The HPTN 046 trial evaluated the efficacy of extended infant nevirapine (NVP) administration for prevention of HIV transmission through breastfeeding. Infants received daily NVP to 6 weeks of age. HIV-uninfected infants (the intent-to-treat group) received daily NVP or placebo up to 6 months of age. We analyzed emergence of NVP resistance in infants who acquired HIV-infection despite prophylaxis.
HIV genotyping was performed using the ViroSeq HIV Genotyping System. Medians and proportions were used to summarize data. Two-sided Fisher’s exact tests were used to evaluate associations between categorical variables.
NVP resistance was detected in 12 (92.3%) of 13 infants who were HIV-infected by 6 weeks and in seven (28%) of 25 infants who were HIV-uninfected at 6 weeks and HIV-infected at 6 months of age (6/8=75% in the NVP arm, 1/17=5.9% in the placebo arm, P=0.001). Among those 25 infants, 4 had mothers who initiated an antiretroviral (ARV) treatment regimen by 6 months postpartum. In all 4 cases, the treatment regimen included a non-nucleoside reverse transcriptase inhibitor (NVP or efavirenz). NVP resistance was detected in all four of those infants by 6 months of age (4/4=100%). In contrast, only three (14.2%) of the remaining 21 HIV-infected infants whose mothers did not initiate ARV treatment developed NVP resistance (P=0.003).
Extended NVP prophylaxis significantly increased the risk of NVP resistance in infants who acquired HIV infection after 6 weeks of age. Treatment of maternal HIV infection was also associated with emergence of NVP resistance in HIV-infected, breastfed infants.
Nevirapine resistance; prevention of mother-to-child transmission; extended nevirapine; HIV
Assays for detecting levels of antiretroviral drugs in study participants are increasingly popular in preexposure prophylaxis (PrEP) trials, since they provide an objective measure of adherence. Current correlation analyses of drug concentration data are prone to bias. In this article, we formulate the causal estimand of prevention efficacy among drug compliers, those who would have had a threshold level of drug concentration had they been assigned to the drug arm of the trial. The identifiability of the causal estimand is facilitated by exploiting the exclusion restriction; that is, drug noncompliers do not acquire any prevention benefit. In addition, we develop an approach to sensitivity analysis that relaxes the exclusion restriction. Applications to published data from 2 PrEP trials, namely the Preexposure Prophylaxis Initiative (iPrEx) trial and the Centre for the AIDS Programme of Research in South Africa (CAPRISA) 004 trial, suggest high efficacy estimates among drug compliers (in the iPrEx trial, odds ratio = 0.097 (95% confidence interval: 0.027, 0.352); in the CAPRISA 004 trial, odds ratio = 0.104 (95% confidence interval: 0.024, 0.447)). In summary, the proposed inferential method provides an unbiased assessment of PrEP efficacy among drug compliers, thus adding to the primary intention-to-treat analysis and correlation analyses of drug concentration data.
causal inference; compliance; exclusion restriction; potential outcome; principal stratification; two-phase sampling
The first antiretroviral drug (Truvada) to be used as a pre-exposure prophylaxis (PrEP) in preventing HIV transmission is about to be approved. Behavioral studies suggest that a portion of users may share anti-retroviral drugs with sex partners, family, or friends. Pill sharing will decrease PrEP efficacy and adherence level, and potentially create an environment favorable for the development of drug resistance. We aim to evaluate the potential impact of pill sharing on the PrEP effectiveness and on the rates of drug-resistance development in heterosexual populations.
A transmission dynamic model was used to assess the population-level impact of oral PrEP. The fractions of new HIV infections prevented (CPF), drug resistance prevalence and the proportion of new infections in which drug-resistant HIV is transmitted (TDR) are evaluated over fixed time periods. The influence of different factors on CPF and TDR is studied through simulations, using epidemic parameters representative of the countries in Sub-Saharan Africa.
