点击显示 收起
Yale University School of Medicine, Department of Epidemiology and Public Health, New Haven
Connecticut Emerging Infections Program, Department of Public Health, Hartford, Connecticut
Division of HIV/AIDS Prevention, National Center for HIV, STD and TB Prevention
Respiratory Diseases Branch, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
University of California at Berkeley, School of Public Health, Berkeley
Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland
Background.
Our goal was to describe trends in invasive pneumococcal disease incidence among persons with acquired immunodeficiency syndrome (AIDS) since the introduction of highly active antiretroviral therapy (HAART).
Methods.
We used time-trend analysis of annual invasive pneumococcal disease incidence rates from a population-based, active surveillance system. Annual incidence rates were calculated for 5 JulyJune periods by use of data from San Francisco county, the 6-county Baltimore metropolitan area, and Connecticut. The numerators were the numbers of invasive Streptococcus pneumoniae infections among persons 1864 years of age with AIDS; the denominators were the numbers of persons living with AIDS, estimated on the basis of AIDS surveillance data.
Results.
The annual incidence of invasive pneumococcal disease declined from 1094 cases/100,000 persons with AIDS (July 1995June 1996) to 467 cases/100,000 persons living with AIDS (July 1999June 2000). The annual percentage changes in incidence were -34%, -29%, -8%, and -1%. Declines were similar by surveillance area, sex, and race/ethnicity. During the final year of the study, the invasive pneumococcal disease incidence in persons with AIDS was half that of the pre-HAART era but was still 35 times higher than that in similarly aged nonHIV-infected adults.
Conclusions.
In the United States, invasive pneumococcal disease incidence declined sharply across a range of subgroups living with AIDS during the period after widespread introduction of HAART. Despite these gains, persons with AIDS remain at high risk for invasive pneumococcal disease.
The introduction of highly active antiretroviral therapy (HAART), which began in 1995, led to dramatic declines in HIV-related morbidity and mortality in the United States. By the end of 1997, when more than one-half of patients with AIDS were receiving HAART [1], AIDS deaths had declined by 70% [2], new cases of AIDS had declined by 25% [3], and the incidence of common opportunistic illnesses had declined by 80% in some cohorts [4]. HAART did not benefit all groups equally, however. National surveys found that African Americans, women, and publicly insured patients were less likely to receive HAART when it first became available and were less likely to continue therapy once they had started [1, 5].
Bacterial pneumonia is a major complication of HIV infection that has been shown to occur as frequently as Pneumocystis carinii pneumonia (PCP) yet earlier during the course of HIV disease [6]. Among HIV-infected persons, as in the general population, Streptococcus pneumoniae is the most common cause of bacterial pneumonia [7]. Invasive pneumococcal diseasedefined as isolation of S. pneumoniae from a normally sterile site, such as blood or cerebrospinal fluidoccurs in 15%30% of patients with pneumococcal pneumonia [7].
In the present analysis, we combined data from a multistate, population-based invasive pneumococcal disease surveillance system with AIDS surveillance data, to examine trends in the incidence of invasive pneumococcal disease among persons with AIDS since the introduction of HAART. We sought to measure, at the population level, the extent to which invasive pneumococcal disease incidence declined after the introduction of HAART and to evaluate whether trends differed across major demographic subgroups living with AIDS.
METHODS
Cases of invasive pneumococcal disease were reported through the Active Bacterial Core Surveillance/Emerging Infections Program network, a population-based surveillance system operating in 10 states. We used cases diagnosed between 1 July 1995 and 30 June 2000 in 3 surveillance areas where the AIDS status of patients could be ascertained throughout the 5-year period: San Francisco County, the 6-county Baltimore metropolitan area, and all 8 counties in Connecticut. The resident population in these 3 areas in 1999 was 6.5 million, or 2.4% of the US population (US Census Bureau data), and included 6% of the estimated number of persons with AIDS in the United States [3]. Overall, persons with AIDS in the 3 surveillance areas were less likely than the national population with AIDS to be Hispanic and were more likely to have injection drug use as a mode of HIV exposure.
