Search

Background: In 2007, Swaziland initiated a hub-and-spoke model for decentralizing antiretroviral therapy (ART) access for HIV-infected children (<15 years old). Decentralization was facilitated through (1) down referral of stable children on ART from overburdened central facilities (hubs) to primary healthcare clinics (spokes) and (2) pediatric ART initiation at spokes (spoke initiation)., Methods: We conducted a nationally representative retrospective cohort study among children starting ART during 2004-2010 to assess effect of down referral and spoke initiation on rates of loss to follow-up (LTFU), death and attrition (death or LTFU). Twelve of 28 pediatric ART hubs were randomly selected using probability-proportional-to-size sampling. Seven selected facilities had initiated hub-and-spoke decentralization by study start; at these facilities, 901 of 1893 hub-initiated and maintained (hub-maintained) children and 495 of 1105 down-referred or spoke-initiated children were randomly selected for record abstraction. At the 5 hub-only facilities, 612 of 1987 children were randomly selected. Multivariable proportional hazards regression was used to estimate adjusted hazard ratios (AHR) for effect of down referral (a time-varying covariate) and spoke initiation on outcomes., Results: Among 2008 children at ART initiation, median age was 5.0 years, median CD4% 12.0%, median CD4 count 358 cells/µL and median weight-for-age Z score -1.91. Controlling for known confounders, down referral was strongly protective against LTFU (AHR: 0.40; 95% confidence interval: 0.20-0.79) and attrition (AHR: 0.46; 95% confidence interval: 0.26-0.83) but not mortality. Compared with hub-only children or hub-maintained children, spoke-initiated children had similar outcomes., Conclusions: Decentralization of pediatric ART through down referral and spoke initiation within a hub-and-spoke system should be continued and might improve program outcomes.

Background: New York City (NYC) pediatricians are now caring for fewer HIV-infected infants and more school age children and adolescents than earlier in the epidemic., Methods: Clinical, laboratory and demographic data were abstracted from medical records at 10 NYC centers participating in the CDC Pediatric Spectrum of HIV Disease project. Pediatric AIDS cases and HIV-related deaths reported to the NYC Department of Health were examined., Results: Median age of HIV-infected children in care increased from 3 years in 1989 to 1991 to 6 years in 1995 to 1998. The number of HIV-infected women giving birth in NYC declined 50% from 1990 to 1997 (1630 to 831); increasing numbers were identified prenatally (14% in 1989; 78% after 1995); and most received prenatal zidovudine prophylaxis (73% in 1997). Estimated perinatal transmission decreased to 10% by 1997. Improved identification of seropositive status in infants was associated with an increased proportion of infected infants receiving Pneumocystis carinii pneumonia (PCP) prophylaxis, 84% in 1997. AIDS free survival was longer for children born 1995 to 1998 than for those born before 1995, P = 0.004. In 1998 among children with advanced immunosuppression (CDC category 3), 66% were prescribed 3 or more antiretroviral medicines and 88% received PCP prophylaxis. Citywide AIDS cases and HIV-related deaths fell precipitously beginning in 1996., Conclusions: Based on the observations of this study, the cohort of NYC HIV-infected children in care is aging, associated with a decline in new HIV infections, high rates of PCP prophylaxis and increased time to AIDS. Falling HIV-related deaths citywide support these observations.

Background: Tuberculosis disease incidence increased sharply in New York City (NYC) in the late 1980s in children and adults. The relationship of tuberculosis disease in adults with the coincident epidemic of immunosuppression caused by HIV disease has been well-documented. This paper examines the relationship of tuberculosis and HIV in children in NYC., Methods: Information on tuberculosis was collected by retrospective chart abstraction in a cohort of HIV-exposed and infected children enrolled in a longitudinal study of HIV. Tuberculosis cases were ascertained by chart review or by matching HIV-infected and -exposed children to NYC Tuberculosis Registry cases. NYC Tuberculosis Registry data on children reported from 1989 to 1995, and not reported as HIV-infected, were used for comparison., Results: Tuberculosis disease was found in 45 (3%) of 1426 HIV-infected children (0.61 per 100 child years of observation) and in 5 (0.5%) of 1085 HIV-exposed uninfected children (0.2 per 100 child years). 30% of children were evaluated for HIV only after presenting with tuberculosis. Children with tuberculosis and HIV were more likely than other age-matched HIV-infected children to have decreased CD4+ T lymphocyte counts (66% vs. 37%, P = 0.02) and more likely than other NYC children with tuberculosis to have culture-confirmed and extrapulmonary tuberculosis. In this series 8 of 21 deaths in HIV-infected children with tuberculosis appeared to be related to tuberculosis., Conclusions: During a period of high tuberculosis incidence in NYC, 3% of HIV-infected children in our cohort had tuberculosis, higher than the rate in uninfected children born to HIV-positive mothers in the same cohort. Because of this association, HIV-infected children with pulmonary illness should be tested for tuberculosis; and all children with tuberculosis should be tested for HIV.

Objective: To determine the incidence of HIV-1-related clinical findings, mortality and predictors of death in a cohort of HIV-exposed infants followed from birth., Methods: Data were collected approximately bimonthly during the first and second year of life and used in Kaplan-Meier and Cox proportional hazards survival analyses to predict time to the development of symptoms and death., Results: One hundred sixteen infected and 396 uninfected infants were followed for a median of 26 months at 7 New York City hospitals from 1986 to 1995. Two or more nonspecific HIV-related symptoms, AIDS or death occurred in 83% of infected children by the first year. Fifty infected infants (43%) developed AIDS and 19 (38%) of these had Pneumocystis carinii pneumonia. Estimated median age at AIDS/death was 30 months and 64% of infected children remained alive and AIDS-free at 1 year. Estimated infant mortality among infected children was 160/1000 live births, and median survival after AIDS was 21 months; 55% of infected children survived > 12 months after diagnosis of AIDS. P. carinii pneumonia was the most common cause of death. Although birth CD4 values did not predict AIDS or death, CD4 counts as early as 6 months of age were highly correlated with both. Thirteen (68%) of 19 infants who remained AIDS-free up to 3 to 6 months of age with CD4 count < or = 1500 cells/microliters subsequently developed AIDS vs. 18 (30%) of 61 with CD4 count > 1500 (P = 0.0001)., Conclusions: Most HIV-1-infected infants develop disease in the first year of life. AIDS or death can be predicted by a threshold CD4 count of 1500 cells/microliters at 3 to 6 months of age.

