Sample and Recruitment
In 2004, we surveyed by telephone patients from four inner-city health centers. These centers were included because they were known to serve large numbers of persons of color and of disadvantaged persons and were willing to implement quality improvement initiatives. Most of these patients were part of a panel of 325 persons that was initially selected from a random sample (of billing records) of patients who were >50 years old and had been seen at one of the health centers in the last year and who continued to participate in our studies. A personalized introductory letter and a letter from the health centers endorsing the project and encouraging participation were sent to each of the sampled patients. A $10 honorarium was offered for completing the survey. Of the panel, 248 participated, 14 refused and 15 were ineligible—for a response rate of 80% [248/(325-15)]. This panel was augmented by a convenience sample that, following HIPAA guidelines, was recruited by the sites using an introductory letter and recruitment sheet that was distributed by office staff during visits. Among these 104 new recruits, 83 interviews were completed, 0 refused and one was ineligible—for a response rate of 81% (83/103). As such, the total sample for this study was 331 patients.
The items in the questionnaire and the layout were designed by a multidisciplinary team using an iterative process after a review of literature about ethical concerns about genetic testing. Items on the questionnaire assessed self-reported demographic characteristics, including race, health behaviors, attitudes and beliefs regarding genetic determinants of disease, genetic testing and religious beliefs as they relate to genetic testing. Given the exploratory nature of the study and the desire to investigate ethical concerns, we did not focus on a particular behavioral theory. However, we have been influenced by and previously used the Theory of Reasoned Action—particularly the components related to beliefs and attitudes about outcomes. The final questionnaire included approximately 36 multiple-choice and three-point Likert scale questions related to this paper, depending on skip patterns. Levitra plus
From August to October 2004, telephone surveys of this sample were performed by trained interviewers using computer-assisted telephone interviewing (CATI) permitting direct data entry during the interview. The CATI system managed the sample of persons to be contacted—that is, it randomly assigned people to be contacted to each interviewer and recorded the outcome of attempts to reach them. The CATI system also directed question sequence, reduced unintentionally skipped questions and provided automatic range checks.
In order to specifically study racial differences between African Americans and Caucasians, we used race as an outcome variable, as has been done elsewhere, and excluded the 17 questionnaires from participants who reported that they were neither African-American nor Caucasian. Attitudes about genetic testing were dichotomized into “agree/neutral” or “disagree.” We chose these categories because only 1-5% of participants gave “neutral” responses to specific attitudinal items. Bivariate associations were examined using Chi-squared tests (or Fisher’s exact tests in the case of small cell counts). Logistic regression analysis was performed to determine variables significantly associated with race. In this multivariate model, we included as independent variables all variables associated in bivariate analyses with the dependent variable at the PO.10 level or if specified a priori (i.e., age). Because attitude about genetic testing specific to racial discrimination was highly correlated (r=0.58, PO.001) with the attitude about genetic testing specific to the handicapped, we only included the former variable in our multivariate model. Although we tested for interactions between education group and each attitude in our model, no significant interactions were found. All statistical analyses were performed using SAS 8.2 statistical software (SAS Inc., Cary, NC). Statistical significance was set at P<0.05 and data were unweighted. You can afford your pills. Buy canadian viagra online
This project was approved by the Institutional Review Board of the University of Pittsburgh.