Juwon Song

Reductions in HIV Risk Among Runaway Youth

Runaway youth are 6–12 times more likely to become infected with HIV than other youth. Using a quasi-experimental design, the efficacy of an HIV prevention program was evaluated over 2 years among 2 groups of runaways: (1) those at 2 shelters who received Street Smart, an intensive HIV intervention program, and (2) youth at 2 control shelters. Street Smart provided youth with access to health care and condoms and delivered a 10-session skill-focused prevention program based on social learning theory to youth. Prior to analysis of the interventions outcomes, propensity scores were used to identify comparable subgroups of youth in the intervention (n = 101) and control conditions (n = 86). Compared to females in the control condition, females in the intervention condition significantly reduced their unprotected sexual acts at 2 years and alcohol use, marijuana use, and the number of drugs used over 12 months. Male adolescents in the intervention condition showed significant reductions in marijuana use over 6 months compared to control youth. Adolescent HIV prevention programs must proactively identify mechanisms for maintaining behavior change over the long-term, and innovative research designs are needed to allow examination of agency-level interventions.

The 18-month impact of an emergency room intervention for adolescent female suicide attempters.

Following a suicide attempt by female adolescents, the impact of a specialized emergency room (ER) care intervention was evaluated over the subsequent 18 months. Using a quasi-experimental design, this study assigned 140 female adolescent suicide attempters (SA), ages 12-18 years, and their mothers (88% Hispanic) to receive during their ER visit either: (a) specialized ER care aimed at enhancing adherence to outpatient therapy by providing a soap opera video regarding suicidality, a family therapy session, and staff training; or (b) standard ER care. The adjustment of the SA and their mothers was evaluated over 18 months (follow-up, 92%) using linear mixed model regression analyses. SAs adjustment improved over time on most mental health indices. Rates of suicide reattempts (12.4%) and suicidal reideation (29.8%) were lower than anticipated and similar across ER conditions. The specialized ER care condition was associated with significantly lower depression scores by the SA and lower maternal ratings on family cohesion. Significant interactions of intervention condition with the SAs initial level of psychiatric symptomatology indicated that the interventions impact was greatest on maternal emotional distress and family cohesion among SA who were highly symptomatic. SAs attendance at therapy sessions following the ER visit was significantly associated with only one outcome--family adaptability. Specialized ER interventions may have substantial and sustained impact over time, particularly for the parents of youth with high psychiatric symptomatology.

Effects of a behavioral intervention to reduce risk of transmission among people living with HIV: The Healthy Living Project randomized controlled study

Context:The US Centers for Disease Control and Prevention (CDC) strongly recommend comprehensive risk counceling and services for people living with HIV (PLH); yet, there are no evidence-based counseling protocols. Objective:To examine the effect of a 15-session, individually delivered, cognitive behavioral intervention on a diverse sample of PLH at risk of transmitting to others. Design:This was a multisite, 2-group, randomized, controlled trial. Participants:Nine hundred thirty-six HIV-infected participants considered to be at risk of transmitting HIV of 3818 persons screened were randomized into the trial. Participants were recruited in Los Angeles, Milwaukee, New York, and San Francisco. Intervention:Fifteen 90-minute individually delivered intervention sessions were divided into 3 modules: stress, coping, and adjustment; safer behaviors; and health behaviors. The control group received no intervention until the trial was completed. Both groups completed follow-up assessments at 5, 10, 15, 20, and 25 months after randomization. Main Outcome Measure:Transmission risk, as measured by the number of unprotected sexual risk acts with persons of HIV-negative or unknown status, was the main outcome measure. Results:Overall, a significance difference in mean transmission risk acts was shown between the intervention and control arms over 5 to 25 months (χ2 = 16.0, degrees of freedom = 5; P = 0.007). The greatest reduction occurred at the 20-month follow-up, with a 36% reduction in the intervention group compared with the control group. Conclusion:Cognitive behavioral intervention programs can effectively reduce the potential of HIV transmission to others among PLH who report significant transmission risk behavior.

Do malpractice concerns, payment mechanisms, and attitudes influence test-ordering decisions?

Greater understanding is needed of nonclinical factors that determine neurologists’ decisions to order tests. The authors surveyed 595 US neurologists and utilized demographic information, attitude scales, and clinical scenarios to evaluate the influence of nonclinical factors on test-ordering decisions. Greater test reliance, higher malpractice concerns, and receiving reimbursement for testing were all associated with a higher likelihood of test ordering. These findings have implications for training needs and suggest malpractice worries may inflate health care costs.

The influence of partner type and risk status on the sexual behavior of young men who have sex with men living with HIV/AIDS.

Objectives: The influence of partner type and risk status on the unprotected sexual behavior of young men living with HIV (YMLH) who have sex with men is examined. Methods: Sexual behavior and sexual partner characteristics of 217 YMLH recruited from adolescent care clinics in 4 AIDS epicenters (Los Angeles, San Francisco, New York, and Miami) were assessed. YMLH were categorized by sexual behavior pattern, and sexual partners were classified by type and risk status. Generalized linear modeling employing overdispersed Poisson distribution was used to analyze the effect of partner type and partner risk status on unprotected sex acts. Results: Most YMLH (62%) reported multiple partners, 26% reported 1 sexual partner, and 12% reported abstinence in the past 3 months. Approximately 34% of polygamous and 28% of monogamous youth engaged in unprotected sex. Monogamous youth were most likely to have unprotected sex with HIV-positive partners. Polygamous youth were most likely to have unprotected sex with HIV-positive partners, irrespective of whether the partner was regular or casual. For polygamous YMLH, unprotected sex did not differ among single-time/new partners with different risk levels. Conclusions: Partner characteristics influence the condom use behavior of YMLH. YMLH make decisions regarding condom use based on perception of their partners risk. Preventive interventions must include skills for acquiring accurate information about partner risk status and education regarding the health risks of unprotected sex with HIV seroconcordant partners.