Without pill sharing, a 70% efficacious PrEP used consistently by 60% of uninfected individuals prevents 52.8% (95% CI 49.4%-56.4%) of all new HIV infections over ten years with drug-resistant HIV transmitted in 2.2% of the new infections. Absolute CPF may vary by 9% if up to 20% of the users share PrEP while the level of TDR and total resistance prevalence may increase by up to 6-fold due to pill sharing in some intervention scenarios.
Pill sharing may increase the PrEP coverage level achieved in the population but it also affects the PrEP efficacy for the users who do not follow the prescribed schedule. More importantly, it creates a pool of untracked users who remain unreached by the effort to avoid sub-optimal PrEP usage by infected people. This increases substantially the potential risk of drug resistance in the population.
HIV; Pill sharing; Pre-exposure prophylaxis
Four clinical trials have shown that oral and topical pre-exposure prophylaxis (PrEP) based on tenofovir may be effective in preventing HIV transmission. The expected reduction in HIV transmission and the projected prevalence of drug resistance due to PrEP use vary significantly across modeling studies as a result of the broad spectrum of assumptions employed. Our goal is to quantify the influence of drug resistance assumptions on the predicted population-level impact of PrEP.
All modeling studies which evaluate the impact of oral or topical PrEP are reviewed and key assumptions regarding mechanisms of generation and spread of drug-resistant HIV are identified. A dynamic model of the HIV epidemic is developed to assess and compare the impact of oral PrEP using resistance assumptions extracted from published studies. The benefits and risks associated with ten years of PrEP use are evaluated under identical epidemic, behavioral and intervention conditions in terms of cumulative fractions of new HIV infections prevented, resistance prevalence among those infected with HIV, and fractions of infections in which resistance is transmitted.
Published models demonstrate enormous variability in resistance-generating assumptions and uncertainty in parameter values. Depending on which resistance parameterization is used, a resistance prevalence between 2% and 44% may be expected if 50% efficacious oral PrEP is used consistently by 50% of the population over ten years. We estimated that resistance may be responsible for up to a 10% reduction or up to a 30% contribution to the fraction of prevented infections predicted in different studies.
Resistance assumptions used in published studies have a strong influence on the projected impact of PrEP. Modelers and virologists should collaborate toward clarifying the set of resistance assumptions biologically relevant to the PrEP products which are already in use or soon to be added to the arsenal against HIV.
There are few studies of the association between placental malaria (PM) and mother-to-child transmission (MTCT) of human immunodeficiency virus-1 (HIV-1), and the results of published studies are inconsistent. To determine the association between PM and MTCT of HIV-1, we performed a secondary analysis of data from a clinical trial of antibiotics to reduce chorioamnionitis. Data regarding 1,662 HIV-1–infected women with live born singleton and first-born twin infants with information regarding PM and infant HIV-1 infection status at birth were analyzed. At the time of the study, women did not have access to antiretroviral drugs for treatment of acquired immunodeficiency syndrome but had received nevirapine prophylaxis to reduce the risk of MTCT of HIV-1. Placental malaria was not associated with the infant HIV-1 infection status at birth ( P = 0.67). Adjustment for maternal plasma viral load and CD4+ cell count did not change these results (odds ratio = 1.06, 95% confidence interval = 0.51–2.20, P = 0.87). Placental malaria was more likely to be related to HIV-1 infection at birth among women with low viral load at baseline (P for interaction = 0.08). In conclusion, PM was not associated with infant HIV-1 infection status at birth. The interaction of maternal plasma viral load, PM, and MTCT of HIV-1 warrants further studies.
HIV prevalence among pregnant women in Southern Africa is extremely high. Epidemiological studies suggest that pregnancy increases the risk of HIV sexual acquisition and that HIV infections acquired during pregnancy carry higher risk of mother-to-child transmission (MTCT). We analyze the potential benefits from extending the availability of effective microbicide to pregnant women (in addition to non-pregnant women) in a wide-scale intervention.
