Jane C. Lindsey

Tutorial in biostatistics methods for interval-censored data

In standard time-to-event or survival analysis, occurrence times of the event of interest are observed exactly or are right-censored, meaning that it is only known that the event occurred after the last observation time. There are numerous methods available for estimating the survival curve and for testing and estimation of the effects of covariates in this context. In some situations, however, the times of the events of interest may only be known to have occurred within an interval of time. In clinical trials, for example, patients are often seen at pre-scheduled visits but the event of interest may occur in between visits. These data are interval-censored. Owing to the lack of well-known statistical methodology and available software, a common ad hoc approach is to assume that the event occurred at the end (or beginning or midpoint) of each interval, and then apply methods for standard time-to-event data. However, this approach can lead to invalid inferences, and in particular will tend to underestimate the standard errors of the estimated parameters. The purpose of this tutorial is to illustrate and compare available methods which correctly treat the data as being interval-censored. It is not meant to be a full review of all existing methods, but only those which are available in standard statistical software, or which can be easily programmed. All approaches will be illustrated on two data sets and compared with methods which ignore the interval-censored nature of the data. We hope this tutorial will allow those familiar with the application of standard survival analysis techniques the option of applying appropriate methods when presented with interval-censored data.

Antiretroviral treatment for children with peripartum nevirapine exposure

BACKGROUND Single-dose nevirapine is the cornerstone of the regimen for prevention of mother-to-child transmission of human immunodeficiency virus (HIV) in resource-limited settings, but nevirapine frequently selects for resistant virus in mothers and children who become infected despite prophylaxis. The optimal antiretroviral treatment strategy for children who have had prior exposure to single-dose nevirapine is unknown. METHODS We conducted a randomized trial of initial therapy with zidovudine and lamivudine plus either nevirapine or ritonavir-boosted lopinavir in HIV-infected children 6 to 36 months of age, in six African countries, who qualified for treatment according to World Health Organization (WHO) criteria. Results are reported for the cohort that included children exposed to single-dose nevirapine prophylaxis. The primary end point was virologic failure or discontinuation of treatment by study week 24. Enrollment in this cohort was terminated early on the recommendation of the data and safety monitoring board. RESULTS A total of 164 children were enrolled. The median percentage of CD4+ lymphocytes was 19%; a total of 56% of the children had WHO stage 3 or 4 disease. More children in the nevirapine group than in the ritonavir-boosted lopinavir group reached a primary end point (39.6% vs. 21.7%; weighted difference, 18.6 percentage-points; 95% confidence interval, 3.7 to 33.6; nominal P=0.02). Baseline resistance to nevirapine was detected in 18 of 148 children (12%) and was predictive of treatment failure. No significant between-group differences were seen in the rate of adverse events. CONCLUSIONS Among children with prior exposure to single-dose nevirapine for perinatal prevention of HIV transmission, antiretroviral treatment consisting of zidovudine and lamivudine plus ritonavir-boosted lopinavir resulted in better outcomes than did treatment with zidovudine and lamivudine plus nevirapine. Since nevirapine is used for both treatment and perinatal prevention of HIV infection in resource-limited settings, alternative strategies for the prevention of HIV transmission from mother to child, as well as for the treatment of HIV infection, are urgently required. (Funded by the National Institutes of Health; ClinicalTrials.gov number, NCT00307151.).

Unfair Treatment, Racial/Ethnic Discrimination, Ethnic Identification, and Smoking Among Asian Americans in the National Latino and Asian American Study

OBJECTIVES We examined the relations of self-report of general unfair treatment and self-report of race/ethnicity-specific discrimination with current smoking among Asian Americans. We investigated whether ethnic identification moderated either association. METHODS Weighted logistic regressions were performed among 1977 Asian Americans recruited to the National Latino and Asian American Study (2002-2003). RESULTS In weighted multivariate logistic regression models including both general unfair treatment and racial/ethnic discrimination, odds of current smoking were higher among Asian Americans who reported high levels of unfair treatment (odds ratio [OR]=2.80; 95% confidence interval [CI]=1.13, 6.95) and high levels of racial/ethnic discrimination (OR=2.40; 95% CI=0.94, 6.12) compared with those who reported no unfair treatment and discrimination, respectively. High levels of ethnic identification moderated racial/ethnic discrimination (F(3) =3.25; P =.03). High levels of ethnic identification were associated with lower probability of current smoking among participants reporting high levels of racial/ethnic discrimination. CONCLUSIONS Our findings suggest that experiences of unfair treatment and racial/ethnic discrimination are risk factors for smoking among Asian Americans. Efforts to promote ethnic identification may be effective in mitigating the influence of racial/ethnic discrimination on smoking in this population.

