Bacterial infections in AIDS patients.

Bacterial infections cause a substantial burden of disease in the HIV-infected population. Like the opportunistic infections associated with AIDS, they are often usually severe. Unlike many AIDS-related opportunistic infections, however, many can be eradicated by therapy with the appropriate antibiotic, although some, like Salmonella sepsis, require prolonged suppressive therapy to prevent recurrences. Most importantly, the possibility of a bacterial infection should be kept in mind when evaluating any acutely ill HIV-infected patient, as therapy of bacterial disease is often curative while untreated infection can lead to severe morbidity and premature death.

Lymph node architecture preceding and following 6 months of potent antiviral therapy: follicular hyperplasia persists in parallel with p24 antigen restoration after involution and CD4 cell depletion in an AIDS patient.

OBJECTIVES: To evaluate changes in architecture, viral RNA, and viral protein over 6 months in lymph nodes from retroviral-naïve HIV-infected persons before and after commencing highly active antiretroviral therapy (HAART). METHODS: Nine antiretroviral-naïve HIV-infected persons had lymph nodes excised at baseline and at 2 and 6-8 months after beginning a four-drug combination regimen containing zidovudine, lamivudine, nevirapine, and indinavir. Two patients had AIDS. Lymph nodes were examined by immunohistochemical staining for Gag p24 HIV, CD3, CD21, CD20, HAM 56, and Ki67 antigens and by in-situ hybridization (ISH) for HIV RNA and H3-histone RNA. RESULTS: Eight of nine baseline lymph nodes showed follicular hyperplasia and germinal center and paracortical mononuclear cell activation. At 2 months, the lymph nodes from seven patients, including the AIDS patients, showed more follicular hyperplasia and activation than their baseline specimens but with decreased mononuclear cell activation. By 6 months, seven lymph nodes were less hyperplastic and activated than their corresponding 2 month specimens. Combined ISH/immunohistochemical staining of baseline lymph nodes revealed productively infected T (CD3) and B (CD20) cells and macrophages (HAM56+). HIV RNA-positive mononuclear cells were infrequent at 2 months, and rare at 6 months. HIV RNA was still associated with follicular dendritic cells (FDC) at 2 months, but not at 6 months. HIV p24-positive antigen in germinal centers persisted through all 6, and the one 8 month specimens. The baseline lymph nodes from one of the AIDS patients was involuted and T cell depleted, whereas the follow-up lymph nodes were hyperplastic with normal T cell levels. CONCLUSION: Follicular hyperplasia and cell activation, possibly caused by persistent viral protein in germinal centers, may help explain why HIV viremia rebounds so rapidly after the interruption of HAART. Restoration of architecture may follow the treatment of patients with AIDS who initially had involuted and CD4 cell-depleted lymph nodes.

Survival differences in patients with AIDS.

To define the clinical, demographic, and behavioral variables that may influence survival in patients with AIDS, we studied 526 patients with AIDS diagnosed through September 1987 who were cared for at a single medical center. A diversity of racial and ethnic backgrounds, ages, both men and women, and all risk behaviors except hemophilia were well represented. The initial AIDS defining diagnosis was the most powerful predictor of survival. The median survival was 12.8 months for patients presenting with Kaposi's sarcoma (p less than 0.001), 10.9 months for patients presenting with Pneumocystis carinii pneumonia (p less than 0.001), and 4.8 months for patients presenting with other infections or neoplasms (p less than 0.02). For the entire series, male sex and younger age were associated with more favorable survival (p less than 0.025). For those presenting with Pneumocystis carinii pneumonia, in addition to younger age (p less than 0.025), black race (p less than 0.025) and the combination of male sex and intravenous drug use (p less than 0.005) were associated with a more favorable survival. Within a setting of comparable clinical care, survival from the point of diagnosis of AIDS is associated most strongly with the initial AIDS diagnosis, but differences in age, gender, race, and risk behavior also exert an influence on survival.

Acquired immune deficiency syndrome occurring within 5 years of infection with human immunodeficiency virus type-1: the Multicenter AIDS Cohort Study.

