albumin (mg/l) to urinary creatinine (mg/dl; ACR). Participants with a valid ACR are classified as having microalbuminuria if this value is not less than 30 mg/g.

The glomerular filtration rate (ml/min/1.73 m2) is estimated by two methods. The first is the MDRD method, based on the standardized creatinine value for NHANES III and NHANES 1999–2000, 2001–2002, 2003–2004, and 2005–2006, separately, based on NCHS recommendations. The formula used to estimate the GFR is as follows (Levey et al.): estimated GFR = 175 * (standardized serum creatinine in mg/dl)**(-1.154) * age**(-0.203) * (0.742 if female) * (1.212 if black).

The second method is based on cystatin C only (Stevens et al.), which is available only for NHANES III, NHANES 1999–2000, and NHANES 2001–2002: estimated GFR = 76.7 * cystatin C **(-1.19).

CKD is defined as an eGFR less than 60 ml/min/1.73 m2, or an eGFR ≥ 60 in the presence of microalbuminuria. CKD stages are defined as follows: Stage 5, eGFR < 15; Stage 4, 15 ≤ eGFR < 30; Stage 3, 30 ≤ eGFR < 60; Stage 2, ACR ≥ 30 and 60 ≤ eGFR ≤ 89; and Stage 1, ACR ≥ 30 and eGFR ≥ 90. These are the standard CKD definitions used in this chapter.

STATISTICAL ANALYSIS

To obtain national estimates of each statistic, odds ratios, sampling weights, and survey design are implemented by SUDAAN (Research Triangle Institute, Research Triangle Park, NC). Standard errors are estimated using the Taylor Series Linearization method for NHANES III and NHANES 1999–2006. GFR is estimated either by the MDRD (creatinine) method or by cystatin C, as indicated in the figure titles. CKD includes Stages 1–5; all other comorbidities are self-reported.

Table 1.f and Figures 1.12–15 present data on awareness, treatment, and control of metabolic markers. Patients are classified as hypertensive if measured systolic blood pressure (BP) is ≥ 140 mmHg (≥ 130 mmHg for CKD or diabetic patients) or measured diastolic BP is ≥ 90 mmHg (≥ 80 mmHg for CKD or diabetic patients), or if the patient self-reports currently taking a prescription to control hypertension. Patients are classified as being aware of hypertension if they report having been told they have high BP, are classified as being treated for hypertension if they report currently taking a prescription to control hypertension, and are considered in control of hypertension if current BP is < 140/< 90 (< 130/< 80 for CKD or diabetic patients).

Control of hypercholesterolemia is assessed in a similar way. Hypercholesterolemia is defined as a measured LDL cholesterol above the ATP III target range (≥ 160 mg/dl for patients with 0–1 risk factors, ≥ 130 mg/dl for patients with two or more risk factors, ≥ 100 mg/dl for patients with coronary heart disease (CHD) and CHD risk equivalents). CKD is classified as a CHD risk equivalent. Awareness of hypercholesterolemia is assessed by self-report of being told by a doctor that blood cholesterol level is high, and a patient is classified as being treated for hypercholesterolemia if he or she reports currently taking a cholesterol medication or dieting to control cholesterol. Control is defined as meeting the ATP III LDL target for the appropriate risk category, as described above. Current control of HDL cholesterol and total cholesterol are also presented here; awareness and treatment, however, are not assessed, since LDL cholesterol is currently the recommended target of therapy.

Diabetic patients are identified by self-report, as described above. Control of diabetes is assessed as a glycosylated hemoglobin (A1c) of less than 7 percent, as recommended by the American Diabetes Association.

In Figures 1.16–18, SAS survey procedures are used and survey design and weights are considered to calculate C-statistic, sensitivity, and specificity.

Renal function measures in adolescents
Chapter Two

The NHANES dataset is described in the methods for Chapter One. For Chapter Two, the population is restricted to participants age 12–17. Measurements are defined as in Chapter One, with the following additions.

Age is defined as the participant’s age at the time of the household interview, and grouped into ages 12–13, 14–15, and 16–17. Race/ethnicity is defined as non-Hispanic white, African American, Hispanic, and other.

Obesity is defined as a z-score for weight-to-height of 1.645 on the CDC growth reference table, 2000. The status of current pregnancy for females participants is based on self-report, while current menstruation is a response of “Having it now” to the question “When did you have your last menstrual period?” The glomerular filtration rate (GFR; ml/min/1.73 m2) is estimated by the Schwartz formula (Schwartz et al.), based on the standardized creatinine value for NHANES 1999–2000, 2001–2002, 2003–2004, and 2005–2006, separately, from NCHS recommendations. The formula used to estimate the GFR is as follows: estimated GFR = k * (height (cm)) / serum creatinine (mg/dl), where the constant k is 0.55 in children age 1–13, 0.70 in adolescent males (because of the presumed increase in male muscle mass), and 0.55 for adolescent females. For analytical methods, please refer the description of Chapter One, above. In Tables 2.b–d, a linear model is fitted to find the association between each risk factor and renal function, either by single predictor (unadjusted) or multivariates (adjusted by all variables in the table). The regression coefficient and its standard error are estimated. Predicted variables are eGFR in Table 2.c, cystatin C in Table 2.c, and log-transformed ACR in Table 2.d. In Table 2.e, a linear regression is fitted to find the impact of each renal function measurement on blood pressure and laboratory abnormalities, either by single renal function measurement (unadjusted) or adjusted by age, gender, and race/ethnicity.

