Application of random effect ordinal regression model for outcome evaluation of two randomized controlled trials.

Cluster randomization is often used in intervention trials, yet when individuals nested within clusters are considered as the units of analysis for outcome evaluation, it cannot be assumed that the observations are statistically independent. Observations that are not statistically independent also result when repeated measures are taken over time for the same individual. Ignoring clustered observations when performing data analysis can lead to the erroneous conclusion that the intervention under study had a statistically significant effect. Moreover, individual responses are often collected on ordinal scales; thus models for continuous or categorical data are usually not appropriate. We applied a random effect ordinal regression model to data sets from two randomized controlled intervention trials that measured graded scale non-independent responses. The first trial compared two school programmes for AIDS prevention in terms of impact (i.e., changes in the frequency of condom use). The second trial used the MOS-HIV questionnaire to measure the quality of life of new AIDS cases four times over a one-year follow-up period (only results of the role-functioning scale are reported). Regarding the first data set, the effect of the intervention was not significant, and the post-intervention frequency of condom use was mainly attributable to the pre-intervention frequency (p<0.01), with no differences among schools. Regarding the second data set, a borderline significant increase in the role-functioning scale scores was observed over the follow-up period; the results differed only slightly by intervention group; a significant (p<0.01) intra-individual correlation of 0.4 was found.     

Community-level HIV intervention in 5 cities: final outcome data from the CDC AIDS Community Demonstration Projects

OBJECTIVES: This study evaluated a theory-based community-level intervention to promote progress toward consistent condom and bleach use among selected populations at increased risk for HIV infection in 5 US cities. METHODS: Role-model stories were distributed, along with condoms and bleach, by community members who encouraged behavior change among injection drug users, their female sex partners, sex workers, non-gay-identified men who have sex with men, high-risk youth, and residents in areas with high sexually transmitted disease rates. Over a 3-year period, cross-sectional interviews (n = 15,205) were conducted in 10 intervention and comparison community pairs. Outcomes were measured on a stage-of-change scale. Observed condom carrying and intervention exposure were also measured. RESULTS: At the community level, movement toward consistent condom use with main (P < .05) and nonmain (P < .05) partners, as well as increased condom carrying (P < .0001), was greater in intervention than in comparison communities. At the individual level, respondents recently exposed to the intervention were more likely to carry condoms and to have higher stage-of-change scores for condom and bleach use. CONCLUSIONS: The intervention led to significant communitywide progress toward consistent HIV risk reduction.     

Exposure to a Community-Level HIV Prevention Intervention: Who Gets the Message

As part of the evaluation of a community-level HIV prevention program for women, this study examined predictors of exposure to print media and community outreach and assessed the relationship between exposure to the intervention and condom use behavior. Data from interviews with 479 women randomly selected from the intervention community in 1995 and 1996 were examined. Analysis of demographic and risk characteristics were conducted to identify predictors of exposure to the project's HIV prevention messages. Additionally, logistic regression analyses were conducted to examine the effects of intervention exposure on condom use, controlling for factors related to exposure. The results revealed that the print media campaign reached the largest number of women. However, women at highest risk did not have high rates of exposure to print media, but had greater exposure to outreach. Exposure to print media had an effect on increased communication with a main partner about condom use, but was not significantly related to condom use last time had sex. There were no significant main effects for exposure to outreach on condom use behavior. An important finding of this analysis was that each intervention strategy was successful in reaching a different portion of the target population and that exposure had differential effects on the condom use behavior of particular segments of the target population.     