The objective of this study is to describe participants in the Multicenter AIDS Cohort Study (MACS) with incident infection due to the human immunodeficiency virus type-1 (HIV-1) in whom AIDS developed by March 1990 and within 5 years of seroconversion (group A). Secondly, behavioral, clinical, and immunologic characteristics of these men are compared to those of matched seroconverters remaining AIDS free (group B). Between entry into the MACS (April 1984-March 1985) and July 1989, 345 seronegative homosexual/bisexual men had HIV-1 antibody; of these men, AIDS developed in 32 by March 1990. The Kaplan-Meier estimates of the proportion of men with incident HIV-1 infection with AIDS were 6 months, 0%; 12 months, 1%; 24 months, 3%; and 48 months, 10%. These 32 men engaged in receptive anal intercourse with more partners before (p less than 0.005) and after seroconversion (p less than 0.005) and reported more sexually transmitted disease preseroconversion (p = 0.05) than did group B. These findings suggest that sexually transmitted co-factors, preseroconversion and/or postseroconversion, play a role in the progression of HIV-1 infection. Alternatively, greater sexual activity could increase the hazard of exposure to a more virulent HIV-1 strain.

Population-based estimates of zidovudine and aerosol pentamidine use in San Francisco: 1987-1989.

From the San Francisco Men's Health Study (SFMHS) and the San Francisco General Hospital Cohort we derived partially population-based estimates of zidovudine (ZDV) use in San Francisco from 1987 to 1989. Data from the SFMHS alone were used to make estimates of aerosol pentamidine (AP) use in 1989. From 1987 to 1989, zidovudine use increased from 36 to 68% in participants with AIDS. In participants with symptomatic HIV infection without AIDS and in those with less than 200 CD4 cells, ZDV use increased initially but then leveled off (from 6 to 25% and 24 to 55%, respectively). Zidovudine use with more than 500 CD4 cells increased from 0.5 to 4%. In 1989 AP use with less than 200 CD4 cells was 42% and with AIDS was 44%. Whereas 50% of participants with AIDS and less than 200 CD4 were using both ZDV and AP, only 14% of those without AIDS but with less than 200 CD4 cells were using both. Surprisingly few members of these cohorts are using therapies proven effective in reducing the morbidity and mortality of HIV infection.

Cervical and vaginal squamous cell abnormalities in women infected with human immunodeficiency virus.

We sought to determine whether women infected with human immunodeficiency virus (HIV) had cervicovaginal cellular changes suggesting lower genital tract neoplasia or human papillomavirus (HPV) infection at a rate different from that in women without HIV infection. In a blinded fashion, cytological preparations of cervicovaginal smears from women infected with the HIV were analyzed and compared to preparations from women at high risk for but not infected with HIV. Eleven of 35 (31%) HIV-infected subjects had evidence of squamous abnormalities compared with 1 of 23 (4%) non-HIV-infected women (p = 0.019). Nine of 35 (26%) HIV-infected women had cytohistological evidence of human papillomavirus (HPV) infection compared to 1 of 23 (4%) non-HIV-infected women (p = 0.072). We conclude that HIV-infected women have a high prevalence of cervical and vaginal cytological abnormalities and evidence of genital HPV infection. Further study is necessary to determine whether there is an increased risk for cervicovaginal neoplastic disorders in women infected with HIV.

Cervical cytologic abnormalities and papillomavirus in women infected with human immunodeficiency virus.

We investigated the relationship of human papillomavirus (HPV) infection of the female genital tract, cervical cytology, and human immunodeficiency virus (HIV) infection in 67 women. Forty-eight women had a history of intravenous drug use, 18 were heterosexual partners of HIV-infected intravenous drug users, and one was a transfusion recipient. Patients received a Pap smear, cervicovaginal lavage for HPV determination by Southern blot, HIV serum antibody by enzyme immunoassay with Western blot confirmation, and thorough screening for other sexually transmitted diseases. Seventeen of the 35 (49%) women seropositive for HIV had HPV infection, compared with 8 of 32 (25%) seronegative women (p less than 0.05). Fourteen of 35 (40%) HIV-positive women had squamous intraepithelial lesions (SIL) on cervical cytology, compared with three of 32 (9%) HIV-negative women (p less than 0.01). Of 22 women with symptomatic HIV infection, 11 (50%) had SIL on cytology; 10 of these 11 were HPV-positive. Among 13 asymptomatic HIV-positive women, only three (23%) had such cytological lesions. Our findings strongly suggest that HIV-induced immunosuppression exacerbates HPV-mediated cervical cytologic abnormalities.

National AIDS incidence trends and the extent of zidovudine therapy in selected demographic and transmission groups.