CKD identified in the claims data
Chapter Three

Figure 3.1 illustrates the burden of new and prevalent diabetes, chronic obstructive pulmonary disease, and cardiovascular disease in the Medicare CKD population. Methods are those described for Figure 9.1. Table 3.a compares the characteristics of prevalent general Medicare, MarketScan, and Ingenix i3 CKD patients by age, gender, comorbidity, and occupation in 2007. Each comorbidity is defined by medical claims (one inpatient or two outpatient) during each calendar year; identification of hepatitis C also requires that the two outpatient claims be at least 30 days apart. Cardiovascular disease is defined as any of the following claims-defined diseases: atherosclerotic heart disease, congestive heart failure, cerebrovascular accident, dysrhythmia, other cardiovascular disease, or peripheral vascular disease. Maps in Figure 3.2 include 2007 patients, age 20–64, in the MarketScan and Ingenix i3 databases, while Figures 3.3–4 include 2007 incident and prevalent CKD patients, age 65 and older, in the general Medicare data. CKD is defined by medical claims (one inpatient or two outpatient) during each calendar year; ESRD patients are excluded. In Figures 3.5–12, the 5 percent Medicare sample includes patients who are age 65 and older, without ESRD, and who survive all of 2006 with Medicare as primary payor (and are not enrolled in Medicare Advantage). The MarketScan and Ingenix i3 cohorts are constructed in a similar fashion, but restricted to patients age 20–64, enrolled in a fee-for-service plan, and without ESRD. Patients with CKD in the prior year are excluded in definitions of incident CKD. Table 3.b and Figures 3.13–18 illustrate the percentage of prevalent CKD patients with different comorbidities, defined as in Table 3.a. For the general Medicare data, patients are age 65 and above, and have both inpatient/outpatient and physician/supplier coverage during the calendar year. MarketScan patients are age 20–64, with fee-for-service coverage during the calendar year. Ingenix i3 patients are age 20–64 and are under coverage with business type classified as commercial. Both hospital days and hospital admissions per patient year are assessed during the calendar year. All patients must survive to the end of each calendar year; ESRD patients are excluded. In Table 3.c and Figures 3.20–25, CKD is defined as follows: Stage 5: eGFR < 15; Stage 4: 15 ≤ eGFR < 30; Stage 3: 30 ≤ eGFR < 60; Stage 3–5: eGFR <60. Comorbidities such as hypertension, cardiovascular disease, chronic obstructive pulmonary disease, hepatitis C, cancer, anemia, liver disease, and hospitalization are defined from claims, and metabolic abnormalities are defined from laboratory test results. Figure 3.19 presents distribution of eGFR by CKD identification codes, including CKD defined using all codes, CKD stage diagnosis code (585.X), diabetes with renal manifestations (250.4), and hypertensive kidney disease with CKD (403.X1).

Care of patients with CKD
Chapter Four

Figure 4.1 shows the cumulative probability of non-CKD patients receiving a first urinary microalbumin measurement. The general Medicare population includes patients continuously enrolled in the Medicare inpatient/outpatient and physician/supplier program during 2006, age 65 or older at the beginning of the year. Patients are excluded if they are enrolled in a managed care program (HMO), acquire Medicare as secondary payor, die, are diagnosed with CKD or ESRD during 2006, have a missing date of birth, or do not live in the 50 states, the District of Columbia, Puerto Rico, or the Territories. Racial and ethnic categories are mutually exclusive.

Ingenix i3 patients are those continuously enrolled in a fee-for-service plan in 2006 and 2007, and age 50–64 during that year. Patients are excluded if they are diagnosed with CKD or ESRD during 2006.

For both populations, patients are followed from January 1 to December 31, 2007 for the first urinary microalbumin measurement. The Kaplan-Meier method is used to calculate the cumulative probability. Patients are censored at death, development of ESRD, and payor status change for the Medicare population, while patients are censored at development of ESRD for the Ingenix i3 population.

CPT codes used to define urinary microalbumin measurement are 82042, 82043, 82044, and 84156. Diabetes and hypertension are defined in 2006. Methods of defining CKD, diabetes, and hypertension are the same as those described for Chapter Six, below.

Figure 4.2 includes patients from the 5 percent Medicare sample, age 66 and older, who survive all of 2007 with Medicare as primary payor and are not enrolled in Medicare Advantage. Patients with CKD or ESRD in 2006 are excluded. The first CKD claim is identified in 2007 by the regular CKD diagnosis codes, excluding 584. The calendar year 2006 is used to define diabetes and cardiovascular disease by the standard method, and the Kaplan-Meier method is used to obtain the cumulative probability. The MarketScan and Ingenix i3 cohorts are constructed in a similar fashion, but restricted to patients age 50–64 who are enrolled in a fee-for-service plan. Figure 4.3 uses a similar cohort, but excludes only ESRD patients. The first nephrologist claim in 2007 is identified for Medicare patients from the physician specialty codes on physician/supplier claims, for MarketScan patients from provider codes on inpatient and outpatient claims, and for Ingenix i3 patients from provider category codes on inpatient, outpatient, or physician/suppler claims. Figure 4.4 is similar 4.3, but restricted to patients with CKD in 2006.

Figure 4.5 includes patients from the 5 percent Medicare sample, age 66 and older, who survive all of 2006 with Medicare as primary payor, are not enrolled in Medicare Advantage, and develop CKD in 2006; patients with CKD in 2005 or ESRD in 2006 are excluded. Calendar year 2006 is used to define CKD by the standard method, while calendar year 2005 is used to define diabetes and cardiovascular disease by the standard method; the Kaplan-Meier method is used to obtain the cumulative probability within one year of CKD diagnosis. MarketScan and Ingenix i3 cohorts are constructed in a similar fashion, but restricted to patients age 50–64 who are enrolled in a fee-for-service plan. The first nephrologist claim, primary care claim, or cardiology claim is identified for Medicare patients from the physician specialty codes on physician/supplier claims, for MarketScan patients from provider codes on inpatient and outpatient claims, and for Ingenix i3 patients from provider category codes on inpatient, outpatient, or physician/supplier claims. Constructed in a similar fashion, Figure 4.6 is restricted to 2006 CKD patients with diabetes, Figure 4.7 to 2006 CKD patients with cardiovascular disease, and Figure 4.8 to 2006 CKD patients with both diabetes and cardiovascular disease.

Figures 4.9–14 include CKD patients in 2002, 2004, and 2006; and show the cumulative probability of testing during one year. The 2007 cohort for Figures 4.9–12 and 4.14 is the same as that described for Figure 4.4; the cohort in Figure 4.13 is similar, but limited to diabetic CKD patients. Patients are followed from January 1, 2007 to December 31, 2007, and the Kaplan-Meier method is used to obtain the cumulative probability. Cohorts and follow-up are similar for 2003 and 2005. Tests are identified from outpatient and physician/supplier claims during the year, and identified as follows: creatinine testing, HCPCS codes 80048, 80050, 80053, 80069, and 82565; calcium/phosphorus testing, HCPCS codes 82310, 80048, 80050, 80053, 80069, and 84100; parathyroid hormone testing, HCPCS code 83970; lipid testing, HCPCS codes 80061, 82465, 83700, 83701, 83715, 83716, 83717, 83718, 83719, 83720,83721, and 84478; glycosylated hemoglobin testing, HCPCS codes 83036 and 83037; and hemoglobin testing, HCPCS codes 85013, 85014, 85018, 85025, 85027, 80050, and 80055.