The surveillance methodology has been described elsewhere [810]. At each site, surveillance personnel regularly contact participating microbiology laboratories to identify cases, defined as isolation of S. pneumoniae from a normally sterile site (blood, cerebrospinal fluid, pleural fluid, peritoneal fluid, pericardial fluid, sinus surgical aspirate, bone, or joint fluid) in a resident of the surveillance area. Laboratory records are reviewed every 6 months to verify completeness of reporting. For all cases, surveillance personnel or hospital infection control staff complete a standardized case-report form documenting demographic and clinical information.
Numerators used to calculate annual incidence rates consisted of the number of reported cases of invasive pneumococcal disease in patients 1864 years of age who were documented as having AIDS (on the basis of the 1993 Centers for Disease Control and Prevention AIDS case definition [11]) before or at the time of diagnosis of invasive pneumococcal disease. JulyJune periods were chosen so that each year would encompass an entire winter respiratory-disease season. Recurrent cases diagnosed at least 7 days after and within 1 year of a previously reported episode were used to estimate recurrence rates but were excluded from calculations of incidence. Seven (1.3%) cases, in patients of unknown race/ethnicity, were excluded from race/ethnicity-specific analyses.
AIDS status was ascertained in 1 of 2 ways. In San Francisco County and the Baltimore metropolitan area, surveillance personnel have routinely collected information on chronic medical conditions from all patients with invasive pneumococcal disease since 1995, including AIDS status, if noted in the medical record. In Connecticut, where AIDS status was not systematically recorded until January 1997, we conducted a match between the Connecticut AIDS registry and the Active Bacterial Core Surveillance database, to identify persons with AIDS diagnosed before or within 30 days of invasive pneumococcal disease onset. This match was performed in accordance with state confidentiality laws.
Denominators for calculating annual incidence rates consisted of the estimated number of persons 1864 years of age living with AIDS (on the basis of the 1993 CDC AIDS case definition [11]) on 31 December of each JulyJune period, adjusted for delays in the reporting of AIDS and death. These data were obtained from the CDC's Division of HIV/AIDS Prevention Surveillance Branch, with the permission of state AIDS surveillance coordinators from each study area. CDC AIDS data-release policies suppressed cells containing 3 persons with AIDS and restricted cross-tabulations to county and 1 additional variable (e.g., race, sex, or age group). To limit the number of small cells, we restricted our analysis to individuals 1864 years of age, who comprised >95% of persons with AIDS in the surveillance areas. Because of our interest in evaluating whether trends differed by race, and in view of prior research in the HIV-infected population suggesting that race is a more important predictor of invasive pneumococcal disease than is sex or age group [12, 13], we stratified by race/ethnicity in our main analysis.
To calculate invasive pneumococcal disease incidence rates in the general population of persons 1864 years of age, we again excluded recurrent cases from the numerator and used 1 January population estimates, interpolated from 1 July census estimates (US Census Bureau data), in the denominator. Incidence in the nonHIV-infected population was approximated by limiting the numerator to cases of invasive pneumococcal disease in patients not known to be HIV infected and subtracting persons living with AIDS from the 1 July census estimate denominators. The number of HIV-infected persons without AIDS was not subtracted from the denominator, since it is unknown; however, doing so would have had a negligible effect on incidence rates.
Annual invasive pneumococcal disease incidence rates per 100,000 persons with AIDS were calculated and plotted by sex and race/ethnicity for the 3 surveillance areas combined, as well as by race/ethnicity for each area separately. Confidence intervals (CIs) for individual rates were calculated on the basis of the normal approximation of the log rates [14]. Incidence trends were analyzed using Poisson regression. The dependent variable, Y, was the count of patients with invasive pneumococcal disease who had AIDS in each stratum. The logarithm of the denominator P, the number of persons with AIDS in each stratum, was entered as an offset in the model to scale invasive pneumococcal disease counts to the population at risk. This is the equivalent of modeling the log incidence rate. We used preliminary models to evaluate trends and to calculate rate ratios by sex and race/ethnicity for the 3 surveillance areas combined. Our main analysis examined whether invasive pneumococcal disease incidence trends differed by race and geographic area. Because of the small number of patients of Hispanic race/ethnicity, this analysis was restricted to non-Hispanic black and non-Hispanic white individuals, hereafter referred to as "blacks" and "whites." Numerators and denominators were grouped in strata defined by the independent variables period (5 JulyJune periods), surveillance area (San Francisco County, Baltimore metropolitan area, or Connecticut), and race (black or white). Model parameters and associated statistics were estimated by maximum likelihood, by use of PROC GENMOD in SAS for Windows (version 8.02; SAS Institute). The statistical significance of the independent variables and their interactions was assessed by use of the likelihood-ratio test [15]. Significance was set at the standard .05 level, and all CIs given are 95%. The variable period, initially parameterized as a set of 4 indicator variables, was reduced to 2 variables when linear contrasts [16] indicated that the last 3 periods could be collapsed into 1 category with no significant loss of fit. The deviance of the final model was 20.2 on 23 df (P = .6, 2 test) suggesting adequate fit [17].