Methods and Findings
A transmission dynamic model was designed to assess the impact of microbicide use in high HIV prevalence settings and to estimate proportions of new HIV infections, infections acquired during pregnancy, and MTCT prevented over 10 years. Our analysis suggests that consistent use of microbicide with 70% efficacy by 60% of non-pregnant women may prevent approximately 40% and 15% of new infections in women and men respectively over 10 years, assuming no additional increase in HIV risk to either partner during pregnancy (RRHIV/preg = 1). It may also prevent 8–15% MTCT depending on the increase in MTCT risk when HIV is acquired during pregnancy compared to before pregnancy (RRMTCT/preg). Extending the microbicides use during pregnancy may improve the effectiveness of the intervention by 10% (RRHIV/preg = 1) to 25% (RRHIV/preg = 2) and reduce the number of HIV infections acquired during pregnancy by 40% to 70% in different scenarios. It may add between 6% (RRHIV/preg = 1, RRMTCT/preg = 1) and 25% (RRHIV/preg = 2, RRMTCT/preg = 4) to the reduction in the residual MTCT.
Providing safe and effective microbicide to pregnant women in the context of wide-scale interventions would be desirable as it would increase the effectiveness of the intervention and significantly reduce the number of HIV infections acquired during pregnancy. The projected benefits from covering pregnant women by the HIV prevention programs is more substantial in communities in which the sexual risk during pregnancy is elevated.
The C868T single nucleotide polymorphism in the CD4 receptor encodes an amino acid substitution of tryptophan for arginine in the third domain. Previous studies suggest that C868T increases the risk of HIV-1 acquisition; however, the influence of this single nucleotide polymorphism (SNP) on disease progression has not been established. The presence of the C868T polymorphism was not statistically significantly associated with HIV-1 disease progression outcomes in a cohort of postpartum Kenyan women.
Because lubricants may decrease trauma during coitus, it is hypothesized that they could aid in the prevention of HIV acquisition. Therefore, safety and anti-HIV-1 activity of over-the-counter (OTC) aqueous- (n = 10), lipid- (n = 2), and silicone-based (n = 2) products were tested. The rheological properties of the lipid-based lubricants precluded testing with the exception of explant safety testing. Six aqueous-based gels were hyperosmolar, two were nearly iso-osmolar, and two were hypo-osmolar. Evaluation of the panel of products showed Gynol II (a spermicidal gel containing 2% nonoxynol-9), KY Jelly, and Replens were toxic to Lactobacillus. Two nearly iso-osmolar aqueous- and both silicone-based gels were not toxic toward epithelial cell lines or ectocervical or colorectal explant tissues. Hyperosmolar lubricants demonstrated reduction of tissue viability and epithelial fracture/sloughing while the nearly iso-osmolar and silicon-based lubricants showed no significant changes in tissue viability or epithelial modifications. While most of the lubricants had no measurable anti-HIV-1 activity, three lubricants which retained cell viability did demonstrate modest anti-HIV-1 activity in vitro. To determine if this would result in protection of mucosal tissue or conversely determine if the epithelial damage associated with the hyperosmolar lubricants increased HIV-1 infection ex vivo, ectocervical tissue was exposed to selected lubricants and then challenged with HIV-1. None of the lubricants that had a moderate to high therapeutic index protected the mucosal tissue. These results show hyperosmolar lubricant gels were associated with cellular toxicity and epithelial damage while showing no anti-viral activity. The two iso-osmolar lubricants, Good Clean Love and PRÉ, and both silicone-based lubricants, Female Condom 2 lubricant and Wet Platinum, were the safest in our testing algorithm.