Figures 4.15–22 include CKD patients in the entry periods of 2002, 2004, and 2006, and show the cumulative probability of medication use during the 12-month study periods of 2003, 2005, and 2007, separately. The study cohort includes MarketScan and Ingenix i3 patients age 20–64; MarketScan patients have fee-for-service coverage during the entry period and medical coverage and drug insurance during the study period, while Ingenix i3 patients have coverage with business type classified as commercial during both the entry and study periods. Both CKD and diabetes are defined by medical claims (one inpatient or two outpatient) during the entry period. CKD of Stages 3–5 is defined by the 585 stage code.

Figures 4.23–24 include 2007 CKD Ingenix i3 patients, age 50–64. Patients survive all of 2007, are enrolled in a fee-for-service plan for the entire year, and use a statin at least once during the year. Calendar year 2007 is used to define CKD, diabetes, and cardiovascular disease by the standard method. For CKD patients on a statin, the controlled total cholesterol is less than 200 mg/dl and the controlled LDL is less than 100 mg/dl. Figures 4.25–26 use a similar cohort, but include only 2007 CKD patients with diabetes and with at least one use of diabetic drugs in 2007. The controlled glycosylated hemoglobin level is less than 7 percent.

Morbidity & mortality
Chapter Five

Hospitalization

New to this year’s ADR, adjusted admission rates in this chapter include adjustment for baseline comorbidities and prior hospitalization in addition to patient demographics. A model-based adjustment method is used with a Poisson assumption, and includes data from the current and previous two years, with respective weights of 1, ¼, and ⅛. Adjusted rates reflect the distribution of a reference cohort: in this case, Medicare patients in 2005. With this method, the parameter estimates from the model are used to calculate an estimated admission rate for each patient in the reference cohort. Adjusted rates are then computed as the weighted average of these individual rates, using as the weight the time at risk of each patient in the reference cohort.

Figure 5.1 compares all-cause hospital admission rates in CKD and non-CKD patients in prevalent Medicare and MarketScan cohorts. The study design consists of a one-year period during which CKD, comorbidities, and prior hospitalization are defined from claims, followed by the cohort year when follow-up for admissions begins on January 1. The Medicare cohort includes patients age 66 and older on December 31 of the prior year, who are residents of the 50 states, the District of Columbia, Puerto Rico, or the Territories, are continuously enrolled in Medicare inpatient/outpatient and physician/supplier coverage, are without HMO coverage, are without ESRD, and who survive the complete year prior to follow-up. The MarketScan cohort includes patients age 50–64 on December 31 of the prior year who remain without ESRD and enrolled in a fee-for-service commercial health plan during the prior year. Patients are followed for admissions from January 1 of the follow-up year, and are censored at ESRD initiation, end of plan coverage, or December 31; Medicare patients are also censored at death. Rates are adjusted for gender, prior hospitalization, diabetes, COPD, hypertension, liver disease, gastrointestinal disease, cancer, anemia, and cardiovascular disease, which is defined by at least one of the following conditions: peripheral vascular disease, CVA/TIA, atherosclerotic heart disease, congestive heart failure, dysrhythmia, and other cardiac disease.

Table 5.a shows predictors of hospitalization in Medicare patients age 66 and older. Study design, censoring, and inclusion criteria generally follow those described above for the Medicare cohort in Figure 5.1, with the additional exclusion of patients with a bridge hospitalization spanning January 1. Groups for CKD, diabetes, and cardiovascular disease are mutually exclusive. Follow-up for first hospital admission starts on January 1 of 2003, 2005, and 2007, and Cox proportional hazards regression models are used. Adjustment factors include those listed for Table 5.a.

Figure 5.2 displays unadjusted and adjusted all-cause admission rates by CKD stage for prevalent Medicare patients, 2007, age 66 and older. Study design, censoring, and inclusion criteria generally follow the description for the Medicare cohort in Figure 5.1. Here, however, CKD patients without indication of CKD stage from 585 ICD-9-CM diagnosis codes are excluded. Follow-up begins on January 1, 2007. Admission rates are adjusted for age, gender, race, prior hospitalization, cardiovascular disease, diabetes, COPD, hypertension, liver disease, gastrointestinal disease, cancer, and anemia. Rates shown by hypertension, cardiovascular disease, or COPD, respectively, include the subgroup of patients with the comorbidity, and are adjusted for all factors except the respective comorbidity. Rates are adjusted using the model-based adjustment method described above, except that for CKD patients, due to the availability of CKD stage data, only 2007 data are used, without weighting the previous two years.

Figures 5.3–6 show adjusted all-cause and cause-specific admission rates by CKD status and dataset. Again, study design, censoring, and inclusion criteria generally follow the description for the Medicare and MarketScan cohorts in Figure 5.1. Additionally, Ingenix i3 data include point prevalent patients on January 1 of the year, continuously enrolled in a fee-for-service or commercial health plan and without ESRD during the prior year, and age 50–64 on December 31 of the prior year. The group labeled “CKD” includes those with claims-based evidence of CKD in the prior year, while “non-CKD” is defined as patients without claims-based evidence of CKD. For the CKD group identified from Ingenix i3 lab data, patients without at least one serum creatinine value during the prior year are excluded. The MDRD equation is used to compute eGFR, and an eGFR less than 60 ml/min/1.73 m2 defines CKD from lab data. One limitation of this application of the eGFR calculation is that race data are unavailable in the Ingenix i3 dataset; therefore, in the absence of race information, the eGFR equation assumes all patients are not African Americans. Rates are adjusted for gender, prior hospitalization, cardiovascular disease, diabetes, COPD, hypertension, liver disease, gastrointestinal disease, cancer, and anemia. Cause-specific rates reflect hospital admissions for the purpose of the stated condition, and are identified by principal ICD-9-CM diagnosis codes for cardiovascular and infectious admissions listed in the description of Figure 6.2 in Volume Two.

Figures 5.7–9 illustrate geographic variations in hospital admissions for pneumonia, bacteremia/septicemia, and urinary tract infection among Medicare CKD patients, point prevalent on January 1, 2007. Included patients are age 66 and older on December 31, 2006, and, during 2006, have a claims-based CKD diagnosis, are without ESRD, and are continuously enrolled in Medicare parts A and B, with no HMO coverage. Residents of Puerto Rico and the Territories are excluded. Follow-up begins on January 1, 2007, and unadjusted admission rates are presented by state. Cause-specific admissions are based on principal ICD-9-CM codes as follows: pneumonia, 480–486 and 487.0; bacteremia/septicemia, 038.0–038.9 and 790.7; and urinary tract infection, 590–590.9, 595–595.4, 597–597.89, 598, 599.0, 601–601.9, 604–604.9, 607.1–2, 608.0, 608.4, 616.1, 616.3–4, and 616.8.