Factors associated with perceptions of,and decisional balance for,condom use with main partner among women at risk for HIV infection

We examined factors associated with women's perceived advantages (pros), perceived disadvantages (cons), and decisional balance (standardized pros score minus standardized cons score) for condom use with main partner. Data from 1,938 young sexually active women who lived in five U.S. cities where the risk for human immunodeficiency virus is high were analyzed by using logistic, ordinal, and multiple linear regression analysis. For the pros scale of condom use, 27% of the women had low scores, and 33% had moderate scores. For the cons scale, 27% had moderate scores, and 5% had high scores. Of the total, 47% had a negative score on the decisional balance measure. Older age, living with a spouse or partner, or binge drinking was associated with lower pros scores and with a negative score on the decisional balance measure. Income from public assistance was associated with higher pros scores. Income from a spouse or partner or a history of sexually transmitted disease was associated with lower pros scores. Multiple sex partners or being at risk for HIV infection (based on perceptions of the main partner's behaviors) was associated with higher cons scores. Income from a job was associated with a positive score on the decisional balance measure. Our analysis identified the characteristics of women who have low pros scores, high cons scores, and negative decisional balance scores. The regression results can inform our work in HIV prevention on whether to focus on the pros, the cons, or both to obtain positive decisional balance scores and increase condom use in situations that warrant protective behaviors.     

Effect of an STD/HIV Behavioral Intervention on Women''s Use of the Female Condom

OBJECTIVES: This study assessed the effectiveness of a sexually transmitted disease (STD)/HIV behavior change intervention in increasing women's use of the female condom. METHODS: A total of 604 women at high risk for STDs and HIV in New York City, Baltimore, Md, and Seattle, Wash, enrolled in a randomized controlled trial of a small-group, skills-training intervention that included information and skills training in the use of the female condom. RESULTS: In a logistic regression, the strongest predictors of use were exposure to the intervention (odds ratio [OR] = 5.5; 95% confidence interval [CI] = 2.8, 10.7), intention to use the female condom in the future (OR = 4.5; 95% CI = 2.4, 8.5), having asked a partner to use a condom in the past 30 days (OR = 2.3; 95% CI = 1.3, 3.9), and confidence in asking a partner to use a condom (OR = 1.9; 95% CI = 1.1, 3.5). CONCLUSIONS: Clinicians counseling women in the use of the female condom need to provide information, demonstrate its correct use with their clients, and provide an opportunity for their clients to practice skills themselves.     

Correlates of condom use with “main” and “concurrent” sex partners among urban youth in post-conflict Liberia

This paper presents baseline findings on correlates of condom use with “main” and “concurrent” sex partners among Liberian youth (ages 15–17) in Liberia. With Liberia having recently emerged from fourteen years of civil war, this paper is one of only a few published reports of the sexual risk behaviors of urban youth residing in Monrovia, Liberia’s capital. The analysis was limited to highly, sexually active Liberian adolescents who reported having sex with both “main” and “concurrent” sex partners in the last three months (n = 214). Data were used from a baseline survey of a community randomized controlled trial of an HIV prevention intervention. Multinomial logistic regression analysis was conducted to assess correlates of condom use among respondents who used condoms with both main and concurrent partners, when compared to respondents who used condoms with main partners only, and when compared to respondents who used condoms with concurrent partners only. Findings suggest that among Liberian adolescents who have both main and concurrent sex partners, males were more likely to use a condom with both partner types when compared to females, relative to those who did not use condoms with both partner types. Greater parent/child communication about sex and use of a condom at first sexual debut were associated with condom use in main only and concurrent only sex partnerships, relative to those who did not use a condom with either partner types.     

郑州市外来务工人员安全套使用干预模式的析因设计分析

目的了解郑州市外来务工人员安全套使用现状,探讨不同干预模式对外来务工者安全套正确使用的影响,为在该人群中推广使用安全套采取有效的干预措施提供依据。方法采用分层整群抽样的方法,从郑州市中原区抽取889例外来务工人员作为调查对象,采取匿名问卷调查的方式进行调查。并运用析因设计的方法观察集中宣教和观看VCD两种干预模式的效果。结果外来务工人员中,62.9%(352/560)的人发生性行为使用过安全套,最近一个月发生性行为时33.2%的人从未使用安全套,只有17.6%的被调查者每次性行为都用。81.9%的人认为正确使用安全套能预防艾滋病,73.2%的外来务工人员在今后的性生活中愿意坚持使用安全套。不同干预方式对安全套正确使用知识得分的影响显示集中宣教的效果有统计学差异(F=21.870,P<0.001),logistic回归分析影响安全套正确使用的因素依次为艾滋病知识知晓程度、婚姻状况、性别。结论外来务工人员安全套使用率低,安全套正确使用和预防艾滋病知识缺乏,集中宣教是推广安全套正确使用的有效干预方法。     