RESULTS
Of the 3857 cases of invasive pneumococcal disease reported in patients 1864 years of age between 1 July 1995 and 30 June 2000 in the 3 surveillance areas, 572 (15%) were documented as having AIDS. The frequency of recurrent disease could be estimated among 476 patients with invasive pneumococcal disease and AIDS whose cases of invasive pneumococcal disease were reported at least 1 year before the end of the study; 28 (5.9%) had a second episode at least 7 days after and within 1 year of the first episode. In contrast, recurrent disease was reported in 26 (1.2%) of 2,221 patients with invasive pneumococcal disease who were not known to be HIV infected (P < .001, 2 test).
After the 33 patients with recurrent disease were removed, 539 patients with invasive pneumococcal disease who had AIDS were available for trend analyses: 180 from San Francisco County, 191 from Connecticut, and 168 from the Baltimore metropolitan area. Seventy-one percent of patients were male (53% black, 34% white, and 12% Hispanic) and 29% were female (77% black, 16% white, and 7% Hispanic). The incidence was higher among women than among men with AIDS (figure 1). During the entire 5 years, the rate ratio (RR) of invasive pneumococcal disease in women compared with men was 1.7 (CI, 1.42.1). The overall incidence was also higher among blacks, compared with that among whites (RR, 2.2 [CI, 1.82.6]) and Hispanics (RR, 1.8 [CI, 1.42.4]). Incidence did not differ significantly between Hispanics and whites (RR, 1.2 [CI, 0.91.6]).
Analysis of sex- and race/ethnicity-specific trends was possible for the 3 surveillance areas combined. Invasive pneumococcal disease incidence declined significantly between the first and third JulyJune periods in all groups (P < .001 in all models, by Poisson regression likelihood-ratio test). No significant differences were observed during the last 2 years of the study (P = .8 in both sex- and race/ethnicity-specific models). Inspection of plotted rates for the combined surveillance areas (figure 1) suggests that declines in incidence occurred earlier among men than among women and among whites than among blacks and Hispanics. However, these differences could not be distinguished from random variation under the Poisson model, on the basis of tests for interaction between period and sex in the analysis of sex-specific trends (P = .86) or on the basis of tests for interaction between period and race in the analysis of race-specific trends (P = .32).
In our primary analysis, we evaluated whether trends in invasive pneumococcal disease incidence in persons with AIDS differed by race and whether this was consistent across the 3 surveillance areas. A final model included the variables period (with the third, fourth, and fifth JulyJune periods collapsed into 1 category), race (black vs. white), and surveillance area (indicator variables for Connecticut and San Francisco County, with the Baltimore metropolitan area as reference), all of which were significant, independent predictors of invasive pneumococcal disease incidence. The addition of the interaction term race*area significantly improved the model fit (P = .01), suggesting that there are significant differences by surveillance area in the RR of invasive pneumococcal disease incidence in blacks versus whites. Specifically, the RR in blacks versus whites was significantly higher in San Francisco County (RR, 3.6 [CI, 2.75.0]) than in Connecticut (RR, 1.8 [CI, 1.32.5]), whereas RRs in Baltimore were intermediate (RR, 2.4 [CI, 1.44.3]). The addition of the interaction terms area*period (P = .36) and race*period (P = .39) did not significantly improve model fit, an indication that declines in invasive pneumococcal disease incidence were similar in all 3 surveillance areas and did not differ significantly by race.
DISCUSSION
We observed a 57% decline in invasive pneumococcal disease incidence among persons with AIDS during the first 5 years after the introduction of HAART. Most of this decline occurred between July 1995 and June 1998, when the use of protease inhibitors expanded rapidly in the United States [1]. During the last 2 years of our study, overall invasive pneumococcal disease incidence stabilized at levels that were roughly one-half of those of the pre-HAART era but that were still 35 times higher than the estimated rate among similarly aged nonHIV-infected adults.