Figures 5.10–18 present cause-specific hospital admission rates from various data sources. Again, for the Medicare (Figures 5.10–12) and MarketScan (Figures 5.13–15) cohorts, study design, censoring, and inclusion criteria follow the description for Figure 5.1. Ingenix i3 data (Figures 5.16–18) include point prevalent patients on January 1 of the year who, during the prior year, are age 50–64 on December 31, continuously enrolled in a fee-for-service or commercial health plan, and without ESRD. Admissions for pneumonia, bacteremia/septicemia, and urinary tract infection are identified by the principal ICD-9-CM diagnosis codes listed for Figures 5.7–9. In Figures 5.14–15, rates in 2002 for female CKD patients are not available due to insufficient events for model-based estimates. Figures 5.10–18 are comparable since they use the same adjustment factors and reference cohort. Rates are presented by gender and adjusted for prior hospitalization, cardiovascular disease, diabetes, COPD, hypertension, liver disease, gastrointestinal disease, cancer, and anemia.

Mortality

Figures 5.19–21 illustrate trends, by CKD status, in unadjusted and adjusted all-cause mortality from 1995 through 2007 by age, gender, and race, respectively. The study cohort for 1995 represents point prevalent Medicare patients on January 1, 1995, age 66 or older. CKD status is identified from 1994 Medicare claims, and the cohort excludes patients enrolled in an HMO, with Medicare as secondary payor, or diagnosed with ESRD in 1994. Follow-up extends from January 1, 1995 to December 31, 1995, and is censored at the ESRD date and the end of Medicare entitlement. Patients not living in the 50 states or the District of Columbia are excluded. Cohorts for 1996–2007 are constructed in a similar manner. Adjusted mortality is based on a Cox regression model and adjusted for demographics, hospitalization in the prior year, and comorbidities and sources of comorbidities defined in the prior year. Medicare patients from 2005 are used as the reference cohort.

Table 5.b shows adjusted relative risks for death in 2003, 2005, and 2007. The cohort definitions are same as those defined in Figures 5.19–21. A proportional hazards model is used to obtain the relative risks, and covariates include age, gender, race, and comorbidities. Reference groups are those age 66–69 at the beginning of each period, females, whites, and non-comorbid patients. Figure 5.22 is based on the results obtained from Table 5.b.

Stroke & mortality

In Figures 5.23–26 and Table 5.c, strokes are identified using ICD-9-CM codes 430.x–434.x and 436.x, and patients are required to have one inpatient claim or two outpatient claims with a diagnosis to be identified as having had a stroke. For Figures 5.23–24, prevalent CKD patients are identified for 2005, and only those without a stroke in 2005 are eligible for an incident stroke in 2006. For those without prevalent CKD in 2005 (and without a stroke in 2005), CKD and stroke are identified in 2006, and any time at risk for a stroke prior to a CKD diagnosis is attributed to the non-CKD group. For Figure 5.25 we use incident ESRD patients in 2006 without any stroke codes in the 12–24 months prior to initiation, and then look at claims starting 12 months prior to initiation. Figure 5.26 uses the same cohort as Figure 5.24.

Cardiovascular disease in CKD patients
Chapter Six

Figure 6.1 illustrates the percentage of patients with incident congestive heart failure (CHF) receiving an echocardiogram, nuclear imaging, or coronary angiography at or up to 90 days after CHF diagnosis. It uses a subset of the population in Table 6.a and Figure 6.2 (described below), restricted to Medicare patients with a first diagnosis of cardiovascular disease or a first procedure/device during the follow-up period in 2007. Any stress test includes stress echocardiograms, stress nuclear imaging, stress test, and stress electrocariograms (ECGs). Patients receiving these tests are identified through procedure codes using the same method described below. Follow-up for testing begins on the CHF diagnosis date and ends at the earliest of death, ESRD diagnosis, change of enrollment status, 90 days after CHF diagnosis, or December 31, 2007. The percentage of patients receiving each test is calculated as the number of patients tested during the follow-up period divided by the total number of patients at the beginning of follow-up. Codes used to define these tests are as follows:

• resting echocardiogram: 93303, 93304, 93307, 93308, 93312-93315, 93317, and 93318 (CPT codes)
• stress echocardiograms: 93350 (CPT codes)
• resting nuclear imaging: 78468, 78472, and 78494 (CPT codes)
• stress nuclear imaging: 78459, 78460, 78461, 78464, 78465, 78472, 78473, 78478, 78480, 78481, 78483, 78491, and 78492 (CPT codes)
• stress test: 89.41–89.44 (ICD-9-CM procedure codes)
• stress ECGs: 93015–93018 (CPT codes)
• coronary angiography and/or catheterization: 37.22–37.23 and 88.53–88.57 (ICD-9-CM procedure codes); 93508, 93510, 93511, 93524, 93526, 93527, 93529, 93531–93533, 93539, 93540, 93543, 93545, and 93555 (CPT codes)

Tables 6.a–b and Figure 6.2 present information on the risk of incident CHF, cerebrovascular accident/transient ischemic attack (CVA/TIA), peripheral arterial disease (PAD), cardiac arrest, acute myocardial infarction (AMI), first percutaneous coronary interventions (PCI), coronary artery bypass graft surgery (CABG), and use of implantable cardioverter defibrillators and cardiac resynchronization therapy with defibrillator (ICD/CRT-D), by CKD stage, in both Medicare and Ingenix i3 patients. Medicare patients include point prevalent patients on December 31, 2006, age 66 and older, residing in the 50 states, the District of Columbia, Puerto Rico, or the Territories, continuously enrolled in Medicare inpatient/outpatient & physician/supplier coverage, and not enrolled in an HMO in 2006 (baseline). Ingenix i3 patients include point prevalent patients on December 31, 2006, age 20–64, who are continuously enrolled in a fee-for-service commercial health plan in 2006. Patients diagnosed with ESRD prior to January 1, 2007, or with changed coverage on January 1, 2007, are excluded. In estimating the risk of incident CHF in 2007, patients diagnosed with CHF at baseline are also excluded. Similar exclusion criteria are applied for each of the other events/procedures. CKD patients and CKD stage are defined using the methods described for Chapter Three.

