 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
CLINICAL SCIENCE |
Division of Kidney, Urologic and Hematologic Diseases, National Institute of Diabetes, Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland; and Division of Renal Diseases and Hypertension, Department of Medicine, George Washington University Medical Center, Washington, DC.
Correspondence to Dr. Paul L. Kimmel, Division of Renal Diseases and Hypertension, Department of Medicine, George Washington University Medical Center, 2150 Pennsylvania Ave NW, Washington, DC 20037. Phone: 202-741-2283; Fax: 202-741-2285; E-mail: pkimmel{at}mfa.gwu.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As the number of US patients living with AIDS increased 57% from 214,711 in 1995 to 337,017 in 2000, and the number of incident ESRD patients increased 29.9% from 72,827 to 94,602, the number of incident HIV-infected patients increased only by 3.5%, from 1133 to 1171. Over this time, the percentage of incident ESRD patients with HIV infection fell from 1.56% to 1.24%. Among black men 25 to 44 yr of age, HIV infection as a proportion of incident ESRD cases fell from 8.5% to 6.2% from 1995 to 2000. The incident rate per million of AIDS or HIV infection in black men aged 25 to 44 fell from 107 in 1995 to 78 per million in 2000. The incidence rate for HIV-infected women in the ESRD program rose 14% while it declined 7% in men. Almost 2000 HIV-infected women, or 28.8% of the population, have initiated therapy for ESRD with hemodialysis. The number of prevalent cases increased in absolute numbers 81.3% from 2687 to 4871 (0.90% to 1.16% of the ESRD program). One-year survival rates for HIV-infected incident ESRD patients increased from 53.1% to 67.1% from 1995 to 2000.
Although these values may be underestimates because of underreporting due to confidentiality concerns and lack of biopsy confirmation, we conclude that although the prevalence of HIV infection is increasing in the US ESRD population, the increase as a proportion of the program is minimal and is due to better survival after development of renal failure. The incidence of HIV infection in the US ESRD program is stable. Highly active antiretroviral therapy may be responsible for the change in epidemiology of HIV infection in the US ESRD program.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In 1996, an editorial by Winston and Klotman (8) that used data from the Health Care Finance Administration (HCFA) warned of the possibility of an epidemic of HIV infection in the US ESRD program. Highly active antiretroviral therapy (HAART) was initiated shortly thereafter, in 1996, with the introduction of the protease inhibitor indinavir in the United States (10). Since the availability of a variety of protease inhibitors and the widespread use of HAART, the prognosis of HIV-infected patients has improved (11), and case reports have indicated therapy can dramatically change the course of classic HIV-associated nephropathy (HIVAN) (12,13). Several studies from different groups have suggested the prognosis of HIV-infected ESRD patients treated with HAART has improved (14–18).
We used data from the United States Renal Data System (USRDS) to determine what the incidence and prevalence of HIV infection and AIDS-associated renal diseases were in the US ESRD program from 1995 to 2000.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
First, since 1983, the HCFA 2728 Medical Evidence Form has included AIDS as a cause of renal failure. This form was revised in 1995 to include comorbid conditions, among which are AIDS and HIV infection. For the calculation of incidence rates, persons were designated as having HIV infection or AIDS (HIV/AIDS) if the principal cause of ESRD was AIDS and/or there was an indication of AIDS or HIV as a comorbid condition. The addition of the HIV/AIDS comorbidity indicator increases the number of patients included by about 20%. Incidence as defined by a principal diagnosis of HIV/AIDS and/or HIV/AIDS as a comorbid indicator was defined as method 1.
It is well known that there can be underreporting of HIV. For example, on the Medical Evidence Form for comorbid conditions, "cannot disclose" is reported for about 70% of cases for AIDS and HIV. Thus, it is likely that the 2728 form will underreport HIV-related renal diseases. To help counter this downward bias, we examined the Medicare billing data for the AIDS ICD9 code (042). By means of a method previously validated by Hebert et al. (19), we assigned an AIDS/HIV code to persons with at least one hospitalization with an AIDS or HIV code or at least two outpatient encounters with an AIDS or HIV code. The latter is done to avoid "rule-out" diagnoses. To estimate prevalence rates of HIV infection and/or AIDS, we used both the 2728 definition (method 1) as well as a more expanded definition that used the billing data (method 2). Thus, for method 2, the prevalence estimates are based on the combination of method 1 and billing data. A person could be identified as having HIV infection or AIDS through either mechanism. The intent of analyses that used method 2, examining the billing data, is to identify the largest number of HIV/AIDS patients.