Nevirapine versus Ritonavir-Boosted Lopinavir for HIV-Infected Children

BACKGROUND Nevirapine-based antiretroviral therapy is the predominant (and often the only) regimen available for children in resource-limited settings. Nevirapine resistance after exposure to the drug for prevention of maternal-to-child human immunodeficiency virus (HIV) transmission is common, a problem that has led to the recommendation of ritonavir-boosted lopinavir in such settings. Regardless of whether there has been prior exposure to nevirapine, the performance of nevirapine versus ritonavir-boosted lopinavir in young children has not been rigorously established. METHODS In a randomized trial conducted in six African countries and India, we compared the initiation of HIV treatment with zidovudine, lamivudine, and either nevirapine or ritonavir-boosted lopinavir in HIV-infected children 2 to 36 months of age who had no prior exposure to nevirapine. The primary end point was virologic failure or discontinuation of treatment by study week 24. RESULTS A total of 288 children were enrolled; the median percentage of CD4+ T cells was 15%, and the median plasma HIV type 1 (HIV-1) RNA level was 5.7 log(10) copies per milliliter. The percentage of children who reached the primary end point was significantly higher in the nevirapine group than in the ritonavir-boosted lopinavir group (40.8% vs. 19.3%; P<0.001). Among the nevirapine-treated children with virologic failure for whom data on resistance were available, more than half (19 of 32) had resistance at the time of virologic failure. In addition, the time to a protocol-defined toxicity end point was shorter in the nevirapine group (P=0.04), as was the time to death (P=0.06). CONCLUSIONS Outcomes were superior with ritonavir-boosted lopinavir among young children with no prior exposure to nevirapine. Factors that may have contributed to the suboptimal results with nevirapine include elevated viral load at baseline, selection for nevirapine resistance, background regimen of nucleoside reverse-transcriptase inhibitors, and the standard ramp-up dosing strategy. The results of this trial present policymakers with difficult choices. (Funded by the National Institute of Allergy and Infectious Diseases and others; P1060 ClinicalTrials.gov number, NCT00307151.).

Alcohol disorders among Asian Americans: associations with unfair treatment, racial/ethnic discrimination, and ethnic identification (the national Latino and Asian Americans study, 2002–2003)

Study objective: To examine history of alcohol abuse/dependence disorder in relation to unfair treatment, racial/ethnic discrimination, and ethnic identification among Asian Americans. Design: Weighted multivariate analyses of cross-sectional national survey data predicting lifetime history of alcohol abuse/dependence disorders. Setting: USA, Asian Americans. Participants: 2007 Asian American adults recruited to the National Latino and Asian American Study (NLAAS; 2002–2003). Results: Controlling for sociodemographic characteristics, Asian Americans who reported experiencing unfair treatment had higher odds of history of alcohol abuse/dependence disorder (OR 5.26, 95% CI 1.90 to 14.56). Participants who reported high levels of ethnic identification had lower odds of history of alcohol abuse/dependence disorders (OR 0.46, 95% CI 0.23 to 0.90). Ethnic identification moderated the influence of racial/ethnic discrimination (p  =  0.097). Among participants with low levels of ethnic identification, racial/ethnic discrimination was associated with greater odds of having a history of alcohol disorder compared with those with high levels of ethnic identification. Conclusions: Social hazards such as unfair treatment and racial/ethnic discrimination should be considered in the development of programmes addressing alcohol disorders among Asian Americans. Interventions that promote ethnic identification in this population may be particularly relevant in mitigating the negative influence of racial/ethnic discrimination on alcohol disorders.