To estimate the incident CHF rate in Medicare patients, patients without CHF at baseline are followed from January 1, 2007, to the earliest of CHF diagnosis date, death, ESRD diagnosis, change of enrollment status, or December 31, 2007. For patients in the Ingenix i3 dataset, follow-up ends on the earliest of CHF diagnosis date, ESRD diagnosis, change of enrollment status, or September 30, 2007. The same method is used for each of the other endpoints. Figure 6.2 presents unadjusted incident rates for each endpoint, by CKD stage. Event and procedure rates are calculated as the total number of patients having an event or procedure divided by the total follow-up time. Table 6.a presents the hazards ratio of each endpoint, by age, gender, race, and CKD stage, for patients in Medicare 5 percent sample data, while Table 6.b presents the same information, by age, gender, and CKD stage, for patients in the Ingenix i3 dataset. The Cox proportional hazards model is used to estimate the hazards ratio. Comorbid conditions at baseline are also included in the model.

Different data sources, codes, and methods are used to identify patients with each cardiovascular disease at baseline and those who had a first diagnosis of cardiovascular disease during the follow-up period, while the same data sources, codes, and methods are used for identifying procedures. According to a previously validated method for using Medicare claims to identify diabetic patients, a patient is diabetic if, within a one-year observation period, he or she has an ICD-9-CM diagnosis code of diabetes on one or more inpatient institutional claims (inpatient hospitalization, skilled nursing facility, or home health agency), or two or more outpatient institutional claims or physician/supplier claims (referred to as the standard method in this chapter). Using this standard method, we identify patients with CHF, CVA/TIA, and AMI at baseline. This method, however, is not used to identify patients with cardiac arrest. A patient with cardiac arrest at baseline is identified through an ICD-9-CM diagnosis code of cardiac arrest on a claim from either inpatient/outpatient institutional claims or physician/supplier claims. ICD/CRT-D is defined through ICD-9-CM procedure codes in inpatient/outpatient claims. PCI and CABG are identified through ICD-9-CM procedure codes in inpatient/outpatient claims or CPT codes in physician/supplier claims. PAD is defined through either diagnosis codes or procedure codes. If PAD is defined through diagnosis codes, we use the standard method. If PAD is defined through procedure codes, we use the same method used for PCI and CABG.

Cardiovascular events of CHF, CVA/TIA, PAD, cardiac arrest, AMI, first PCI and CABG surgery, and first use of ICD/CRT-D are defined as the date of the first appearance of diagnosis or procedure codes in the claims during the follow-up period. For each event except AMI, the data sources and methods used to define the event are the same as those used in defining the condition at baseline. The event of AMI is defined as the first appearance of the diagnosis code on an inpatient claim. The same codes are used to define PAD, PCI, CABG, and ICD/CRT-D at baseline and during follow-up, while different codes are used for CHF, CVA/TIA and AMI. Codes used to identify cardiovascular disease, procedures, and device use are as follows:

• CHF: 398.91, 422.xx, 425.x, 428.xx, 402.x1, 404.x1, 404.x3, and V42.1 for condition at baseline (ICD-9-CM diagnosis codes); 398.91, 425.x, 428.xx, 402.x1, 404.x1, and 404.x3 for event during follow-up (ICD-9-CM diagnosis codes)
• PAD: 440–444, 447, and 557 (ICD-9-CM diagnosis codes); 84.0, 84.1, 84.91, 39.25, 39.26, and 39.29 (ICD-9-CM procedure codes); 24900, 24920, 25900, 25905, 25920, 25927, 27295, 27590, 27591, 27592, 27598, 27880, 27881, 27882, 27888, 27889, 28800, 28805, 34900, 35131, 35132, 35141, 35142, 35151, 35152, 34051, 34151, 34201, 34203, 34800–34834, 35081–35103, 35331, 35341, 35351, 35355, 35361, 35363, 35371, 35372, 35381, 35450, 35452, 35454, 35456, 35459, 35470, 35471, 35472, 35473, 35474, 35480, 35481, 35482,35483, 35485, 35490, 35491, 35492, 35493, 35495, 35521, 35531, 35533, 35541, 35546, 35548, 35549, 35551, 35556, 35558, 35563, 35565, 35566, 35571, 35583, 35585, 35587, 35621, 35623, 35646, 35647, 35651, 35654, 35656, 35661, 35663, 35665, 35666, and 35671 (CPT codes)
• CVA/TIA: 430–438 for condition at baseline; 430–437 for event during follow-up (ICD-9-CM diagnosis codes) cardiac arrest: 427.4 and 427.5 (ICD-9-CM diagnosis codes)
• AMI, 410 and 412 for condition at baseline; 410, 410.X0, and 410.X1 for event during follow-up (ICD-9-CM diagnosis codes)
• PCI: 00.66, 36.01, 36.02, 36.05, and 36.06 (ICD-9-CM procedure codes); 92980–92982, 92984, and 92995–92996 (CPT codes)
• CABG surgery: 36.1x (ICD-9-CM procedure codes); 33510–33523 and 33533–33536 (CPT codes)
• ICD: 37.94 (ICD-9-CM procedure code)
• CRT-D: 00.51 (ICD-9-CM procedure code)

Figures 6.3–10 shows geographic variations in incident rates for each cardiovascular disease/procedure. For the 2007 maps, the population is a subset of the population used in Table 6.a and Figure 6.2, restricted to those diagnosed with CKD in 2006 and residing in the 50 states and the District of Columbia. Rates are incident events or first procedures per 1,000 patient years in 2007. Similar methods are applied to identify point prevalent Medicare patients with CKD on December 31, 1996, to evaluate the event/procedure rate in 1997. For ICD/CRT-D, point prevalent Medicare patients with CKD on December 31, 2002, are used because the ICD-9-CM procedure code for CRT-D was not available prior to January 1, 2002.

Figures 6.11–13 use a subset of the population in Table 6.a and Figure 6.2, restricted to Medicare patients who had a first diagnosis of cardiovascular disease or received a first procedure/device during the follow-up period in 2007. These figures evaluate diagnostic testing, survival, physician care, and per person per month costs after the first cardiovascular event or procedure during the follow-up period in 2007. CKD patients and CKD stage are defined through ICD-9-CM diagnosis codes on inpatient/outpatient or physician/supplier claims from January 1, 2006, to the date of first cardiovascular disease diagnosis, procedure, or device use in 2007, using the same method described for Chapter Three.