These findings extend earlier observations based on data from the San Francisco County surveillance site where invasive pneumococcal disease incidence among persons with AIDS declined 50% between 1994 and mid-1997 [18]. Our data, collected through June 2000, document a lasting change in invasive pneumococcal disease risk in this population and show that the magnitude and timing of declines were similar among men and women; among non-Hispanic whites, non-Hispanic blacks, and Hispanics; and across the 3 geographic areas studied.
That HAART is the primary cause of these trends is supported by the timing of its introduction into clinical practice, its demonstrated protective effect against bacterial infections, and the absence of alternative explanations. Protease inhibitors were approved for commercial release in late 1995 (saquinavir) and early 1996 (ritonavir and indinavir) [19] and were rapidly adopted as the standard of care for patients with HIV infection. Among patients with AIDS who were enrolled in the Adult Spectrum of Disease Study, use of triple therapy went from near 0% during the fourth quarter of 1995 to 60% by the end of 1997 [20]. Similarly, in a comprehensive national survey, 17% of patients with a CD4+ T cell count <500 cells/mm3 had used a HAART regimen by January 1996, 59% had used a HAART regimen by January 1997, and 85% had used a HAART regimen by January 1998 [21]. During this same time, dramatic declines occurred in AIDS mortality and in the incidence of PCP and other opportunistic illnesses [22]. Our data confirm that, among persons with AIDS, invasive pneumococcal disease followed similar trends, in contrast to the relatively stable incidence observed in the nonHIV-infected population.
Pneumococcal polysaccharide vaccine has been recommended for patients with HIV in the United States since the 1980s [23, 24] and, if administered early enough during the course of HIV disease, has been suggested to cut the risk of invasive pneumococcal disease by one-half or more [12, 25, 26], although it was not found to be efficacious in a randomized controlled trial conducted among HIV-infected adults in Uganda [27]. The vaccine is unlikely to have greatly influenced invasive pneumococcal disease trends during our study, however, because vaccination rates remained low and stable throughout the 1990s and relatively few patients with HIV were vaccinated when their CD4+ T cell counts were high [12]. The pneumococcal conjugate vaccine could not have affected our main findings, either, since its licensure for use in children occurred during the final 6 months of our analysis. Similarly, trimethoprim-sulfamethoxazole prophylaxis to prevent PCP has been shown to protect against bacterial pneumonia in some studies [28, 29] but not in others [12], and utilization did not change significantly during our study period [4, 20]. Use of clarithromycin or azithromycin to prevent disseminated infection with Mycobacterium avium complex (MAC) did increase during the mid-1990s [20]; however, these drugs are associated with only modest protection against invasive pneumococcal disease and are recommended for prevention of disseminated MAC only in those patients whose CD4+ T cell count is <50 cells/mm3 [30]. Neither vaccination nor antimicrobial prophylaxis are likely explanations for the abrupt declines in the incidence of invasive pneumococcal disease observed between 1995 and 1997.
Data from clinical studies document a strong protective effect of HAART against bacterial infections, including invasive pneumococcal disease. Among patients with AIDS at a Baltimore clinic, protease inhibitor regimens were associated with a nearly 50% reduction in the risk of all-cause bacterial pneumonia [13]. Similarly, the Adult Spectrum of Disease Study reported that triple therapy had a strong, independent, protective effect against pneumococcal infections (adjusted RR, 0.5) after known risk factors, including CD4+ T cell counts, were controlled for [12]. In contrast to PCP and other common HIV-related illnesses, which are controlled primarily by cell-mediated immunity, control of encapsulated bacteria such as S. pneumoniae depends to a greater extent on the humoral immune response [7, 13]. Our results from population-level surveillance add to a body of clinical evidence of the beneficial effects of HAART on this arm of the immune system [31].
During the last 2 years of our study, invasive pneumococcal disease incidence among individuals with AIDS was 35 times higher than that among nonHIV-infected persons. Moreover, recurrent invasive pneumococcal disease, most of which is due to new infections rather than to relapses [32], was 5 times more frequent among patients with AIDS than among nonHIV-infected patients. Studies conducted before 1996 reported that HIV-infected persons had a risk of invasive pneumococcal disease 25150 times higher than that among similarly aged nonHIV-infected adults [3336]. Our data show that, despite the halving of invasive pneumococcal disease incidence after the introduction of HAART, persons with AIDS remain at extraordinarily high risk.