Condom use: slippage, breakage, and steps for proper use among adolescents in alternative school settings

BACKGROUND : School‐based human immunodeficiency virus (HIV)/sexually transmitted infection (STI), and pregnancy prevention programs often focus on consistent and correct condom use. Research on adolescents' experience using condoms, including condom slippage/breakage, is limited. This exploratory study examines proper condom use and the occurrence of condom slippage/breakage among alternative school youth. METHODS : Data are from an HIV/STI prevention trial for youth in continuation school settings (N = 776). Analyses included separate hierarchical logistic regression analyses to explore the relationship between potential correlates and each outcome variable. RESULTS : Students' use of steps for proper condom use varied—73.8% put on the condom before sexual contact, 71.1% squeezed air from the tip, and 92.0% unrolled the condom fully. Notably, 28.5% reported condom slippage/breakage. Results from the regression analyses showed that 4 sets of variables (demographic, substance use, sexual risk behaviors, and condom psychosocial factors) were associated with putting on a condom before sexual contact; none of the variable sets were associated with the other 2 condom steps measured. For slippage/breakage, the demographic and sexual risk behaviors were significant correlates; steps for proper condom use approached statistical significance (p = .058). CONCLUSIONS : This study extends the limited research on how adolescents use condoms, and highlights important targets for prevention interventions.     

Threats of Cross-Contamination on Effects of a Sexual Risk Reduction Intervention: Fact or Fiction

This study sought to empirically evaluate the extent and impact of cross-contamination on the effects of a STI/HIV intervention trial previously shown to be effective in reducing high-risk sexual behaviors among African–American adolescent females. Participants were recruited through community health agencies in the Southeastern United States and comprised 522 sexually active 14- to 18- year-old African–American females who completed self-administered questionnaires and face-to-face interviews at baseline, 6- and 12-month time points. Participants were randomized to a STI/HIV risk reduction group or a general health promotion group. The STI/HIV intervention group participated in four group sessions addressing constructs such as HIV knowledge, communication, condom use self-efficacy and condom use behaviors. The control group participated in four group sessions focused on general health topics. The study setting afforded multiple opportunities for cross-talk between intervention and control group participants. Consistent condom use, defined as condom use during every vaginal sex act, was the primary outcome measure. Other outcome measures included various sexual behaviors, observed condom application skills and psychosocial variables associated with HIV preventive behaviors. Approximately 73% of participants reported some level of cross-talk. Linear and binary GEE models assessing the impact of the STI/HIV intervention on contaminated vs. uncontaminated control group participants indicated no differential effects of the intervention. Furthermore, equivalence tests demonstrated that contaminated and uncontaminated control groups were equivalent. Findings from this study provide empirical evidence suggesting that behavioral and psychosocial outcomes may be resistant to cross-contamination in randomized controlled trials testing safer sex interventions among African–American adolescent females.     

Quantile estimation following non-parametric phase I clinical trials with ordinal response

A non-parametric multi-dimensional isotonic regression estimator is developed for use in estimating a set of target quantiles from an ordinal toxicity scale. We compare this estimator to the standard parametric maximum likelihood estimator from a proportional odds model for extremely small data sets. A motivating example is from phase I oncology clinical trials, where various non-parametric designs have been proposed that lead to very small data sets, often with ordinal toxicity response data. Our comparison of estimators is performed in conjunction with three of these non-parametric sequential designs for ordinal response data, two from the literature and a new design based on a random walk rule. We also compare with a non-parametric design for binary response trials, by keeping track of ordinal data for estimation purposes, but dichotomizing the data in the design phase. We find that a multidimensional isotonic regression-based estimator far exceeds the others in terms of accuracy and efficiency. A rule by Simon et al. (J. Natl. Cancer Inst. 1997; 89:1138-1147) yields particularly efficient estimators, more so than the random walk rule, but has higher numbers of dose-limiting toxicity. A small data set from a leukemia clinical trial is analysed using our multidimensional isotonic regression-based estimator.     