After mid-1987 fewer than the expected number of cases of AIDS were reported in the United States in some demographic and transmission groups but not in others. Gay men (regardless of intravenous drug use), adults with hemophilia, and transfusion recipients exhibited fewer cases than expected based on previously reliable models. These favorable trends could not be explained by assuming earlier cessation of human immunodeficiency virus (HIV) infection. Favorable AIDS incidence trends were not found in heterosexual intravenous drug users or in persons infected through heterosexual contact. White gay men from New York City, Los Angeles, and San Francisco experienced markedly favorable trends, whereas little changes was observed for nonwhite gay men from nonurban areas. AIDS incidence trends were quantitatively consistent with the fraction of AIDS-free persons with severe immunodeficiency who received zidovudine in three cohorts. Gay men in San Francisco used zidovudine more frequently than did adults with hemophilia, while little was used by intravenous drug users in New York City. Data describing the initial national distribution of zidovudine (March 31-September 18, 1987) indicated relatively high use by patients with severe immunodeficiency in those groups, such as urban white gay men, that subsequently experienced fewer cases of AIDS than expected. Available data suggest that zidovudine, perhaps in combination with other therapies, has been one factor contributing to favorable AIDS incidence trends in some groups. Broader application of therapy might further retard the incidence of AIDS, especially in intravenous drug users, persons infected through heterosexual contact, minorities, women, and persons diagnosed outside major metropolitan areas.

Impact of immunosuppression on health care use by men in the Multicenter AIDS Cohort Study (MACS).

The effects of human immunodeficiency virus type 1 (HIV-1) serostatus, AIDS, and level of immunosuppression on health service use were examined in the Multicenter AIDS Cohort Study. Data on self-reported hospitalizations, outpatient medical services (non-emergency room) and emergency room care during the preceding 6 months were collected for 3,447 homosexual/bisexual men returning for their 14th and/or 15th semiannual visits in Chicago, Baltimore, Los Angeles, and Pittsburgh. AIDS-free seropositive men with CD4+ cells < 200/microliters were more likely to be hospitalized [odds ratio (OR) = 2.3, 95% confidence limits (CL) = 1.4, 3.8] and use outpatient medical care (OR = 7.9, 95% CL = 4.9, 12.6), compared with seronegative men. Increased outpatient care was initiated at the earliest stages of HIV-1 infection, even when CD4+ cells were > 500/microliter. Dramatic increases in outpatient care for each level of immunosuppression were observed. HIV-1-related symptoms were associated with increased hospitalizations (OR = 4.8, 95% CL = 3.2, 7.3), use of outpatient medical services (OR = 3.3, 95% CL = 1.9, 5.6), and emergency room care (OR = 3.1, 95% CL = 2.1, 4.6). Persons with AIDS and < or = 50 CD4+ cells/microliter most likely to be hospitalized (OR = 8.1; 95% CL = 4.4, 14.9). No significant difference (p > 0.05) in emergency room use was observed according to HIV-1 serostatus, AIDS, or immunosuppression, after adjusting for insurance and clinical symptoms. To the extent that CD4+ cell counts are used as one of the criteria for an AIDS diagnosis and such a diagnosis broadens available benefits to persons with HIV disease, the pattern of health care services described here will be important for health care providers and planners.

Smoking prevalence during pregnancy for women who are and women who are not Medicaid-funded.

Maternal smoking has been related to a number of adverse pregnancy outcomes. Although maternal smoking prevalence has decreased over time, certain populations have retained a high smoking prevalence and remain at high risk for adverse pregnancy outcomes. This study used the Washington State First Steps Program Database to estimate the difference in maternal smoking prevalence between mothers whose prenatal or delivery care was Medicaid-funded and mothers whose care was not Medicaid-funded. We evaluated differences in maternal smoking prevalence between these two groups by marital status, race, adequacy of prenatal care, and age. Among the Medicaid-funded mothers, the age-adjusted maternal smoking prevalence was 44.4% versus 16.3% for those not Medicaid-funded. Among married Medicaid-funded mothers, the smoking prevalence was 2.6 times higher in whites, 1.4 times higher in blacks, and 1.8 times higher in American Indians than for married mothers not funded by Medicaid. Among single Medicaid-funded mothers, the rate was 1.4 times higher in whites and 1.7 times higher in blacks. Differences in smoking prevalence were most apparent among older mothers. For single white and single black mothers, the smoking prevalence increased with increasing maternal age among both Medicaid-funded and other women. Adequacy of prenatal care also influences smoking prevalence. For white and black mothers, the maternal smoking prevalence was lower for those receiving adequate prenatal care than for mothers of the same race who received inadequate prenatal care. The increased maternal smoking prevalence in older single mothers and the higher maternal smoking prevalence in women with Medicaid-funded deliveries suggest that infants born to these mothers may be particularly susceptible to smoking-related health effects.