Overall, during our study, invasive pneumococcal disease incidence among blacks with AIDS was twice as high as that among whites and Hispanics with AIDS. This is consistent with previous findings in both the HIV-infected [29, 37] and general [10, 35, 36, 38] populations. Multivariate analysis of the full Adult Spectrum of Disease Study database (71,116 person-years) suggests that race-associated differences in invasive pneumococcal disease risk are only partly explained by known risk factors. After immune suppression, antiretroviral therapy, mode of HIV transmission, and alcoholism are controlled for, invasive pneumococcal disease incidence is still 50% higher among blacks than among whites [12].
We also found invasive pneumococcal disease incidence to be higher among women with AIDS than among men with AIDS. This was unexpected, because studies in the general population consistently find the incidence to be higher among men [18, 36]. The difference may have been partly due to ascertainment bias in the AIDS denominator. Women at the same level of immunodeficiency may be less likely than men to receive an AIDS diagnosis. Confounding by race and other factors is also likely to be at play. The Adult Spectrum of Disease Study, for example, reported higher rates of pneumococcal disease among HIV-infected women (980 cases/100,000 persons) than among HIV-infected men (780 cases/100,000 persons) in univariate analysis but not in multivariate analysis [12]. Women with AIDS may face more-frequent exposure to S. pneumoniae, through contact with children. It is also possible that they have not benefited from HIV therapy to the extent that men have. A national survey found that 22% of HIV-infected women had not received HAART by January 1998, compared with 13% of HIV-infected men [1, 5].
The main question we addressed was whether trends in invasive pneumococcal disease incidence among persons with AIDS after the introduction of HAART differed by race/ethnicity. Disparities in access to and utilization of HAART [1, 5] could, for example, result in a widening of the gap in invasive pneumococcal disease incidence between blacks and whites. Our results suggest that this was not the case. In terms of reducing invasive pneumococcal disease, the benefits of HAART were seen across a variety of AIDS subgroups.
Our study had several limitations. We did not collect information on antiretroviral therapy use in patients with invasive pneumococcal disease or in the underlying population with AIDS and are therefore unable to demonstrate, at the individual level, an association between HAART and lower disease risk. Our conclusion that HAART is the most likely explanation for the abrupt declines in invasive pneumococcal disease incidence is based on the coincident timing of these 2 events and on consistency with clinical research. Because of AIDS data-release policies, we also were unable to stratify the data by both race and sex in the surveillance areaspecific analyses. Finally, we do not know how representative the 3 surveillance areas are of other regions, although examination of AIDS mortality data shows that AIDS deaths as a proportion of persons living with AIDS in the 3 surveillance areas mirrored national trends, with the sharpest drop in mortality occurring between 1995 and 1998 [3944].
Despite declines in invasive pneumococcal disease incidence since the introduction of HAART, S. pneumoniae continues to cause significant morbidity and mortality among persons with HIV infection and AIDS. The steady increase in the size of the HIV-infected population [45]as well as the disproportionate impact of the epidemic on African Americans, who are at higher risk for invasive pneumococcal diseasesuggest that the burden of invasive pneumococcal disease may increase in the future. In sub-Saharan Africa and Asia, where the HIV epidemic has hit hardest, pneumococcal disease accounts for considerably greater morbidity and mortality. Efforts to accelerate access to antiretroviral drugs among patients with AIDS in developing countries could have a substantial impact on the prevention of invasive pneumococcal disease worldwide.
Acknowledgments
We are grateful for the assistance of Carolyn Wright (Respiratory Diseases Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention); members of the Active Bacterial Core Surveillance of the Emerging Infection Program Network; Joe Posid and Jianmin Li (Division of HIV/AIDS PreventionSurveillance and Epidemiology, National Center for HIV, STD and TB Prevention, Centers for Disease Control and Prevention); Kenneth Carley (AIDS Epidemiology, Department of Public Health, Hartford, Connecticut); and the Connecticut, California, and Maryland state AIDS surveillance staff.