Three human immunodeficiency virus type 1 (HIV-1)-exposed children who had repeatedly positive DNA polymerase chain reaction (PCR) tests for HIV in > or = 5 samples before seroreversion to HIV-negative status are reported. The children belong to a cohort of 210 infants who were born to HIV-infected mothers and were tested at intervals of 1 to 3 months by HIV viral culture, PCR, and p24 antigen; only the PCR was positive in > or = 5 samples in the children reported here. Their clinical features were indistinguishable from other seroreverters. All three children had a transient drop in CD4:CD8 ratio to < 1.0. The transiently positive DNA PCR in HIV-exposed infants may indicate either that HIV infection was eliminated by a strong host immune response or that infection was caused by an attenuated/defective strain of virus.

This analysis sought to identify characteristics of pregnant human immunodeficiency virus type 1 (HIV-1)-infected women that predict mother-to-child HIV-1 transmission. Pregnant and immediately postpartum women at risk for HIV were enrolled at obstetric and pediatric care settings in New York City from 1986 to 1992. Demographic and behavioral characteristics, clinical illness, T lymphocyte subsets, immunoglobulin concentration and syphilis serology were collected on the women. Infants were followed to determine HIV infection classification according to Centers for Disease Control and Prevention criteria for HIV-1 in children. Transmission rates were calculated for women who gave birth more than 15 months before the analysis. Of 172 HIV-1-infected women with known outcome 49 (28%) had infected infants. The transmission rate (TR) was significantly higher among women with < 280 CD4+ cells/microliters (lowest CD4+ quartile) than with CD4+ counts > 280 (48% vs. 22%; P = 0.004; odds ratio, 3.4; 95% confidence interval (1.5, 7.8)); a similar trend was seen by CD4+% quartile. No difference in TR was seen comparing women by CD8+ count quartile but marginally higher TR was seen among women with CD8+% > or = 51% than with CD8+% < 51% (TR = 41% vs. 24%; P = 0.076; odds ratio, 2.2; confidence interval (1.0, 5.1)). The highest TR, 62% was seen in women with both CD8+ count above the median and CD4+ count in the lowest quartile. No significant difference in TR was seen between women with and without HIV-related illness, although the TR was 53% among women hospitalized in the previous year for pneumonia compared with 25% in others (P = 0.03). TR was somewhat lower in women who delivered by cesarean section than vaginally (entire cohort: 18% vs. 32%, P = 0.11; prenatal enrollees only, 17% vs. 38%, P = 0.045). No factor or combination of factors was both highly sensitive and specific for predicting mother-to-child HIV transmission. A possible relationship between transmission and mode of delivery deserves further investigation.

A simplified human immunodeficiency virus 1 (HIV-1)-specific IgA capture enzyme immunoassay (IgA-CEIA) was evaluated and compared with IgA-Western blot assay for early diagnosis of HIV-1 infection in infants born to seropositive women. A total of 232 coded sera collected prospectively from 70 infants were tested. All 25 sera from 10 HIV-1-negative infants born to seronegative mothers (negative controls) were negative by both assays. All 111 sera from 37 seroreverting, uninfected infants were negative by IgA-CEIA (specificity, 100%), whereas 110 of 111 sera were negative by IgA-Western blot assay (specificity, > 99%). Overall IgA-CEIA detected HIV-IgA in 20 (87%) of 23 infected infants, and IgA-Western blot assay detected HIV-IgA in 21 (91.3%) of 23 infants; specimen-wise agreement between the 2 assays was > 80%. Analysis of results by age group indicated that after 2 months of age both assays were equivalent with sensitivity ranging from 60 to 80%. Quantitative data provided by IgA-CEIA suggests that the bulk of HIV-1 IgA synthesis in most HIV-1-infected infants occurs after 2 months of age.

More than one-half of the children born to women with human immunodeficiency virus (HIV) infection are not infected with HIV. Most of these children, although born antibody-positive, lose maternal antibody and remain asymptomatic. However, it has been unclear how many antibody-negative children of HIV-infected women may truly be infected despite the loss of passively acquired maternal antibody. One hundred nine children who lost maternal antibody after birth to HIV-infected women recruited in four United States maternal HIV transmission studies were examined for HIV infection. Polymerase chain reaction (PCR) was used to determine whether children had HIV proviral DNA in peripheral blood mononuclear cells. A total of 268 samples from 109 children were tested. Clinical status and other laboratory findings were also evaluated. The median age at last follow-up was 25 months (range, 12 to 48 months). One hundred seven (98%) children were negative by PCR on all samples tested. None (95% confidence interval, 0.0 to 1.9%) of 109 children had a repeatedly positive PCR. Two children had single positive PCR results that could not be confirmed on subsequent testing. No other laboratory or clinical findings supported HIV infection in either of these children. The loss of HIV antibody in an asymptomatic child born to an HIV-infected woman strongly suggests that the child is not infected with HIV. Single positive PCR results, particularly in the absence of other clinical or laboratory evidence of HIV infection, should not be used alone to diagnose HIV infection.