Adherence to Antiretrovirals Among US Women During and After Pregnancy

Background:Antiretrovirals (ARVs) are recommended for maternal health and to reduce HIV-1 mother-to-child transmission, but suboptimal adherence can counteract its benefits. Objectives:To describe antepartum and postpartum adherence to ARV regimens and factors associated with adherence. Methods:We assessed adherence rates among subjects enrolled in Pediatric AIDS Clinical Trials Group Protocol 1025 from August 2002 to July 2005 on tablet formulations with at least one self-report adherence assessment. Perfectly adherent subjects reported no missed doses 4 days before their study visit. Generalized estimating equations were used to compare antepartum with postpartum adherence rates and to identify factors associated with perfect adherence. Results:Of 519 eligible subjects, 334/445 (75%) reported perfect adherence during pregnancy. This rate significantly decreased 6, 24, and 48 weeks postpartum [185/284 (65%), 76/118 (64%), and 42/64 (66%), respectively (P < 0.01)]. Pregnant subjects with perfect adherence had lower viral loads. The odds of perfect adherence were significantly higher for women who initiated ARVs during pregnancy (P < 0.01), did not have AIDS (P = 0.02), never missed prenatal vitamins (P < 0.01), never used marijuana (P = 0.05), or felt happy all or most of the time (P < 0.01). Conclusions:Perfect adherence to ARVs was better antepartum, but overall rates were low. Interventions to improve adherence during pregnancy are needed.

Virologic and Immunologic Outcomes after 24 Weeks in HIV Type 1-Infected Adolescents Receiving Highly Active Antiretroviral Therapy

BACKGROUND Adolescents represent the fastest growing demographic group of new human immunodeficiency virus (HIV) infections in the United States. At present, there is little information available about their response to therapy. METHODS We studied 120 adolescents infected via high-risk behaviors who began receiving highly active antiretroviral therapy (HAART), to determine their virologic and immunologic response to therapy. RESULTS Subjects were enrolled at 28 sites of the Pediatric Acquired Immunodeficiency Syndrome Clinical Trials Group. After 16-24 weeks of HAART, 59% of subjects had reproducible undetectable virus loads, according to repeat measurements (virologic success). As enumerated by flow-cytometric analysis, increases in levels of CD4 helper cells (both naive and memory) and decreases in levels of CD8 suppressor cells were observed. Partial restoration of some immunologic parameters for patients who did not achieve virologic success was also observed, but to a more limited extent than for adolescents with virologic success. Adherence to HAART was the only predictor of achieving undetectable virus loads. CONCLUSIONS Adolescents have the capacity to improve their immunologic status with HAART. Lower than expected success in virologic control is related to lack of adherence, and efforts to improve treatment outcome must stress measures to assure adherence to medication.

Correlates of opportunistic infections in children infected with the human immunodeficiency virus managed before highly active antiretroviral therapy.

BACKGROUND Opportunistic infections (OIs) are an important cause of morbidity and mortality in children infected with HIV. However, few data are available regarding the overall prevalence, incidence and immunologic correlates associated with these diseases in the pediatric HIV population. The Pediatric AIDS Clinical Trials Group (PACTG) has conducted multicenter studies in HIV-infected children since 1988 and through these studies has collected prospective data on the immunologic and virologic status of study participants and recorded complications, including infectious diseases, related to HIV infection and its treatments. Therefore data were analyzed from across 13 PACTG studies, performed before treatment with highly active antiretroviral therapy was given, to determine the rates of various infectious complications and the immunologic correlates, specifically CD4 cell counts, associated with these diseases. RESULTS OIs were tabulated from 3331 HIV-infected children who participated in 13 clinic trials undertaken before highly effective antiretroviral therapy was available. Five OIs occurred at event rates of >1.0 per 100 patient years (person years): serious bacterial infections, 15.1; herpes zoster, 2.9; disseminated Mycobacterium avium complex (DMAC), 1.8; Pneumocystis carinii pneumonia, 1.3; and tracheobronchial and esophageal candidiasis, 1.2. Six other OIs evaluated, cytomegalovirus (CMV) disease, cryptosporidiosis, tuberculosis, systemic fungal infections, toxoplasmosis and progressive multifocal leukoencephalopathy, occurred at event rates of <1.0 per 100 person years. Pneumonia (11.1 per 100 person years) and bacteremia (3.3 per 100 person years) were the most common bacterial infections. An AIDS-defining OI before entry was a risk factor for the development of a new OI during a trial. Bacterial infections, herpes zoster and tuberculosis occurred frequently at all stages of HIV infection; whereas DMAC, P. carinii pneumonia, CMV and other OIs occurred primarily in children with severe immunosuppression. CONCLUSIONS The frequency of OIs in HIV-infected children in the pre-highly active antiretroviral therapy era varies with age, pathogen, prior OI and immunologic status. Analysis of CD4 counts at the time of DMAC, CMV and PCP provide validation for current prophylaxis guidelines in children > or =2 years old. This information on infectious complications of pediatric HIV will be especially valuable for contemporary management of HIV infection that is poorly responsive to highly active antiretroviral therapy.