The timing of mother-to-child transmission (MTCT) of HIV is critical in understanding the dynamics of MTCT. It has a great implication to developing any effective treatment or prevention strategies for such transmissions. In this paper, we develop an imputation method to analyze the censored MTCT timing in presence of auxiliary information. Specifically, we first propose a statistical model based on the hazard functions of the MTCT timing to reflect three MTCT modes: in utero, during delivery and via breastfeeding, with different shapes of the baseline hazard that vary between infants. This model also allows that the majority of infants may be immuned from the MTCT of HIV. Then, the model is fitted by MCMC to explore marginal inferences via multiple imputation. Moreover, we propose a simple and straightforward approach to take into account the imperfect sensitivity in imputation step, and study appropriate censoring techniques to account for weaning. Our method is assessed by simulations, and applied to a large trial designed to assess the use of antibiotics in preventing MTCT of HIV.
HIV/AIDS; mixture models; mother to child transmission of HIV; multiple imputation
The integrated discrimination improvement (IDI) index is a popular tool for evaluating the capacity of a marker to predict a binary outcome of interest. Recent reports have proposed that the IDI is more sensitive than other metrics for identifying useful predictive markers. In this article, the authors use simulated data sets and theoretical analysis to investigate the statistical properties of the IDI. The authors consider the common situation in which a risk model is fitted to a data set with and without the new, candidate predictor(s). Results demonstrate that the published method of estimating the standard error of an IDI estimate tends to underestimate the error. The z test proposed in the literature for IDI-based testing of a new biomarker is not valid, because the null distribution of the test statistic is not standard normal, even in large samples. If a test for the incremental value of a marker is desired, the authors recommend the test based on the model. For investigators who find the IDI to be a useful measure, bootstrap methods may offer a reasonable option for inference when evaluating new predictors, as long as the added predictive capacity is large.
biological markers; bootstrap confidence interval; prediction; risk assessment; sampling distribution; sampling error; selection bias; type I error
There is conflicting evidence regarding the effects of breast-feeding on maternal mortality from human immunodeficiency virus type 1 (HIV-1) infection, and little is known about the effects of breast-feeding on markers of HIV-1 disease progression.
HIV-1–seropositive women were enrolled during pregnancy and received short-course zidovudine. HIV-1 RNA levels and CD4 cell counts were determined at baseline and at months 1, 3, 6, 12, 18, and 24 postpartum and were compared between breast-feeding and formula-feeding mothers.
Of 296 women, 98 formula fed and 198 breast-fed. At baseline, formula-feeding women had a higher education level and prevalence of HIV-1–related illness than did breast-feeding women; however, the groups did not differ with respect to CD4 cell counts and HIV-1 RNA levels. Between months 1 and 24 postpartum, CD4 cell counts decreased 3.9 cells/µL/month (P< .001), HIV-1 RNA levels increased 0.005 log10 copies/mL/month (P = .03), and body mass index (BMI) decreased 0.03 kg/m2/month (P< .001). The rate of CD4 cell count decline was higher in breast-feeding mothers (7.2 cells/µL/month) than in mothers who never breast-fed (4.0 cells/µL/month) (P = .01). BMI decreased more rapidly in breast-feeding women (P = .04), whereas HIV-1 RNA levels and mortality did not differ significantly between breast-feeding and formula-feeding women.
Breast-feeding was associated with significant decreases in CD4 cell counts and BMI. HIV-1 RNA levels and mortality were not increased, suggesting a limited adverse impact of breast-feeding in mothers receiving extended care for HIV-1 infection.
Much of the burden of morbidity affecting women of childbearing age in sub-Saharan Africa occurs in the context of HIV-1 infection. Understanding patterns of illness and determinants of disease in HIV-1–infected mothers may guide effective interventions to improve maternal health in this setting.
We describe the incidence and cofactors of comorbidities affecting peripartum and postpartum HIV-1–infected women in Kenya. Women were evaluated by clinical examination and standardized questionnaires during pregnancy and for up to 2 years after delivery.