This method will identify the maximum number of patients with HIV infection or AIDS in the ESRD database. However, it is also true that not all ESRD patients in the USRDS database will have billing data. First, for persons who are not Medicare entitled at the time of renal failure, there is a 90 d waiting period until Medicare coverage begins. Second, there is also no bill data for Medicare beneficiaries who are enrolled in health maintenance organizations. Most importantly, however, is the effect of legislative provisions for secondary payers. Secondary payer means that if a person has other insurance, that other insurer pays for care, and Medicare is the secondary payer. Originally, this provision extended for 12 mo after the start of ESRD. It was subsequently extended to 18 mo and then 30 mo. Beginning in 1997, the period was made indefinite. As a result, Medicare is the primary payer for a decreasing percentage of patients, 74% in 1995 and 63% in 2000. Thus, there will be an unknown number of potential patients with HIV/AIDS that cannot be identified either through HCFA 2728 or billing data.
Regional variation in the incidence and prevalence of ESRD associated with HIV/AIDS was examined by identifying high-impact AIDS Metropolitan Statistical Areas (MSA) from the CDC public use data site (http://wonder.cdc.gov/). Fifteen MSA (Atlanta, GA; Baltimore, MD; Boston, MA; Chicago, IL; Dallas, TX; Fort Lauderdale, FL; Houston, TX; Los Angeles, CA; Miami, FL; New York, NY; Newark, NJ; Philadelphia PA; San Diego, CA; San Francisco, CA; and Washington, DC), accounting for half of all AIDS cases since 1980, were selected.
Mortality was calculated with 90 d after the date of renal failure used as the start date, per the USRDS methodology. A standard life table method was used for time until death with censoring at July 1, 2001. All calculations were performed by SAS software, version 8.2.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
|
When incidence in the ESRD program is categorized as diabetic disease, hypertensive disease, glomerulonephritis, polycystic kidney disease, other, unknown and missing, AIDS- and HIV-associated renal diseases are the fifth leading discrete cause of chronic kidney disease (method 1) (Figure 7). Renal disease with AIDS or HIV infection accounted for a cumulative 1.3% of all incident ESRD from 1995 to 2000 (Figure 7). Renal disease with AIDS or HIV infection accounted for 3.9% of all incident ESRD in African Americans from 1995 to 2000 (method 1). Renal disease with AIDS or HIV infection accounted for 18% of all incident ESRD in African American men aged 25 to 44 in 1995, but only 13% in 2000 (method 1).
|
One-year survival rates for HIV-infected incident ESRD patients increased from 53.1% to 67.1% from 1995 to 2000 (method 1) (Figure 8).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These prevalence values may be underestimates because of concerns regarding confidentiality in different jurisdictions, and because of lack of biopsy confirmation (2,21). However, there is no reason to believe that underreporting would have changed during the time period covered by this study. Therefore, the trends are likely to be accurate.
Although the prevalence of renal disease in HIV-infected patients in the US ESRD program has increased, this is largely due to increased survival after the development of kidney failure. The incidence of HIV infection in the ESRD program remained unchanged and is decreasing as a percentage of all ESRD cases.
Although the incidence of the complication is decreasing, HIV infection remains an important risk factor for the development of ESRD. In 1995, the rate of ESRD among people living with AIDS, 5227 per million, was approximately 20-fold as great as that in the general population, 267 per million. In 2000, although the rate of ESRD among people living with AIDS had decreased 33.7% to 3467 per million, it was still approximately tenfold greater than that in the general population, 337 per million.