References
1. Shapiro MF, Morton SC, McCaffrey DF, et al. Variations in the care of HIV-infected adults in the United States: results from the HIV Cost and Services Utilization Study. JAMA 1999; 281:230515. First citation in article
2. Chiasson MA. New York City AIDS case mortality rates in the era of potent antiretroviral therapy. J Urban Health 2000; 77:2557. First citation in article
3. Centers for Disease Control and Prevention. HIV/AIDS surveillance report. Vol. 10. Atlanta: CDC, 1997:143. First citation in article
4. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med 1998; 338:85360. First citation in article
5. Cunningham WE, Markson LE, Andersen RM, et al. Prevalence and predictors of highly active antiretroviral therapy use in patients with HIV infection in the United States. HCSUS Consortium. HIV Cost and Services Utilization. J Acquir Immune Defic Syndr 2000; 25:11523. First citation in article
6. Osmond DH, Chin DP, Glassroth J, et al. Impact of bacterial pneumonia and Pneumocystis carinii pneumonia on human immunodeficiency virus disease progression. Pulmonary Complications of HIV Study Group. Clin Infect Dis 1999; 29:53643. First citation in article
7. Janoff EN, Breiman RF, Daley CL, Hopewell PC. Pneumococcal disease during HIV infection: epidemiologic, clinical, and immunologic perspectives. Ann Intern Med 1992; 117:31424. First citation in article
8. Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. Active Surveillance Team. N Engl J Med 1997; 337:9706. First citation in article
9. Schuchat A, Hilger T, Zell E, et al. Active bacterial core surveillance of the Emerging Infections Program Network. Emerg Infect Dis 2001; 7:929. First citation in article
10. Robinson KA, Baughman W, Rothrock G, et al. Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 19951998: opportunities for prevention in the conjugate vaccine era. JAMA 2001; 285:172935. First citation in article
11. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep 1992; 41:119. First citation in article
12. Dworkin MS, Ward JW, Hanson DL, Jones JL, Kaplan JE. Pneumococcal disease among human immunodeficiency virusinfected persons: incidence, risk factors, and impact of vaccination. Clin Infect Dis 2001; 32:794800. First citation in article
13. Sullivan JH, Moore RD, Keruly JC, Chaisson RE. Effect of antiretroviral therapy on the incidence of bacterial pneumonia in patients with advanced HIV infection. Am J Respir Crit Care Med 2000; 162:647. First citation in article
14. Holford TR. Multivariate methods in epidemiology. New York: Oxford University Press, 2002. First citation in article
15. McCullagh P, Nelder J. Generalized linear models. 1st ed. London: Chapman and Hall, 1983. First citation in article
16. Kleinbaum D, Kupper L, Muller K. Applied regression analysis and other multivariable methods. 2nd ed. Boston: PWS-Kent, 1988:26096. First citation in article
17. Selvin S. Statistical analysis of epidemiologic data. 2nd ed. New York: Oxford University Press, 1996. First citation in article
18. Nuorti JP, Butler JC, Gelling L, Kool JL, Reingold AL, Vugia DJ. Epidemiologic relation between HIV and invasive pneumococcal disease in San Francisco County, California. Ann Intern Med 2000; 132:18290. First citation in article
19. Perdue BE, Weidle PJ, Everson-Mays RE, Bozek PS. Evaluating the cost of medications for ambulatory HIV-infected persons in association with landmark changes in antiretroviral therapy. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 17:35460. First citation in article
20. McNaghten AD, Hanson DL, Jones JL, Dworkin MS, Ward JW. Effects of antiretroviral therapy and opportunistic illness primary chemoprophylaxis on survival after AIDS diagnosis. Adult/Adolescent Spectrum of Disease Group. AIDS 1999; 13:168795. First citation in article
21. Bozzette SA, Berry SH, Duan N, et al. The care of HIV-infected adults in the United States. HIV Cost and Services Utilization Study Consortium. N Engl J Med 1998; 339:1897904. First citation in article
22. Kaplan JE, Hanson D, Dworkin MS, et al. Epidemiology of human immunodeficiency virusassociated opportunistic infections in the United States in the era of highly active antiretroviral therapy. Clin Infect Dis 2000; 30(Suppl 1):S514. First citation in article