Five hundred thirty-five women were enrolled in the cohort (median CD4 count of 433 cells/mm3) and accrued 7736 person-months of follow-up. During 1-year follow-up, the incidence of upper respiratory tract infections was 161 per 100 person-years, incidence of pneumonia was 33 per 100 person-years, incidence of tuberculosis (TB) was 11 per 100 person-years, and incidence of diarrhea was 63 per 100 person-years. Immunosuppression and HIV-1 RNA levels were predictive for pneumonia, oral thrush, and TB but not for diarrhea; CD4 counts <200 cells/mm3 were associated with pneumonia (relative risk [RR] = 2.87, 95% confidence interval [CI]: 1.71 to 4.83), TB (RR = 7.14, 95% CI: 2.93 to 17.40) and thrush. The risk of diarrhea was significantly associated with crowding (RR = 1.86, 95% CI: 1.19 to 2.92) and breast-feeding (RR = 1.71, 95% CI: 1.19 to 2.44). Less than 10% of women reported hospitalization during 2-year follow-up; mortality risk in the cohort was 1.9% and 4.8% for 1 and 2 years, respectively.
Mothers with HIV-1, although generally healthy, have substantial morbidity as a result of common infections, some of which are predicted by immune status or by socioeconomic factors. Enhanced attention to maternal health is increasingly important as HIV-1–infected mothers transition from programs targeting the prevention of mother-to-child transmission to HIV care clinics.
HIV/AIDS; HIV-1 progression; maternal health; morbidity; postpartum; pregnancy; prevention of mother-to-child transmission; women
Excessive non-subcutaneous fat deposition may impair the functions of surrounding tissues and organs through the release of inflammatory cytokines and free fatty acids.
We examined the cross-sectional association between non-subcutaneous adiposity and calcified coronary plaque, a non-invasive measure of coronary artery disease burden.
Participants in the Multi-Ethnic Study of Atherosclerosis underwent CT assessment of calcified coronary plaque. We measured multiple fat depots in 398 white and black participants (47% men and 43% black), ages 47–86 years, from Forsyth County, NC during 2002–2005, using cardiac and abdominal CT scans. In addition to examining each depot separately, we also created a non-subcutaneous fat index using the standard scores of non-subcutaneous fat depots.
A total of 219 participants (55%) were found to have calcified coronary plaque. After adjusting for demographics, lifestyle factors and height, calcified coronary plaque was associated with a one standard deviation increment in the non-subcutaneous fat index (OR = 1.41; 95% CI: 1.08, 1.84), pericardial fat (OR = 1.38; 95% CI: 1.04, 1.84), abdominal visceral fat (OR = 1.35; 95% CI: 1.03, 1.76), but not with fat content in the liver, intermuscular fat, or abdominal subcutaneous fat. The relation between non-subcutaneous fat index and calcified coronary plaque remained after further adjustment for abdominal subcutaneous fat (OR = 1.40; 95% CI: 1.00, 1.94). The relation did not differ by gender and ethnicity.
The overall burden of non-subcutaneous fat deposition, but not abdominal subcutaneous fat, may be a correlate of coronary atherosclerosis.
We present a Bayesian model to estimate the time-varying sensitivity of a diagnostic assay when the assay is given repeatedly over time, disease status is changing and the gold standard is only partially observed. The model relies on parametric assumptions for the distribution of the latent time of disease onset and the time-varying sensitivity. Additionally, we illustrate the incorporation of historical data for constructing prior distributions. We apply the new methods to data collected in a study of mother-to-child transmission of HIV and include a covariate for sensitivity to assess whether two different assays have different sensitivity profiles.
Bayesian models; mother-to-child transmission of HIV; Time-varying sensitivity
As prevention of mother-to-child transmission of HIV (PMTCT) programs and HIV treatment programs rapidly expand in parallel, it is important to determine factors that influence the transition of HIV-infected women from maternal to continuing care.
This study aimed to determine rates and co-factors of accessing HIV care by HIV-infected women exiting maternal care. A cross-sectional survey of women who had participated in a PMTCT research study and were referred to care programs in Nairobi, Kenya was conducted.