Recent work has closely linked the pathogenesis of HIVAN to infection of renal cells (1,2,12,13,21,22). Case reports have emphasized the possible reversibility of established HIVAN in patients treated with HAART (12,13). Recent epidemiologic reports have suggested that HAART is associated with decreased prevalence of renal disease and amelioration of progression in the general population and in HIV-infected patients with chronic renal disease (1,2,20,21,23–27). The increased prevalence and the decreased incidence of HIV infection in the US ESRD program may reflect earlier and more widespread treatment of patients with HIV infection with HAART, as well as improved survival of HIV-infected ESRD patients.
Although AIDS- and HIV-associated disease is the fifth leading discrete cause of ESRD, it is clear that as a proportion of the ESRD program, the association of HIV infection with ESRD is stable and that there is no epidemic increase of new HIV-associated renal diseases. (Missing, other, and unknown categories each comprise a larger number of patients than those listed as HIV/AIDS.) Our data on survival corroborate a recent report from Ahuja et al. (15). The quite dramatic 39.7% increase in survival has had a substantial effect on the epidemiology of the ESRD program. The finding cannot be plausibly attributed to improvement in renal replacement therapeutic techniques. Recent studies, however, have suggested that medications such as HAART and angiotensin-converting enzyme inhibitors, which might be associated with improved renal function and survival in this population are not maximally utilized in the ESRD program, suggesting survival statistics might be amenable to even greater improvement in the future (18,28,29). These findings are even more salient because it is well appreciated that newer antiretroviral agents can also cause nephropathy (1,2,17,27).
The percentage increase in prevalence of HIV infection in the ESRD program was greater for black than white patients, and a disproportionate number of HIV-infected patients in the incident ESRD population are men of African descent. Although the huge, 50-fold disparity in the incidence of HIV-associated ESRD could be associated with socioeconomic factors, including intravenous drug use and access to care and antiretroviral drugs (1,2,16,17,21), it is conceivable that genetic differences between populations in the expression of renal chemokine receptors (30), involved as secondary mediators of HIV infection (30,31), play a pathogenic role in the epidemiologic differences noted (2,21,31–33). This hypothesis is bolstered by the finding of clustering of renal disease in the families of HIV-infected ESRD patients (34). Special care must be taken regarding the evaluation of young African American men (and in light of the present findings older African American patients as well) with chronic kidney disease to make a diagnosis of an HIV-associated renal disease and institute appropriate therapy.
The prevalence rate for HIV-infected women in the ESRD program almost doubled or tripled (Tables 1 and 2
) during the study period, depending on the analytic technique used. The incidence of HIV-infected women in the ESRD program rose while it declined in men (Table 3), reflecting the demographics of the epidemic in the general population (35). Almost 2000 women, or more than a quarter of the cumulative population of HIV-infected patients, have initiated therapy for ESRD with hemodialysis, emphasizing that although the disease is more common in men, it certainly occurs in women. The relative risk for developing ESRD is 0.42 for women compared with men. Gender is not a reason to rule out a diagnosis of HIVAN or other HIV-associated renal disease (1,2).
It is almost certain that the true burden of AIDS in the ESRD population is underestimated in the USRDS database. The HCFA 2728 reporting form shows that in many areas, reporting of AIDS and/or HIV infection is prevented by patient confidentiality. We attempted to overcome this limitation by examining billing data for indications of AIDS diagnoses. However, if another payer is responsible for health care costs, then billing data will not be available. Of the 1347 prevalent patients identified in 1995 (Table 1), 968, or 71.9%, had Medicare entitlement at the time of enrollment in the ESRD program. Of these patients, 84% had subsequent AIDS billing, compared with only 6% of the persons without initial Medicare entitlement. We thus conclude that the 2728 AIDS/HIV designation has relatively good sensitivity.
The specificity of the 2728 criteria is much less certain. It is evident from the regional variation that there is underreporting of AIDS/HIV in certain states, notably California. Thus, the use of billing data likely adds to the case ascertainment.
It is not likely that a "gold standard" for the presence of AIDS/HIV in the universe of patients with ESRD will be developed. Patient confidentiality issues in a number of states will prevent the identification of many patients through the 2728 mechanism (36). The identification through billing data is also problematic because many patients have other sources of insurance coverage. Efficiencies in data collection and abstraction, such as the Vision/SIMS system in Medicare, will improve data transmission and accuracy, but they do not override the confidentiality issues raised above. ESRD networks are exploring the use of data collection through the E-lab system, but the system does not mandate HIV testing.
