Study Population and Procedures

The sample comprised pregnant and postpartum women (<3 months after delivery) enrolled in the Massachusetts Women, Infants and Children (WIC) Program in 2001. WIC clients are not counseled about HIV testing or referred for prenatal HIV testing.

We used a two-step sampling procedure. First, we selected WIC clinics in Massachusetts in the areas with the highest percentage of HIV-positive women of childbearing age. Selected clinics had an enrollment of > 100 women who were pregnant or up to three months postpartum. From this initial pool of 20 clinics, six clinics agreed to be in the study.
viagra soft tabs online

Table 1. Self-Reported Demographic and Social Characteristics of Respondents by Race/Ethnicity-Massachusetts, 2001

All participating clinics were in urban areas, five from the eastern part of the state and one from the western part of the state. Together they provided access to a sample that mirrored the racial/ethnic composition of the state’s female HIV-infected population in 2001, which was 36% non-Hispanic black, 34% non-Hispanic white, 28% Hispanic and 2% other race or unknown.
cheap female viagra

Second, WIC staff approached eligible Hispanic, black and white women and explained the study. Excluded were those clients who were not fluent in English, Spanish or Haitian Creole (the study languages). A $10 gift coupon was offered to respondents. Women provided written informed consent and then completed a self-administered questionnaire that took approximately 15 minutes. The Massachusetts Department of Public Health Institutional Review Board approved the study protocol.

MEASURES

The survey included questions from the CDC Perinatal Guidelines Evaluation Project HIV and Preg nancy Study surveys and CDC’s 1999 Pregnancy Risk Assessment Monitoring System (PRAMS) survey, including CDC PRAMS questions on partner-perpetrated physical and emotional abuse. Questions were pilot-tested for comprehension with women who were representative of the study sample.

Table 2. Self-Reported HIV-Related Test Barriers, Beliefs and Knowledge by Race/Ethnicity—Massachusetts, 2001

We assessed women’s reports of HIV counseling during their most recent pregnancy or shortly after giving birth by a “yes” response to one of six items. Women’s report of HIV testing (outcome variable) was documented if it occurred at any point during pregnancy.

Shown in Table 2 are the items measuring women’s self-report of HIV-related testing barriers, beliefs and knowledge. Women were asked if they knew their HIV status but were not asked to report their serosta-tus because of possible conflicts with Massachusetts confidentiality law. Source your medication needs online. Buy generic cymbalta home delivered

Statistical Analysis

Race and ethnic differences were calculated by cross-tabulation using the Chi-squared test (Tables 1 and 2). The 95% confidence intervals around the proportion of women counseled about the HIV test and tested show the range of values within which the percentage of women counseled and tested would lie (Table 3).

Table 3. Probability of Being Counseled about HIV Test and Tested for HIV by Site of Care and Race

Multivariable linear regression analyses were used to estimate proportional differences in prenatal HIV (Zerit drug еreating HIV infection when used in combination with other medicines) testing for the total sample and stratified by race. For ease of interpretation, proportional differences are reported as estimates of percents, with 95% confidence intervals derived from standard errors of beta estimates.

In the total sample model (Model 1: All women), all terms were entered into the linear regression model simultaneously. Each racial category was coded as a binary variable, with white women serving as the reference group. Other covariates included (♦indicates reference group): age (14-21, 22-29, 30-47); currently married (yes, no); monthly household income (<$ 1,000, $1,000 to $1,999, >$2,000, unknown); recent victim partner violence (yes, no); belief that all pregnant women should be tested for HIV (yes, no); pregnancy status at the time of data collection (first trimester, second trimester, third trimester, trimester unknown, postpartum); and three types of prenatal care sites [community health center, hospital outpatient clinic, private/health maintenance organization (HMO)].

Next, we conducted race-specific analyses using the same covariates in each model (Model 2: Black/Haitian women, Model 3: Hispanic women, Model 4: White women). Since the likelihood of being tested for HIV (Viramune tablets treating HIV infection) was, in part, dependent on receiving counseling, we ran all models with and without prenatal HIV (Medication Viramune treating HIV infection) counseling as a covariate.