A median of 17 months following referral, women were located by peer counselors and interviewed to determine whether they accessed HIV care and what influenced their care decisions. Fisher’s exact test was used to assess the association between client characteristics and access to care.
Peer counselors traced 195 (82%) residences, where they located 116 (59%) participants who provided information on care. Since exit, 50% of participants had changed residence, and 74% reported going to the referral HIV program. Reasons for not accessing care included lack of money, confidentiality, and dislike of the facility. Women who did not access care were less likely to have informed their partner of the referral (p=0.001), and were less likely believe that highly active antiretroviral therapy (HAART) is effective (p<0.01). Among those who accessed care, 33% subsequently discontinued care, most because they did not qualify for HAART. Factors cited as barriers to access included stigma, denial, poor services, and lack of money. Factors that were cited as making care attractive included health education, counseling, free services, and compassion.
A substantial number of women exiting maternal care do not transit to HIV care programs. Partner involvement, a standardized referral process and more comprehensive HIV education for mothers diagnosed with HIV during pregnancy may facilitate successful transitions between PMTCT and HIV care programs.
PMTCT; access; HIV
Co-infection with herpes simplex virus type 2 (HSV-2) has been associated with increased HIV-1 RNA levels and immune activation, two predictors of HIV-1 progression. The impact of HSV-2 on clinical outcomes among HIV-1 infected pregnant women is unclear.
HIV-1 infected pregnant women in Nairobi were enrolled antenatally and HSV-2 serology was obtained. HIV-1 RNA and CD4 count were serially measured for 12–24 months postpartum. Survival analysis using endpoints of death, opportunistic infection (OI), and CD4<200 cells µL, and linear mixed models estimating rate of change of HIV-1 RNA and CD4, were used to determine associations between HSV-2 serostatus and HIV-1 progression.
Among 296 women, 254 (86%) were HSV-2-seropositive. Only 30 (10%) women had prior or current genital ulcer disease (GUD); median baseline CD4 count was 422 cells µL. Adjusting for baseline CD4, women with GUD were significantly more likely to have incident OIs (adjusted hazard ratio (aHR) 2.79, 95% CI: 1.33–5.85), and there was a trend for association between HSV-2-seropositivity and incident OIs (aHR 3.83, 95% CI: 0.93–15.83). Rate of change in CD4 count and HIV-1 RNA did not differ by HSV-2 status or GUD, despite a trend toward higher baseline HIV-1 RNA in HSV-2-seropositive women (4.73 log10 copies/ml vs. 4.47 log10 copies/ml, P = 0.07).
HSV-2 was highly prevalent and pregnant HIV-1 infected women with GUD were significantly more likely to have incident OIs than women without GUD, suggesting that clinically evident HSV-2 is a more important predictor of HIV-1 disease progression than asymptomatic HSV-2.
Assess population attributable fractions (PAFs) for late postnatal transmission (LPT) of human immunodeficiency virus-1 (HIV-1) in a cohort of HIV-1-exposed infants.
We used data established from a risk factor analysis of LPT (negative HIV-1 results through the 4-6 week visit, but positive assays thereafter through the 12-month visit) from a perinatal clinical trial conducted in three sub-Saharan countries. PAFs were calculated as the proportions of excess LPTs attributed to identified risk factors.
For the cohort of 1317 infants, 206 (15.6%) had only low maternal CD4+ counts (< 200 cells/mm3), 332 (25.2%) had only high maternal plasma viral loads (VLs) (> 50 000 copies/mL), and 81 (6.2%) had both low CD4+ counts and high VLs. Their PAFs were 26.0% [95% confidence interval (CI), 12.0%-36.0%], 37.0% (95% CI, 22.0%-51.0%) and 16.0% (95% CI, 6.0%-25.0%), respectively.
Our PAF analysis illustrates the public health impact of the substantial proportion of LPTs accounted for by high-risk women with both low CD4+ counts and high VLs. In light of these results, access to and use of antiretroviral therapy (ART) by high-risk HIV-1-infected pregnant women is essential. Additional strategies to reduce LPT for those not meeting criteria for ART should be implemented.