Morphologic and metabolic abnormalities in vertically HIV-infected children and youth.

Objective:To compare the distribution of lipid and glucose abnormalities and altered fat distribution among vertically HIV-infected patients and controls. Design:Cross-sectional multicenter study on HIV-infected (HIV-positive) patients, 7–24 years of age, stratified by Tanner stage and protease inhibitor use (protease inhibitor, n = 161 and non- protease inhibitor, n = 79) and seronegative controls (HIV-negative, n = 146). Methods:Measurements included fasting lipids, glucose, insulin, 2-h oral glucose tolerance test, dual-energy X-ray absorptiometry, anthropometry, and antiretroviral therapy and medical histories. Multiple linear regression models were used to compare distributions between HIV-positive and HIV-negative groups. Results:Both HIV-positive groups had long exposures to antiretroviral therapy. Protease inhibitor and nonprotease inhibitor groups had similar current CD4 cell count and HIV-1 RNA, but the protease inhibitor group had lower nadir CD4 cell count, higher peak HIV-1 RNA, and more advanced Centers for Disease Control disease stage. In adjusted analyses, both HIV-positive groups had significantly lower mean Z scores for height, weight, BMI, and total and limb fat than the HIV-negative group. Mean triglycerides were significantly higher and high-density lipoprotein cholesterol lower in both HIV-positive groups relative to the HIV-negative group. The protease inhibitor group also had significantly higher mean total, low-density lipoprotein, and non-high density lipoprotein cholesterol. Mean fasting insulin was higher in both HIV-positive groups, and 2-h glucose and insulin were higher in the protease inhibitor group. Ritonavir was associated with increasing dyslipidemia and altered glucose metabolism. Conclusion:In a large group of vertically HIV-infected children and youth with extensive antiretroviral therapy exposure, height, weight, and total and limb fat were lower than in controls. There was a high prevalence of lipid abnormalities among those on protease inhibitors and evidence of developing insulin resistance, factors that may accelerate lifetime risk for cardiovascular disease.

Neurodevelopmental Functioning in HIV-Infected Infants and Young Children Before and After the Introduction of Protease Inhibitor–Based Highly Active Antiretroviral Therapy

OBJECTIVES. The purpose of this work was to examine the effects of HIV infection and the impact of highly active antiretroviral treatment with protease inhibitors on neurodevelopmental functioning during the first 3 years of life. PATIENTS AND METHODS. Pediatric AIDS Clinical Trials Group 219/219C is a longitudinal cohort study that has enrolled HIV-infected (HIV+) and HIV-exposed but uninfected (HIV−) infants and children since 1993. Longitudinal profiles of neurodevelopmental functioning as measured by the Bayley Scales of Infant Development were compared by HIV-infection status before and after the availability of highly active antiretroviral therapy with a protease inhibitor and within infants with Bayley tests available before and after initiating protease inhibitor therapy. RESULTS. In the pre–protease inhibitor era, mean mental and motor scores in HIV+ (n = 54) infants <1 year of age were significantly lower than those among HIV− infants (n = 221) and remained lower up to 2 years of age. After protease inhibitors became available, mean mental and motor functioning of HIV+ infants (n = 91) <1 year of age were still significantly lower than those of HIV− infants (n = 838). However, against a background of declining scores among the HIV− infants, there was evidence of limited improvement in the HIV+ infants relative to their uninfected peers. Among infants who had Bayley II evaluations before and after starting a protease inhibitor, there was a trend to improved mental and motor scores after initiation of protease inhibitor therapy. CONCLUSIONS. The suppression of systemic viral replication and subsequent substantial improvements in survival and immunologic status brought about by highly active antiretroviral therapy have been followed by limited improvements in neurodevelopmental functioning in young children. Additional longitudinal research is needed to better understand the role of antiretroviral therapy as well as the impact of genetic and environmental factors on neurodevelopmental functioning in children affected by HIV.