Subjects were participants in another survey, which evaluated the preventive practices in minority women. A full description of the cross-sectional study is published elsewhere. A convenience sample of women aged 50 and above were recruited for the survey from the inpatient and outpatient units of a teaching hospital that predominately serves minority patients. Internal medicine house staff provided patient care under the direct supervision of internal medicine faculty. The questionnaires were administered by trained volunteer physicians between January 1998 to April 1999. All subjects received the same standard face-to-face interview. Approximately 95% of the women agreed to participate in the study. Among the 354 women asked to participate, 203 were African Americans, and 134 were Hispanics. The participants were specifically asked whether they were using any form of HRT at the time of the survey. If they responded “yes”, they were considered as currently using HRT. Among the 337 minority women, only 333 women responded whether they were currently using HRT or not. For further analysis only 333 minority women were included.

The principal dependent variable examined in this analysis was current estrogen use. Demographic variables studied were age, occupation, insurance, marital status, education and perceived health status. Lifestyle characteristics examined were cigarette smoking, alcohol use and regular exercise. In addition, their medical conditions and medications were also noted. The subjects were asked whether they have a regular medical provider or not, the location of the interview, and whether they were inpatient or outpatient. Preventive practices assessed were monthly self-breast examination, mammogram (mammogram within two years), Pap smear (ever had a Pap smear), stool occult blood examination, sigmoidoscopy and calcium use. cheap levitra online

Statistical Analysis

Descriptive analyses were first used to examine the characteristics of the sample. Bivariate analyses were conducted to determine the relationship between HRT and sociodemographic, health and preventive practices. The Pearson Chi-square statistics was used to assess simple bivariate relationships. To assess for multicollinearity, a correlation analysis was also conducted between each of the independent variables, and the interrelations among them were modest. Factors with P<0.25 at bivariate level and without a problem of collinearity were entered into a multivariate logistic regression model. Multiple logistic regression analyses were conducted to examine the independent impact of each independent variable on HRT, adjusting for other potential confounding variables such as age, health status and regular activity. Logistic regression results are presented as odds ratios and 95% confidence intervals. In all cases, a two-tailed PO.05 was considered to be statistically significant. The goodness fit of the models was assessed by Hosmer-Lemeshow statistics, and there was no evidence of a lack of fit in the selected model (p=0.392). The statistical program for Social Sciences SPSS version 11.5 was used for data analysis. fosamax 70 mg