Demographic characteristics, drug use, and sexual behavior of i.v. drug user with AIDS in Bronx, New York.

Intravenous (i.v.) drug users are a key factor in the transmission of human immunodeficiency virus (HIV) infection, yet epidemiologic information about this population, especially those with acquired immunodeficiency syndrome, is scarce. The demographic characteristics, drug use behavior, and sexual practices of i.v. drug users who developed AIDS were prospectively studied at the Montefiore Medical Center from October 1984 to February 1988. The early wave of i.v. drug users with AIDS was characterized by poverty, minority overrepresentation (more than 80 percent were black or Hispanic), and initiation of i.v. drug use at an early age (median age 19 years). Injection of drugs and sharing of needles was frequent. Most had used so-called shooting galleries, but only for a minority of injection episodes. Heroin or cocaine use was almost universal, nearly always accompanied by abuse of another substance, usually alcohol or marijuana. Fewer than a third had ever participated in a methadone maintenance program, but more than 40 percent had been in prison since 1978. All patients had been sexually active, often with partners who were not i.v. drug users. The research suggests a complex interaction existing between high-risk demographic characteristics, drug use practice, and certain types of sexual behavior, all of which contributed to the early spread of HIV infection in this population. Efforts that are directed toward interrupting i.v. drug user-related transmission of HIV need to include consideration of these characteristics.

Cellular and anatomical reservoirs of HIV-1 in patients receiving potent antiretroviral combination therapy.

The eradication of human immunodeficiency virus 1 (HIV-1) from infected persons is the ultimate goal of HIV therapeutic interventions. Great strides have been made in developing potent antiretroviral regimens that greatly suppress HIV-1 replication. Despite these therapeutic advances, major obstacles remain to eradicating HIV-1. Reservoirs of HIV-1 have been identified that represent major impediments to eradication. Conceptually, there are 2 types of sanctuaries for HIV-1, cellular and anatomical. Cellular sanctuaries may include latent CD4+ T cells containing integrated HIV-1 provirus; macrophages, which may express HIV-1 for prolonged periods; and follicular dendritic cells, which may hold infectious HIV-1 on their surfaces for indeterminate lengths of time. The key anatomical reservoir for HIV-1 appears to be the central nervous system. An understanding of the nature of HIV within these reservoirs is critical to devising strategies to hasten viral eradication.

Studies in subjects with long-term nonprogressive human immunodeficiency virus infection.

BACKGROUND: In a small percentage of persons infected with human immunodeficiency virus type 1 (HIV-1), there is no progression of disease and CD4+ T-cell counts remain stable for many years. Studies of the histopathological, virologic, and immunologic characteristics of these persons may provide insight into the pathogenic mechanisms that lead to HIV disease and the protective mechanisms that prevent progression to overt disease. METHODS AND RESULTS: We studied 15 subjects with long-term nonprogressive HIV infection and 18 subjects with progressive HIV disease. Nonprogressive infection was defined as seven or more years of documented HIV infection, with more than 600 CD4+ T cells per cubic millimeter, no antiretroviral therapy, and no HIV-related disease. Lymph nodes from the subjects with nonprogressive infection had significantly fewer of the hyperplastic features, and none of the involuted features, characteristic of nodes from subjects with progressive disease. Plasma levels of HIV-1 RNA and the viral burden in peripheral-blood mononuclear cells were both significantly lower in the subjects with nonprogressive infection than in those with progressive disease (P = 0.003 and P = 0.015, respectively). HIV could not be isolated from the plasma of the former, who also had significantly higher titers of neutralizing antibodies than the latter. There was viral replication, however, in the subjects with nonprogressive infection, and virus was consistently cultured from mononuclear cells from the lymph nodes. In the lymph nodes virus "trapping" varied with the degree of formation of germinal centers, and few cells expressing virus were found by in situ hybridization. HIV-specific cytotoxic activity was detected in all seven subjects with nonprogressive infection who were tested. CONCLUSIONS: In persons who remain free of disease for many years despite HIV infection the viral load is low, but viral replication persists. Lymph-node architecture and immune function appear to remain intact.