The lack of complete billing data for all persons suggests that our enumeration was less than complete. Still, there is no reason to hypothesize that the level of reporting has changed significantly during these years, so we believe that the trend data are probably accurate.
In addition, the validity of this analysis is bolstered because it is consistent with CDC dialysis facility surveillance data, using different ascertainment strategies (1,2,5). Tokars et al. (5) used a questionnaire mailed to all chronic hemodialysis units licensed by the HCFA in conjunction with the HCFA annual facility survey for calendar year 2000, covering, among other issues related to hemodialysis practices and infectious diseases, the number of patients with HIV infection. They recorded a 0.7% prevalence of AIDS and a total 1.4% prevalence of HIV infection in 1995, and a 0.4% prevalence of AIDS and a total 1.5% prevalence of HIV infection in 2000. By use of the USRDS database, we found AIDS or HIV infection accounted for a cumulative incidence of 1.3% from 1995 to 2000, and a prevalence of 1.16% in 2000. Thus, we believe the trend data reported here to be accurate.
Another inherent limitation of this study is the inability to link information regarding antiretroviral drug prescription and use with outcomes. Such data, however, are only available in much smaller population samples (28). Another limitation is that we cannot ascertain that the patients have HIVAN, compared with other renal diseases (1,2). We also cannot accurately ascertain the number of hepatitis C virus (HCV)-infected patients coinfected with HIV, who may have different prognostic outcomes (37), by use of these databases. The 2728 form, which codes for underlying renal disease, and the comorbid conditions section do not include HCV infection, although they do include HIV/AIDS. Billing data would only delineate HCV infection if it were specifically treated by a nonnephrologist Medicare billing physician. Analysis of diagnostic data in the USRDS regarding HCV infection in the HIV-infected population (data not shown), not surprisingly, led to counterintuitive results, indicating different patient populations designated in the database. Many HIV-infected ESRD patients with HCV as a coexisting infection may not be coded as such on the 2728 forms. More extensive use of biopsies in this population would also improve the ability to establish trends in large databases. However, our definitions have been designed to be inclusive and present a worst-case analysis regarding HIV infection in the ESRD program. Under these constraints and limitations, it is clear with more prevalent use of HAART and better preventive measures, HIV infection is unlikely to be epidemic in the US ESRD program.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. C. Odden, R. Scherzer, P. Bacchetti, L. A. Szczech, S. Sidney, C. Grunfeld, and M. G. Shlipak Cystatin C Level as a Marker of Kidney Function in Human Immunodeficiency Virus Infection: The FRAM Study Arch Intern Med, November 12, 2007; 167(20): 2213 - 2219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. McKenzie, S. L. Hendrickson, W. A. Briggs, R. A. Dart, S. M. Korbet, M. H. Mokrzycki, P. L. Kimmel, T. S. Ahuja, J. S. Berns, E. E. Simon, et al. NPHS2 Variation in Sporadic Focal Segmental Glomerulosclerosis J. Am. Soc. Nephrol., November 1, 2007; 18(11): 2987 - 2995. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. I. Choi, R. A. Rodriguez, P. Bacchetti, D. Bertenthal, P. A. Volberding, and A. M. O'Hare Racial Differences in End-Stage Renal Disease Rates in HIV Infection versus Diabetes J. Am. Soc. Nephrol., November 1, 2007; 18(11): 2968 - 2974. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D. Cohen and P. L. Kimmel HIV-associated renal diseases in Africa a desperate need for additional study Nephrol. Dial. Transplant., August 1, 2007; 22(8): 2116 - 2119. [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Jaffe and P. L. Kimmel Chronic Nephropathies of Cocaine and Heroin Abuse: A Critical Review Clin. J. Am. Soc. Nephrol., July 1, 2006; 1(4): 655 - 667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. Schwartz, L. A. Szczech, M. J. Ross, M. E. Klotman, J. A. Winston, and P. E. Klotman Highly Active Antiretroviral Therapy and the Epidemic of HIV+ End-Stage Renal Disease J. Am. Soc. Nephrol., August 1, 2005; 16(8): 2412 - 2420. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