An information-motivation-behavioral skills (IMB) model-based HIV prevention intervention for truck drivers in India

The incidence of HIV/AIDS in India is increasing drastically, and truck drivers are seen as critical sources of HIV transmission due to their high rates of unprotected sex with multiple partners. An intervention based on the Information-Motivation-Behavioral Skills (IMB) model was compared to an information-only control condition in a randomized trial. IMB constructs were assessed among 250 male truck drivers immediately prior to and following implementation of the intervention, and sexual and condom use behaviors were assessed approximately 10 months later. The intervention consisted of a single-session group workshop with 5 interactive activities designed to address HIV prevention-related IMB constructs and to motivate condom use. Findings showed mixed support for the effectiveness of the intervention. There was an effect of the IMB intervention on attitudes, norms, behavioral skills, and intentions specific to condom use with marital partners, but no effects on constructs related to non-marital partners. There was some evidence of greater condom use with marital and non-marital partners at behavioral follow-up for participants in the IMB condition, and effects on condom use with marital partners were mediated by changes in IMB constructs. These findings provide initial evidence for the effectiveness of theoretically-based approaches to HIV prevention in India.     

POWER for reproductive health: results from a social marketing campaign promoting female and male condoms.

PURPOSE: To evaluate effects of a 6-month social marketing campaign on awareness of, attitudes toward and use of female as well as male condoms for 15-25 year-old-women. METHODS: Using a time-space sampling methodology, we conducted a cross-sectional survey of 3407 women at pre-campaign in 12 western U.S. neighborhoods on female and male condom awareness, attitudes, and use. Six of the 12 study neighborhoods were randomly selected to receive the POWER social marketing campaign designed to impact condom knowledge, attitudes, and use. The campaign was followed with another cross-sectional survey of 3,003 women in all 12 study neighborhoods on condom knowledge, attitudes, use and awareness of POWER materials. We compared pre-and post-campaign surveys to determine the efficacy of POWER and conducted post hoc analyses on post-campaign data to determine if exposure to POWER was related to higher levels of positive condom attitudes and norms and condom use. RESULTS: We found no differences between neighborhoods with and without the POWER campaign with regard to our primary outcomes. To diagnose reasons for this null effect, we examined outcomes post hoc examining the influence of POWER exposure. Post hoc analyses show some evidence that exposure to POWER was associated with condom use. In the context of the nested trial, this raises concerns that post test only evaluations are limited. CONCLUSIONS: Establishing the efficacy of a social marketing campaign is challenging. This group randomized trial showed a null effect. Social marketing campaigns may need to have more media channels and saturation before they can show behavioral effects. Using a nested design with randomization at the community level and probability sampling introduces rigor not commonly seen in evaluations of social marketing campaigns.     

Effectiveness of female and male condoms in preventing exposure to semen during vaginal intercourse: a randomized trial

Objectives

Comparison of male condom (MC) vs. female condom (FC) with respect to self-reported mechanical and acceptability problems and semen exposure using prostate-specific antigen (PSA) as an objective biological marker and evaluation of the effect of an educational intervention on self-reported problems and semen exposure, by condom type.

Design

Randomized crossover trial.

Methods

Four hundred women attending a family planning clinic in Brazil were randomized and either received in-clinic instruction or were encouraged to read the condom package insert; all used two FCs and two MCs. We measured the rates of self-reported user problems with MC and FC use and the rates of semen exposure during use (assessed by testing vaginal fluid for PSA).