Figure 6.11 presents survival, by CKD stage, after each cardiovascular event or procedure. For survival after CHF, incident CHF patients in 2007 are included. Follow-up begins on the CHF diagnosis date and ends on the earliest of death, ESRD diagnosis, or December 31, 2007. Using the model-based adjustment method (described in the section on statistical methods in the Volume Two appendix), survival probabilities are estimated with the Cox proportional hazards model, and adjusted for age. The reference group includes point prevalent Medicare patients on December 31, 2006, who are age 66 and older; residing in the 50 states, the District of Columbia, Puerto Rico, or the Territories; continuously enrolled in Medicare inpatient/outpatient and physician/supplier coverage; and not enrolled in an HMO in 2006. Patients diagnosed with ESRD prior to January 1, 2007, or with changed coverage on January 1, 2007, are excluded.

Figure 6.12 illustrates physician care of CKD patients during and after hospitalization for each cardiovascular event and procedure. To evaluate physician care during and after hospitalization for CHF, included patients are those whose first diagnoses of CHF in 2007 are from inpatient claims, and who are diagnosed with CKD at the time of hospital admission, discharged alive, not diagnosed with ESRD, and do not have a change in enrollment status at the time of discharge. The time period for evaluating care after hospitalization begins on the day after discharge and ends on the earliest of death, ESRD diagnosis, change of enrollment status, 90 days after discharge, or December 31, 2007. A similar method is used for the other events or procedures. For first PCI and first use of ICD/CRT-D, patients with first claims from outpatient institutional claims or physician/supplier claims are also included in the analysis. For these patients, the time period for evaluating physician care after the procedure begins on the day after the claim-from date on the outpatient or physician/supplier claims and ends on the earliest of death, ESRD diagnosis, change of enrollment status, 90 days after the procedure, or December 31, 2007. Patients receiving primary, cardiology, and nephrology care during and up to 90 days after hospitalization are examined to identify those with a first diagnosis of CHF or AMI, receiving their first PCI or CABG surgery, or receiving their first ICD/CRT-D device. Neurology is added for patients with CVA/TIA and cardiac arrest, while radiology and vascular surgery are added for those with PAD. The physician specialty codes on physician/supplier claims are used to identify primary care (01, general practice; 08, family practice; and 11, internal medicine), cardiology (06), nephrology (39), neurology (13), radiology (94), and vascular surgery (77).

Figure 6.13 presents, by CKD stage and for Medicare patients, per person per month (PPPM) inpatient/outpatient and physician/supplier expenditures related to cardiovascular diagnoses and procedures. The time period for calculating the cost begins on the first cardiovascular diagnosis or procedure date and ends on the earliest of death, ESRD diagnosis, change of enrollment status, or December 31, 2007. The actual Medicare PPPM payment is calculated by dividing the total Medicare payment in the follow-up period by the total follow-up time.

Transition to ESRD
Chapter Seven

Figures 7.2–6 include incident ESRD patients (Medicare patients are limited to those age 67 and older). For Figures 7.3–6, physician specialty is identified from claims; physician visits in Figure 7.3 include those to a primary care physician, cardiologist, or nephrologist, while in Figure 7.6 primary care represents family practice, general practice, and internal medicine. Inpatient and outpatient locations are identified by location code or the source of the claim, depending on the dataset.

Figures 7.7–12 include incident ESRD patients in 2003, 2005, and 2007, and show the percentage of patients with at least one test during the eight quarters before the first ESRD service date. Tests are identified from outpatient and physician/supplier claims during the two years, as follows: parathyroid hormone testing, HCPCS code 83970; creatinine testing, HCPCS codes 80048, 80050, 80053, 80069, and 82565; lipid testing, HCPCS codes 80061, 82465, 83715, 83716, 83717, 83718, 83719, 83720,83721, and 84478; hemoglobin testing, HCPCS codes 85013, 85014, 85018, 85025, 85027, 80050, and 80055; and glycosylated hemoglobin testing, HCPCS codes 83036 and 83037. For glycosylated hemoglobin testing, patients must be defined with diabetes during the 24 months before incident ESRD. The ESRD cohort includes patients age 67 and older; the MarketScan cohort includes all ESRD patients with fee-for-service coverage during the study period, and the Ingenix i3 cohort includes all ESRD patients under coverage with business type classified as commercial.

Figures 7.13–20 show the percentage of patients on specific drugs during the eight quarters prior to and the one quarter after ESRD initiation. The cohort includes MarketScan and Ingenix i3 incident ESRD patients age 20–64. MarketScan patients have fee-for-service coverage and drug insurance during the nine quarters, while Ingenix i3 patients have coverage with business type classified as commercial during the same period.

Table 7.a and Figures 7.21–23 illustrate the percentage of patients on specific drugs prior to and after ESRD initiation, and uses the same study cohort as Figures 7.13–20. Figure 7.23 shows medication use in the quarter after ESRD initiation in patients using the medication three quarters prior to ESRD. All diabetic drugs (insulin, sulfonylureas, and potassium-sparing diuretics) are analyzed for ESRD patients with diabetes, which is defined by medical claims (one inpatient or two outpatient) during the year of ESRD incidence.

Figures 7.24–26 include ESRD patients who have Medicare as primary payor coverage (Medicare) or are eligible (MarketScan and Ingenix i3) for two years prior to ESRD; Figure 7.27 is limited to Medicare patients, because data on dialysis type is not available in the other datasets. Medicare patients include those age 67 or older at initiation. Methods used to identify vascular access insertions and to calculate rates are the same as in those used in Volume Two, Chapter Five. Vascular access procedures are obtained from Part B revenue line item files in the Medicare population and from physician’s claims in the Ingenix I3 and Marketscan data. Data from Part B and physician claims files can reflect procedures done in either the inpatient or outpatient setting

Acute kidney injury
Chapter Eight

In this chapter, patients with an acute kidney injury (AKI) hospitalization are identified from inpatient claims by the presence of ICD-9-CM code 584.x. AKI that requires dialysis (AKI-D) is identified by the additional presence of any of the following: ICD-9-CM procedure codes 39.95 and 54.98; ICD-9-CM diagnosis codes V45.1, V56.0, and V56.1; CPT codes 90935, 90937, 90945, and 90947; and revenue codes 0800–0809. Patients with ESRD diagnosed before the AKI hospitalization discharge are omitted, except as indicated. For patients with multiple AKI hospitalizations through the years, the first one in the timeframe is counted. The event rate is estimated as the number of events per 1,000 patient years at risk.