Predictors for failure of Pneumocystis carinii pneumonia prophylaxis. Multicenter AIDS Cohort Study.

OBJECTIVE: To identify clinical and epidemiological factors associated with failure of Pneumocystis carinii pneumonia (PCP) prophylaxis in those receiving primary and secondary prophylaxis. DESIGN: Longitudinal cohort study of participants infected with human immunodeficiency virus type 1 in the Multicenter AIDS Cohort Study who used PCP prophylaxis regimens after their T-helper lymphocyte counts had decreased to less than 0.200 x 10(9)/L (200/microL). MAIN OUTCOME MEASURE: Occurrence or recurrence of PCP. RESULTS: A total of 476 participants reported taking one or more of the following regimens: trimethoprim-sulfamethoxazole (TMP-SMX), dapsone, and/or aerosolized pentamidine--367 as primary prophylaxis and 109 as secondary prophylaxis after a previous episode of PCP. A total of 92 (20%) developed PCP despite prophylaxis. The mean failure rates per person-year of follow-up were 16.0% for those receiving primary prophylaxis and 12.1% for those receiving secondary prophylaxis (P = .19). Median times to death after initiation of primary or secondary prophylaxis were 2.0 and 1.2 years, respectively. The main predictor for failure of PCP prophylaxis was profound T-helper lymphocytopenia; 86% of failures occurred after T-helper cell counts decreased to less than 0.075 x 10(9)/L and 76% occurred after counts decreased to less than 0.050 x 10(9)/L. In multivariate time-dependent analysis, when compared with counts between 0.100 x 10(9)/L and 0.200 x 10(9)/L, the risk ratio for failure with counts less than 0.050 x 10(9)/L was 2.90 (P < .001). Once T-helper cell counts were considered, fever was the only other health status indicator that predicted subsequent PCP (ie, a time-dependent risk ratio of 2.22; P = .01). Use of TMP-SMX as the prophylaxis regimen was protective but did not eliminate failure (ie, a time-dependent risk ratio of 0.55; P = .03). CONCLUSIONS: These findings strongly support identifying improved methods of PCP prophylaxis once T-helper cell counts decrease to less than 0.075 x 10(9)/L or 0.100 x 10(9)/L. Given this severe degree of immunosuppression, an inherently more effective regimen against P carinii is required.

HIV infection in homosexual and bisexual men 18 to 29 years of age: the San Francisco Young Men's Health Study.

OBJECTIVES: Recent studies suggest very high human immunodeficiency virus (HIV) infection rates in some populations of younger homosexual men, but these studies may represent only particularly high-risk populations. The current study obtained population-based data on the HIV epidemic in young homosexual/bisexual men. METHODS: A household survey of unmarried men 18 through 29 years of age involved a multistage probability sample of addresses in San Francisco. A follow-up interview and HIV test for men who were HIV negative at baseline were completed; the median follow-up was 8.9 months. RESULTS: Sixty-eight of 380 homosexual/bisexual men (17.9%) tested HIV seropositive. Sixty-three percent of men reported one or more receptive anal intercourse partners in the previous 12 months, and 41% of those men did not use condoms consistently. The HIV seroincidence rate among those seronegative at first study was 2.6% per year. CONCLUSIONS: HIV infection rates in young homosexual men in San Francisco are lower than those in the early 1980s; however, the rate of infection in these men, most of whom became sexually active after awareness of AIDS had become widespread, threatens to continue the epidemic in the younger generation at a level not far below that of a decade ago.

Epidemiologic patterns of upper respiratory illness and Pneumocystis carinii pneumonia in homosexual men.

The relationship between self-reported upper respiratory illness symptoms (URI) and human immunodeficiency virus Type 1 (HIV-1) was examined in homosexual men using semiannual visits from 1984 to 1988. Temporal and geographic patterns of Pneumocystis carinii pneumonia (PCP) diagnosis in these men during the same time period are also described. URI, including acute sinusitis, was reported more often by 916 HIV-1-seropositive participants than by 2,161 seronegative participants (32.21 versus 28.86% p less than 0.001). For 387 seropositive subjects who progressed to acquired immunodeficiency syndrome (AIDS), the proportion reporting URI peaked one visit pre-AIDS at a level significantly higher than matched control subjects (0.45 versus 0.28, p less than or equal to 0.001). The peak was higher for those with PCP as an initial diagnosis. Reported URI peaked in winter and troughed in summer, and PCP diagnosis rates peaked and troughed 4 months later, respectively. Cities with the highest reported rates of URI also had the highest proportions of AIDS cases with PCP as an initial diagnosis. No temporal or geographic patterns were observed for other HIV-1-related symptoms or non-PCP AIDS diagnoses. These patterns suggest the possibility of a person-to-person transmission of P. carinii similar to that of other respiratory pathogens, which would imply a need to consider stricter methods to prevent nosocomial transmission of this pathogen in inpatient and outpatient settings. Further investigation of these issues is needed.