Results

The educational intervention group reported fewer problems with either condom as compared with the control group (p=.0004, stratified by condom type). In both groups, self-reported problems were more frequent with FC use than with MC use (p<.0001, stratified by intervention). The educational intervention did not significantly reduce semen exposure. Overall, semen exposure occurred more frequently with FC use (postcoital PSA, >1 ng/mL; 22%) than with MC use (15%); the difference, however, was small and nonsignificant for high PSA levels (≥150 ng/mL; 5.1% for FC vs. 3.6% for MC).

Conclusions

In this study, the FC was less effective than the MC in preventing semen exposure during use and led more frequently to self-reported user problems. Both devices were highly protective against “high-level” semen exposure, as measured by postcoital PSA levels in vaginal fluid. In-clinic education may reduce user problems and increase acceptability and use of both devices.     

Social and Proximate Determinants of the Frequency of Condom Use Among African,Caribbean, and Other Black People in a Canadian City: Results from the BLACCH Study

African, Caribbean, and other Black (ACB) people are a priority group for HIV prevention in Canada, but little is known about condom use in this population. This exploratory community-based research project addresses this gap in knowledge. 125 sexually active ACB people completed a questionnaire covering condom use and social determinants of health. The data were analyzed using ordinal logistic regression and mediation analyses. 20.5 % of sexually active ACB adults used condoms consistently. Male gender, wealth, unstable immigration classes, and unsecure employment statuses were independently associated with more frequent condom use. Proximate determinants mediating these relationships included: not having a cohabiting regular partner, not disliking condoms, and having a history of unwanted sex. The proximate determinants mediated 85.7–97.6 % of the effects of the social determinants. These results link social context and proximate factors with condom use. They can be used to design evidence-informed interventions for ACB people.     

Condom promotion in women attending inner city general practices for cervical smears: a randomized controlled trial

BACKGROUND: Although condom promotion schemes have been widely piloted in UK general practices, there have been no rigorous evaluations of their effectiveness. OBJECTIVES: To see if a practice-based sexual health education intervention increases the number of women having smears who are given condoms and advice on avoiding STDs. To see if this low cost intervention affects subsequent condom use. METHOD: We conducted a cluster randomized trial of condom promotion in 1382 women aged <35 years attending 28 South London general practices for cervical smear tests. RESULTS: More women in intervention than control practices reported receiving advice on avoiding sexually transmitted infections (27% versus 10%) and being given condoms (28% versus 1%, P < 0.05). However, there was no difference in subsequent condom use, even in the 22% of women reporting >/=2 sexual partners in the previous year. CONCLUSIONS: To provide evidence of effectiveness, future interventions may need to be more intensive or focus on higher risk groups.     

Ordinal regression models for epidemiologic data

Health status is often measured in epidemiologic studies on an ordinal scale, but data of this type are generally reduced for analysis to a single dichotomy. Several statistical models have been developed to make full use of information in ordinal response data, but have not been much used in analyzing epidemiologic studies. The authors discuss two of these statistical models--the cumulative odds model and the continuation ratio model. They may be interpreted in terms of odds ratios, can account for confounding variables, have clear and testable assumptions, and have parameters that may be estimated and hypotheses that may be tested using available statistical packages. However, calculations of asymptotic relative efficiency and results of simulations showed that simple logistic regression applied to dichotomized responses can in some realistic situations have more than 75% of the efficiency of ordinal regression models, but only if the ordinal scale is collapsed into a dichotomy close to the optimal point. The application of the proposed models to data from a study of chest x-rays of workers exposed to mineral fibers confirmed that they are easy to use and interpret, but gave results quite similar to those obtained using simple logistic regression after dichotomizing outcome in the conventional way.     