Breast feeding; late postnatal transmission; prevention of mother to child transmission/vertical transmission; risk factors; viral load
Inflammatory markers predict coronary heart disease (CHD). However, associations with coronary artery calcium (CAC), a marker of subclinical CHD, are not established.
We examined cross-sectional associations of C-reactive protein (CRP), interleukin-6 (IL-6) and fibrinogen with CAC presence (Agatston score > 0 by computed tomography) in 6,783 Multi-Ethnic Study of Atherosclerosis (MESA) participants.
In all participants, those in the highest, compared to lowest, quartile of CRP had a relative risk (RR, 95% confidence interval) of 1.13 (1.06-1.19; p<0.01) for CAC in age, sex and ethnicity adjusted models. For highest versus lowest quartiles, relative risks were 1.22 (1.15-1.30; p<0.01) for IL-6 and 1.18 (1.11-1.24; p<0.01) for fibrinogen. Adjusting for CHD risk factors (smoking, diabetes, blood pressure, obesity and dyslipidemia) attenuated RRs. RRs for CAC were 1.05 (0.99-1.12; p=0.63) for CRP, 1.12 (1.06-1.20; p<0.01) for IL-6 and 1.09 (1.02-1.16; p=0.01) for fibrinogen in multivariable adjusted models. Results were similar for men and women and across ethnic groups.
Inflammatory markers were weakly associated with CAC presence and burden in MESA. Our data support the hypothesis that inflammatory biomarkers and CAC reflect distinct pathophysiology.
Atherosclerosis; Calcium; Inflammation; Population
It is common in longitudinal studies for scheduled visits to be accompanied by as-needed visits due to medical events occurring between scheduled visits. If the timing of these as-needed visits is related to factors that are associated with the outcome but are not among the regression model covariates, naively including these as-needed visits in the model yields biased estimates. In this paper, the authors illustrate and discuss the key issues pertaining to inverse intensity rate ratio (IIRR)-weighted generalized estimating equations (GEE) methods in the context of a study of Kenyan mothers infected with human immunodeficiency virus type 1 (1999–2005). The authors estimated prevalences and prevalence ratios for morbid conditions affecting the women during a 1-year postpartum follow-up period. Of the 484 women under study, 62% had at least 1 as-needed visit. Use of a standard GEE model including both scheduled and unscheduled visits predicted a pneumonia prevalence of 2.9% (95% confidence interval: 2.3%, 3.5%), while use of the IIRR-weighted GEE predicted a prevalence of 1.5% (95% confidence interval: 1.2%, 1.8%). The estimate obtained using the IIRR-weighted GEE approach was compatible with estimates derived using scheduled visits only. These results highlight the importance of properly accounting for informative follow-up in these studies.
data analysis; data interpretation, statistical; epidemiologic methods; follow-up studies; generalized estimating equation; generalized linear model; longitudinal studies; models, statistical
We present a new joint longitudinal and survival model aimed at estimating the association between the risk of an event and the change in and history of a biomarker that is repeatedly measured over time. We use cubic B-splines models for the longitudinal component that lend themselves to straight-forward formulations of the slope and integral of the trajectory of the biomarker. The model is applied to data collected in a long term follow-up study of HIV infected infants in Uganda. Estimation is carried out using MCMC methods. We also explore using the deviance information criteria, the conditional predictive ordinate and ROC curves for model selection and evaluation.
HIV/AIDS; disease progression; mother-to-child transmission; joint longitudinal and survival models; biomarker change
There are limited data regarding the relative merits of biomarkers as predictors of mortality or time to initiation of antiretroviral therapy (ART).