23. Recommendations of the immunizations practices advisory committee: pneumococcal polysaccharide vaccine. MMWR Morb Mortal Wkly Rep 1989; 38:648, 736. First citation in article
24. Kaplan JE, Masur H, Holmes KK. Guidelines for preventing opportunistic infections among HIV-infected persons2002. Recommendations of the US Public Health Service and the Infectious Diseases Society of America. MMWR Recomm Rep 2002; 51:152. First citation in article
25. Gebo KA, Moore RD, Keruly JC, Chaisson RE. Risk factors for pneumococcal disease in human immunodeficiency virusinfected patients. J Infect Dis 1996; 173:85762. First citation in article
26. Breiman RF, Keller DW, Phelan MA, et al. Evaluation of effectiveness of the 23-valent pneumococcal capsular polysaccharide vaccine for HIV-infected patients. Arch Intern Med 2000; 160:26338. First citation in article
27. French N, Nakiyingi J, Carpenter LM, et al. 23-valent pneumococcal polysaccharide vaccine in HIV-1-infected Ugandan adults: double-blind, randomised and placebo controlled trial. Lancet 2000; 355:210611. First citation in article
28. Hirschtick RE, Glassroth J, Jordan MC, et al. Bacterial pneumonia in persons infected with the human immunodeficiency virus. Pulmonary Complications of HIV Infection Study Group. N Engl J Med 1995; 333:84551. First citation in article
29. Navin TR, Rimland D, Lennox JL, et al. Risk factors for community-acquired pneumonia among persons infected with human immunodeficiency virus. J Infect Dis 2000; 181:15864. First citation in article
30. Detels R, Munoz A, McFarlane G, et al. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. Multicenter AIDS Cohort Study Investigators. JAMA 1998; 280:1497503. First citation in article
31. Feikin DR, Feldman C, Schuchat A, Janoff EN. Global strategies to prevent bacterial pneumonia in adults with HIV disease: an evidence-based review. Lancet Infect Dis 2004; 4:44555. First citation in article
32. McEllistrem MC, Mendelsohn AB, Pass MA, et al. Recurrent invasive pneumococcal disease in individuals with human immunodeficiency virus infection. J Infect Dis 2002; 185:13648. First citation in article
33. Redd SC, Rutherford GW 3rd, Sande MA, et al. The role of human immunodeficiency virus infection in pneumococcal bacteremia in San Francisco residents. J Infect Dis 1990; 162:10127. First citation in article
34. Schuchat A, Broome CV, Hightower A, Costa SJ, Parkin W. Use of surveillance for invasive pneumococcal disease to estimate the size of the immunosuppressed HIV-infected population. JAMA 1991; 265:32759. First citation in article
35. Bennett NM, Buffington J, LaForce FM. Pneumococcal bacteremia in Monroe County, New York. Am J Public Health 1992; 82:15136. First citation in article
36. Breiman RF, Spika JS, Navarro VJ, Darden PM, Darby CP. Pneumococcal bacteremia in Charleston County, South Carolina. A decade later. Arch Intern Med 1990; 150:14015. First citation in article
37. Guerrero M, Kruger S, Saitoh A, et al. Pneumonia in HIV-infected patients: a case-control survey of factors involved in risk and prevention. AIDS 1999; 13:19715. First citation in article
38. Nuorti JP, Butler JC, Farley MM, et al. Cigarette smoking and invasive pneumococcal disease. Active Bacterial Core Surveillance Team. N Engl J Med 2000; 342:6819. First citation in article
39. Centers for Disease Control and Prevention. HIV/AIDS surveillance reports. Vol. 10. Atlanta, GA: Centers for Disease Control and Prevention, 1998. First citation in article
40. Centers for Disease Control and Prevention. HIV/AIDS surveillance reports. Vol. 11. Atlanta, GA: Centers for Disease Control and Prevention, 1999. First citation in article
41. Centers for Disease Control and Prevention. HIV/AIDS surveillance reports. Vol. 12. Atlanta, GA: Centers for Disease Control and Prevention, 2000. First citation in article
42. Connecticut Department of Public Health. HIV/AIDS surveillance annual report. 31 December 2000. First citation in article
43. San Francisco Department of Public Health. Quarterly AIDS surveillance report. March 2001. First citation in article
44. Maryland Department of Health and Mental Hygiene. The Maryland 2000 HIV/AIDS annual report. 31 December 2000. First citation in article
45. Karon JM, Fleming PL, Steketee RW, De Cock KM. HIV in the United States at the turn of the century: an epidemic in transition. Am J Public Health 2001; 91:10608. First citation in article