Figure 8.1 displays the percentage of patients hospitalized for AKI or AKI-D in a given year. The cohort includes general Medicare patients age 66 or older on December 31 of the cohort year, continuously enrolled in Medicare inpatient/outpatient and physician/supplier coverage, with no HMO coverage, and who survive and are without ESRD in the cohort year.

Characteristics of AKI patients

Figure 8.2, Table 8.a, and Figures 8.5–6 describe the demographic characteristics of patients suffering AKI. The study cohort includes the general Medicare patients described for Figure 8.1 (Figure 8.2 uses the 2007 cohort), along with MarketScan and Ingenix i3 patients age 20–64 on December 31 of the cohort year who are enrolled in a fee-for-service plan. Contrast is received during the two weeks prior to AKI hospitalization, and ACE-Is/ARBs and statins are taken in the three months prior to AKI hospitalization

Figures 8.3–4 use the same cohort describe for Figure 8.1. Figure 8.3 shows the type of dialysis used by hospitalized AKI-D patients. Modality is defined as follows: peritoneal dialysis, CPT codes 90945 or 90947 and 49420; continuous venous to venous hemodialysis (CVVHD), dialysis with CPT codes 90945 or 90947 but without 49420; intermittent hemodialysis (IHD), dialysis with CPT codes 90935 or 90937 and intermittent in the first three days; and daily hemodialysis (DHD), dialysis with CPT codes 90935 or 90937 and with three consecutive dialysis sessions in the first three days. To define modality, we first determine if there is any peritoneal dialysis during the period of the AKI event, and then look for continuous dialysis to identify hemodialysis or DHD. Those who are not identified by the above methods are categorized as having an unknown dialysis type. Figure 8.4 shows the percentage of hospitalized AKI patients who receive contrast in the two weeks prior to AKI admission.

Overall rates of acute kidney injury

Figure 8.7 presents rates of acute kidney injury, and Figure 8.b shows the adjusted hazard ratios for AKI hospitalization, adjusted for age, gender, and race. The study cohort for both includes point prevalent general Medicare patients on January 1, 2007, age 66 and older on December 31, 2007, along with point prevalent MarketScan and Ingenix i3 patients age 20–64 on December 31, 2007. Patients with ESRD before January 1, 2007, are excluded. Each patient is followed from January 1, 2007, to the earliest of death (for Medicare patients only), ESRD diagnosis, change of enrollment, or December 31, 2007.

Figure 8.8 presents the rate of AKI for patients with CKD at the entry period, Figure 8.9 the AKI rate for patients without CKD at the entry period, and Table 8.c the adjusted hazard ratios for AKI hospitalization, adjusted for age, gender, race, and comorbidities (anemia, ASHD, other cardiac disease, CHF, CKD, CVA/TIA, PVD, COPD, gastrointestinal disease, liver disease, dysrhythmia, cancer, hypertension, and diabetes), defined in the entry year 2006. The cohort for each includes general Medicare patients (described as for Figure 8.1) and MarketScan and Ingenix i3 patients, 2006. Each patient is followed from January 1, 2007 to the earliest of death (for Medicare patients only), ESRD diagnosis, change of enrollment, or December 31, 2007. CKD is identified with one or more Part A institutional claims (inpatient hospitalization, skilled nursing facility, or home health agency), or two or more Part A institutional claims (outpatient) or physician/supplier claims, with codes as follows: 016.0, 095.4, 189.0, 189.9, 223.0, 236.91, 250.4, 271.4, 274.1, 283.11, 403.x1, 404.x2, 404.x3, 440.1, 442.1, 447.3, 572.4, 580–583, 585-588, 591, 642.1, 646.2, 753.12–753.17, 753.19, 753.2, and 794.4. The codes for the remaining comorbidities are the same as those indicated elsewhere.

Figures 8.10–11 illustrate geographic variations in unadjusted rates of AKI and AKI-D in 2007 for general Medicare patients. The cohort is constructed as in Figure 8.7, but is restricted to those residing in the 50 states and the District of Columbia.

Patient care & outcomes

The methods for identifying the types of physician visits in Figures 8.12 and 8.15 are the same as those described in the methods for Chapter Seven. In Figure 8.15, multiple physician claims for the same specialty during the same inpatient stay are counted only once.

Testing in Figures 8.13–14 is identified as follows: creatinine testing, HCPCS codes 80048, 80050, 80053, 80069, and 82565; microalbumin testing: CPT codes 82042, 82043, 82044, and 84156.

Figure 8.16–17 examine the use of ACE-Is/ARBs and statins before and after AKI hospitalization in both AKI and AKI-D patients, and include 2007 Ingenix i3 patients, identified as in Figure 8.2. Medication treatment is identified in the three months before and after admission for AKI.

Figure 8.18 demonstrates the probability of patients receiving follow-up testing after hospitalization for AKI. Testing includes glycosylated hemoglobin (A1c) testing and eye examinations (diabetic patients only), along with influenza vaccinations. Patients with AKI are identified from the Ingenix i3 dataset in 2006, and are followed from the AKI admission until ESRD diagnoses, end of fee-for-service coverage, or one year after AKI admission. Diabetic status is determined from the one-year entry period before the AKI admission. Codes for testing are as follows: A1c tests, CPT code 83036 (at least two tests); eye examinations, CPT codes 92002, 92004, 92012, 92014, 92018, 92019, 92225, 92226, 92230, 92235, 92240, 92250, 92260, 92287, 67101, 67105, 67107, 67108, 67110, 67112, 67141, 67145, 67208, 67210, 67218, 67227, and 67228; and influenza vaccinations, CPT codes 90724, 90737, 90645, 90646, 90647, 90648, 90655, 90656, 90657, 90658, 90659, and 90660. The Kaplan-Meier method is used to calculate the cumulative unadjusted probability of testing during the one-year follow-up period.

Figures 8.19–20 display outcomes among survivors of an episode of hospitalized AKI, by AKI and CKD status. General Medicare patients age 66 and older on December 31, 2006, continuously enrolled in Medicare coverage, and without HMO coverage are included, and we identify administrative claims for AKI in 2006. CKD is defined by searching claims in the one year prior to the AKI admission. Patients are followed from AKI discharge (for those with AKI) or January 1, 2007 (for those without AKI), to the earliest of death, ESRD diagnosis, end of Medicare coverage, or one year after AKI discharge. The Kaplan-Meier method is used to estimate the probability of death in Figure 8.19 and the probability of ESRD in Figure 9.20.