Intergenerational benefits of family-based HIV interventions.

The longitudinal impact of a family-based intervention on grandchildren of parents with HIV (PWH) is evaluated. Because PWH and their daughters demonstrated gains over 6 years when randomized to a coping skills intervention compared with a control condition, the adjustment of the PWHs grandchildren was also compared across conditions. Grandchildren in the intervention condition reported significantly fewer internalizing and externalizing behavioral symptoms compared with grandchildren in the control condition. There is weak evidence that grandchildren in the intervention condition had higher scores on measures of cognitive development and more positive home environments. These results suggest that there are possibly long-term, intergenerational benefits of an intervention for families coping with HIV.

Choosing an appropriate number of factors in factor analysis with incomplete data

When we conduct factor analysis, the number of factors is often unknown in advance. Among many decision rules for an appropriate number of factors, it is easy to find approaches that make use of the estimated covariance matrix. When data include missing values, the estimated covariance matrix using either complete cases or available cases may not accurately represent the true covariance matrix, and decision based on the estimated covariance matrix may be misleading. We discuss how to apply model selection techniques using AIC or BIC to choose an appropriate number of factors when data include missing values. In the simulation study, it is shown that the suggested methods select the correct number of factors for simulated data with known number of factors.

Handling Baseline Differences and Missing Items in a Longitudinal Study of HIV Risk Among Runaway Youths

Many longitudinal studies in field settings present challenges due to selection bias and incomplete data. A motivating example is provided by an intervention study aimed at preventing HIV transmission among runaway youths housed at shelters in New York City. Two shelters with 167 youths received the intervention, and two shelters with 144 youths received the control treatment. The number of unprotected sexual acts in the prior three months for each youth was assessed at a baseline interview and (to the extent possible) at five follow-up time points. Among observed items, there is strong evidence of a lack of balance on baseline characteristics between the intervention and control groups; meanwhile, beyond occasional missing items among participants interviewed at baseline, there were three items about baseline characteristics added after the study began, resulting in a few items being missing on a large percentage of the study sample. Here, we outline two strategies for handling the complexities of this data set, both of which make use of propensity scores to address the imbalances across treatment groups. One approach relies on available cases and ad hoc choices to simplify the steps leading up to a linear mixed model analysis in SAS PROC MIXED; the other approach uses multiple imputation strategies to reflect uncertainty due to missing values in an analogous linear mixed model analysis. Ultimately we did not find substantial qualitative differences in this setting between the available-case and imputed-data approaches. But in both cases, we find that the considerable imbalance on covariates between treatment arms constrains the ability to draw inferences about the intervention effect, suggesting the importance of evaluating propensity-score distributions in quasi-experimental intervention research.

Score tests for zero-inflation and overdispersion in two-level count data

In a Poisson regression model in which observations are either clustered or represented by repeated measurements of counts, the number of observed zero counts is sometimes greater than the expected frequency by the Poisson distribution and overdispersion may remain even after modeling excess zeros. The zero-inflated negative binomial (ZINB) mixed regression model is suggested to analyze such data. Previous studies have proposed score statistics for testing zero-inflation and overdispersion separately in correlated count data. Here, we also deal with simultaneous score tests for zero-inflation and overdispersion in two-level count data by using the ZINB mixed regression model. Score tests are suggested for (1) zero-inflation in the presence of overdispersion, (2) overdispersion in the presence of zero-inflation, and (3) zero-inflation and overdispersion simultaneously. The level and power of score test statistics are evaluated by a simulation study. The simulation results indicate that score test statistics may occasionally underestimate or overestimate the nominal significance level due to variation in random effects. This study proposes a parametric bootstrap method to overcome this problem. The simulation results of the bootstrap test indicate that score tests hold the nominal level and provide good power.

Pattern-Mixture Model of the Cox Proportional Hazards Model with Missing Binary Covariates

When fitting a Cox proportional hazards model with missing covariates, it is inefficient to exclude observations with missing values in the analysis. Furthermore, if the missing-data mechanism is not Missing Completely At Random(MCAR), it may lead to biased parameter estimation. Many approaches have been suggested to handle the Cox proportional hazards model when covariates are sometimes missing, but they are based on the selection model. This paper suggest an approach to handle Cox proportional hazards model with missing covariates by using the pattern-mixture model (Little, 1993). The pattern-mixture model is expressed by the joint distribution of survival time and the missing-data mechanism. In the pattern-mixture model, many models can be considered by setting up various restrictions, and different results under various restrictions indicate the sensitivity of the model due to missing covariates. A simulation study was conducted to show the sensitivity of parameter estimation under different restrictions in a pattern-mixture model. The proposed approach was also applied to mouse leukemia data.