Quasi-Likelihood Analysis of Patient Satisfaction with Medical Care

In health services research studies, a score for satisfaction with care is often determined for a patient by summing several items, each measured on a Likert scale with ordered response options indicating satisfaction with some aspect of medical care (e.g., five categories ordered 1-poor to 5-excellent). A common goal is to determine patient and physician level predictors of patient satisfaction using regression analysis. The large number of categories in the ordinal summed response variable may present obstacles to traditional analytic methods for ordinal data such as the proportional odds model for cumulative logits. Further, linear regression is generally known to be inappropriate. Quasi-likelihood methods provide a flexible and tractable alternative modelling procedure. Weak assumptions about the measurement scale may be made by estimating parameters that define a family of link functions. A quasi-likelihood analysis of data from a study of elderly patients' satisfaction with communication with their primary care physician is presented. Although several factors are significantly related to satisfaction, a diagnostic plot based upon cumulative deviances reveals inadequacy of fit for the patients with the lowest observed satisfaction.     

安徽省农村已婚妇女避孕套使用情况及其影响因素

目的了解安徽省艾滋病高发地区和低发地区农村已婚妇女的避孕套使用现状及其影响因素,为制定相应的避孕套正确使用的健康教育及行为干预的策略与措施提供依据。方法对安徽省艾滋病高发地区临泉县和低发地区岳西县3 817名农村已婚妇女进行问卷调查,采用单因素分析与Logistic回归分析其影响因素。结果农村已婚妇女避孕套使用率仅为6.2%,其中临泉县为4.7%,岳西县为7.6%。避孕套相关知识的知晓率为60.8%;其中临泉县为61.4%,岳西县为60.1%。文化程度越高、艾滋病知识和避孕套知识了解得越多、社会性别意识越强避孕套使用率越高。结论安徽省农村地区已婚妇女对避孕套知识缺乏了解,避孕套使用率低,对其开展有针对性的使用避孕套相关健康教育和行为干预迫在眉睫。     

市场营销在社区吸毒人群艾滋病行为干预中的应用

目的：目前国内艾滋病高危人群的行为干预均在探索阶段。该研究旨在探索社区吸毒人群预防艾滋病的行为干预模式。方法：干预方法主要是以同伴为基础的社区健康教育、建立医学咨询服务点、避孕套及注射器市场营销。结果：1998－1999年四川省性病艾滋病防治协会在冰山州西昌市3个术开展了社区吸毒人群艾滋病干预项目。社区人群及吸毒人员干预后艾滋病知识均有明显提高（P＜0．01）。吸毒人员均放弃了共用注射器行为,但避孕套的推广使用不理想。结论：以同伴为基础的社区健康教育和避孕套及注射器营销在预防控制艾滋病方面是有效的手段,但避孕套的推广工作有待进一步改进。     

Efficacy of an HIV/STI Sexual Risk-Reduction Intervention for African American Adolescent Girls in Juvenile Detention Centers: A Randomized Controlled Trial

Few HIV/STI interventions exist for African American adolescent girls in juvenile detention. The objective was to evaluate the efficacy of an intervention to reduce incident STIs, improve HIV-preventive behaviors, and enhance psychosocial outcomes. We conducted a randomized controlled trial among African American adolescent girls (13–17 years, N = 188) in juvenile detention from March 2011 to May 2012. Assessments occurred at baseline and 3- and 6-months post-randomization and included: audio computer-assisted self-interview, condom skills assessment, and self-collected vaginal swab to detect Chlamydia and gonorrhea. The Imara intervention included three individual-level sessions and four phone sessions; expedited partner therapy was offered to STI-positive adolescents. The comparison group received the usual care provided by the detention center: STI testing, treatment, and counseling. At the 6-month assessment (3-months post-intervention), Imara participants reported higher condom use self-efficacy (p < 0.001), HIV/STI knowledge (p < 0.001), and condom use skills (p < 0.001) compared to control participants. No significant differences were observed between trial conditions in incident Chlamydia or gonorrhea infections, condom use, or number of vaginal sex partners. Imara for detained African American adolescent girls can improve condom use skills and psychosocial outcomes; however, a critical need for interventions to reduce sexual risk remains.