Figures 8.21–22 present unadjusted rates and hazard ratios of ESRD and death by AKI and CKD status, using the same cohort as in Figures 8.19–20. Patients are followed from AKI admission (for those with AKI) or January 1, 2007 (for those without) to the earliest of death, ESRD diagnosis, end of Medicare coverage, or one year after AKI admission. Hazard ratios are estimated using Cox proportional hazard models, adjusted for age, gender, and race.

Table 8.d includes all patients in the 5 percent Medicare sample who are age 66 and older and who have hospitalized AKI events in 2006 or 2007. Hospitalized AKI is defined by ICD-9-CM diagnosis code 584 in inpatient claims, and discharge status in these claims is used to determine death status during hospitalization. The previous year of each AKI event is used to define diabetes, hypertension, and CKD. A logistic model is used to obtain the odds ratios of in-hospital death, adjusting for age, gender, diabetes, hypertension, and CKD (and, in the Medicare cohort, for race). The MarketScan cohort is constructed in a similar fashion, but restricted to patients age 20–64 who are enrolled in a fee-for-service plan.

Table 8.e includes all patients in the 5 percent Medicare sample who are age 66 and older and who have hospitalized AKI events in 2005 or 2006. Hospitalized AKI is again defined by ICD-9-CM diagnosis code 584 in inpatient claims, and the previous year of each AKI event is used to define diabetes, hypertension, and CKD. Patients are followed from the discharge date until the earliest of one year, ESRD, death, December 31, 2007, or loss of insurance coverage. The Cox model is used to obtain hazard ratios of death after discharge, adjusting for age, gender, race, diabetes, hypertension, and CKD.

The Medicare cohort in Table 8.f is same as that used in Table 8.e. The MarketScan and Ingenix i3 cohorts are constructed in a similar fashion, but restricted to patients age 20–64 who are enrolled in a fee-for-service plan. The Cox model is used to obtain hazard ratios of ESRD after discharge, with the same adjustments used in Table 8.d.

The cohorts and adjustments in Table 8.g are same as those used in Table 8f, and the Cox model is used to obtain hazard ratios of recurrent AKI hospitalization after discharge. Recurrent AKI events are defined as those more than 30 days after the first AKI event.

Table 8.h is similar to Table 8.d, but includes only hospitalized AKI patients with dialysis. Hospitalized AKI is defined as in Table 8.d, while dialysis is identified by any of the following codes: CPT codes 90935, 90937, 90945, and 90947; ICD-9-CM procedure codes 39.95 and 54.98; revenue codes 0800–0809; and ICD-9-CM diagnosis codes V45.1, V56.0, and V56.1. Table 8.i is similar to Table 8.e, Table 8.j is similar to Table 8.f, and Table 8.k is similar to Table 8.g. Tables 8.i–k, however, include only hospitalized AKI patients with dialysis.

Costs of CKD
Chapter Nine

Populations

Using the methodology described below, Figure 9.1 compares populations and costs in 2007 for Medicare patients (based on the 5 percent Medicare sample) and MarketScan patients, while Figures 9.2–7, 9.12–13, and Table 9.a compare costs for these populations.

Figures 9.8–11 are based on the Medicare and MarketScan incident 2006 ESRD populations during the transition to ESRD. Medicare patients here are age 67 and older, with Medicare as primary payor for the entire transition period (six months before through six months after the initiation of renal replacement therapy), and not enrolled in a managed care program (HMO) during the transition period. The MarketScan patients include those younger than 65 and continuously enrolled in a fee-for-service plan for the entire transition period.

The general Medicare population includes persons age 65 and older who survive for at least the last three months of year one, are continuously enrolled in Medicare inpatient/outpatient and physician/supplier coverage for this time period, are not enrolled in an HMO, and do not have ESRD during year one. Costs for this portion of the cohort are aggregated for year two, with censoring at the earliest of death, development of ESRD, change in payor status, or the end of year two. In addition, the cohorts include those who survive at least three months during year two, are enrolled at least three months during the year in Medicare inpatient/outpatient and physician/supplier coverage, are not enrolled in an HMO, and do not have ESRD during year two. Costs are also aggregated for year two for this portion of the cohort.

Important comorbidities (diabetes mellitus, chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD), and cardiovascular disease (CVD)) are determined for these cohorts from Medicare claims using a previously validated method. A patient is defined as having one of these comorbidities if, within the one-year observation period (year one or year two), he or she has a qualifying ICD-9-CM diagnosis code on one or more institutional claims (inpatient, skilled nursing facility, or home health agency) or two or more institutional outpatient claims and/or physician/supplier claims. Qualifying diagnosis codes are as follows: diabetes, 250.xx, 357.2, 362.0x, and 366.41; CKD, 016.0, 095.4, 189.0, 189.9, 223.0, 236.91, 250.4, 271.4, 274.1, 283.11, 403.x1, 404.x2, 404.x3, 440.1, 442.1, 447.3, 572.4, 580–588, 591, 642.1, 646.2, 753.12–753.17, 753.19, 753.2, and 794.4; COPD, 491–494, 496, 510; and CVD, 398.91, 402.x1, 404.x1, 404.x3, 410–414, 420–421, 422.xx, 423–424, 425.x, 426–427, 428.xx, 429, 430–438, 440–444, 447, 451–453, 557, 785.0–785.3, V42.1, V42.2, V43.3, V45.0, V45.81, V45.82, and V53.3. Costs are presented for the 1992–1993 through 2006–2007 cohorts. The cost year is always year two of the cohort.

The MarketScan population includes patients age 50–64, and is constructed in the same fashion as that described for the Medicare population, requiring continuous enrollment in a fee-for-service health plan. Patients identified as having ESRD are excluded. The cohorts are from 1999–2000 to 2006–2007.

Cost categories

Costs are categorized in several ways. For Figures 9.1–5, costs are simply total claims-based expenditures, while those in 9.6–11 are claims-based expenditures expressed per person per month (PPPM). Costs for Figures 9.12–13 are defined by the type of claim — inpatient, outpatient, and physician/supplier — and expressed as PPPM expenditures. MarketScan also has a separate claim set for drug claims, and these claims are included in the outpatient category. Costs are further broken down for Table 9.a, using diagnosis-related groupings (DRGs) for inpatient claims; revenue codes, current procedural terminology (CPT) codes, and healthcare common procedure coding system (HCPCS) codes for outpatient claims; and CPT, HCPCS, provider specialty, and place of service codes for physician/supplier claims.