We evaluated the usefulness of the CD4 cell count, CD4 cell percentage (CD4%), human immunodeficiency virus type 1 (HIV-1) load, total lymphocyte count (TLC), body mass index (BMI), and hemoglobin measured at 32 weeks’ gestation as predictors of mortality in a cohort of HIV-1–infected women in Nairobi, Kenya. Sensitivity, specificity, positive predictive value (PPV), and area under the receiver operating characteristic (ROC) curve (AUC) were determined for each biomarker separately, as well as for the CD4 cell count and the HIV-1 load combined.
Among 489 women with 10,150 person-months of follow-up, mortality rates at 1 and 2 years postpartum were 2.1% (95% confidence interval [CI], 0.7%–3.4%) and 5.5% (95% CI, 3.0%–8.0%), respectively. CD4 cell count and CD4% had the highest AUC value (>0.9). BMI, TLC, and hemoglobin were each associated with but poorly predictive of mortality (PPV, <7%). The HIV-1 load did not predict mortality beyond the CD4 cell count.
The CD4 cell count and CD4% measured during pregnancy were both useful predictors of mortality among pregnant women. TLC, BMI, and hemoglobin had a limited predictive value, and the HIV-1 load did not predict mortality any better than did the CD4 cell count alone.
To determine the utility of Total Lymphocyte Count (TLC) in predicting the 12 month mortality in HIV infected Ugandan children; to correlate TLC and CD4 cell %.
This is a retrospective data analysis of clinical and laboratory data collected prospectively on 128 HIV infected children in the HIVNET 012 trial.
TLC and CD4 cell % measurements were obtained at birth, 14 weeks and 12, 24, 36, 48, and 60 months of age and assessed with respect to risk of death within 12 months.
Median TLC/ul (CD4 cell %) were 4150 (41%) at birth, 4900 (24%) at 12 months, 4300 (19%) at 24 months, 4150 (19 %) at 36 months, 4100 (18%) at 48 months and 3800 (20%) at 60 months. The highest risk of mortality within 12 months was 34–37% at birth and declined to 13–15% at 24 months regardless of TLC measurement. The correlation between CD4 cell % and TLC was extremely low overall (r = 0.01).
The TLC did not predict a risk of progression to death within 12 months and therefore TLC alone may not be a useful surrogate marker for determining those children in greatest need for antiretroviral therapy in HIV infected Ugandan children.
Total Lymphocyte Count; HIV; Africa; children
Morbidity and mortality patterns among pregnant women and their infants (before antiretroviral therapy was widely available) determines HIV-1 diagnostic, monitoring, and care interventions.
Data from mothers and their infants enrolled in a trial of antibiotics to reduce mother-to-child-transmission of HIV-1 at 4 sub-Saharan African sites were analyzed. Women were enrolled during pregnancy and follow-up continued until the infants reached 12 months of age. We describe maternal and infant morbidity and mortality in a cohort of HIV-1-infected and HIV-1-uninfected mothers. Maternal and infant factors associated with mortality risk in the infants were assessed using Cox proportional hazard modeling.
Among 2292 HIV-1-infected mothers, 166 (7.2%) had a serious adverse event (SAE) and 42 (1.8%) died, whereas no deaths occurred among the 331 HIV-1 uninfected mothers. Four hundred twenty-four (17.8%) of 2383 infants had an SAE and 349 (16.4%) died before the end of follow-up. Infants with early HIV-1 infection (birth to 4 – 6 weeks) had the highest mortality. Among infants born to HIV-1-infected women, maternal morbidity and mortality (P = 0.0001), baseline CD4 count (P = 0.0002), and baseline plasma HIV-1 RNA concentration (P < 0.0001) were significant predictors of infant mortality in multivariate analyses.
The high mortality among infants with early HIV-1 infection supports access to HIV-1 diagnostics and appropriate early treatment for all infants of HIV-1-infected mothers. The significant association between stage of maternal HIV-1 infection and infant mortality supports routine CD4 counts at the time of prenatal HIV-1 testing.
HIV-1 infection; infant mortality; maternal morbidity and mortality; sub-Saharan Africa; pregnant women