Changes in survival after acquired immunodeficiency syndrome (AIDS): 1984-1991.

In a prospective cohort of 2,647 human immunodeficiency virus type 1 (HIV-1) seropositive homosexual men enrolled in Baltimore, Chicago, Los Angeles, and Pittsburgh, 891 developed clinical acquired immunodeficiency syndrome (AIDS) between June 1984 and January 1992. Cox proportional hazards models were used to examine temporal trends in survival after AIDS for specific diagnoses, controlling for level of immunosuppression at diagnosis, age, race, and geographic location. Median survival time following AIDS onset increased from 11.6 months in 1984-1985 to 19.5 months in 1988-1989; for those diagnosed in 1990-1991, the median survival time dropped to 17.2 months. Trends in improved survival were diagnosis-specific. Survival after Pneumocystis carinii pneumonia consistently improved from 1984 to 1991 (p < 0.001). Compared with men diagnosed in 1984-1985, those diagnosed with P. carinii pneumonia in 1990-1991 had one-tenth the hazard of dying. For men with > or = 100 helper T-lymphocytes (CD4+ cells) when diagnosed with Kaposi's sarcoma, the relative hazards (95% confidence intervals) of dying after Kaposi's sarcoma were 0.8 (0.42-1.60) in 1986-1987, 0.7 (0.34-1.58) in 1988-1989, and 0.6 (0.19-1.61) in 1990-1991 compared with those diagnosed before 1986. Men with < 100 CD4+ cells when diagnosed with Kaposi's sarcoma did not demonstrate a consistent change in their subsequent survival. After a nonsignificant (p > 0.05) initial improvement in prognosis, there has not been a significant improvement in survival for men who presented with other opportunistic infections. Observed increases in overall survival probably relate to improved treatment of patients who develop P. carinii pneumonia. Limited improvement in survival following other AIDS diagnoses indicates the need for developing effective treatment against these diseases.

Trends in the incidence of outcomes defining acquired immunodeficiency syndrome (AIDS) in the Multicenter AIDS Cohort Study: 1985-1991.

Incidence of clinical outcomes defining acquired immunodeficiency syndrome (AIDS) may be expected to change as a consequence of progressive immunosuppression and use of chemoprophylaxis before the onset of AIDS. Using Poisson regression methods, we examined trends in the incidence of initial and secondary AIDS-defining illnesses from 1985 to 1991 among 2,627 homosexual men participating in the Multicenter AIDS Cohort Study who were seropositive for human immunodeficiency virus type 1. The incidence of Pneumocystis carinii pneumonia rose steeply until 1987 but has declined since then (p < 0.001), while the other AIDS-defining conditions have showed significant (p < or = 0.039) upward trends. Trends for Kaposi's sarcoma, lymphoma, neurologic disease, and cytomegalovirus/herpes simplex virus infections were explained by progressive immunosuppression, but residual downward and upward trends were present for P. carinii pneumonia and other opportunistic infections (bacterial, fungal, and protozoal infections and wasting syndrome). Despite selection bias, those receiving P. carinii pneumonia chemoprophylaxis showed a significantly lower incidence of P. carinii pneumonia (relative risk = 0.32, 95% confidence interval 0.16-0.63), and the time trends of P. carinii pneumonia were explained by progressive immunosuppression and use of prophylaxis. No significant effects on all other diagnoses were seen in those selected to receive antiretroviral therapy. Secondary diagnoses showed a strongly significant (p < 0.001) increase in non-P. carinii pneumonia and non-Kaposi's sarcoma among those with initial diagnoses of Kaposi's sarcoma. Overall, the trend observed in the incidence of other opportunistic infections underscores the need for developing and testing new strategies to curtail or delay the onset of these diseases.