We conducted a community-based, cross-sectional interviewer-administered survey of women 40 years of age and older who spoke English or Haitian Creole in eastern Massachusetts neighborhoods having a high proportion of Haitian residents. We used an area probability sample to select subjects. After constructing a sampling frame by using the City of Boston and Cambridge Assessing Department’s Property Parcel Data for fiscal year 1997, we interviewed Haitian community leaders to identify housing blocks with high concentrations of Haitian families. Blocks estimated by two or more informants to contain 10 Haitian households or at least 20% Haitian households were included in the survey universe.
In each block, all housing units were enumerated, and 12 units were randomly selected for interview. Interviewers who visited housing units having Haitian residents also asked them to suggest blocks where other Haitian families were likely to reside.
Survey Instrument and Data Collection
We developed a survey instrument that included questions on ethnicity, age and education, as well as selected knowledge, attitudes and belief variables in relation to the utilization of screening. Researchers adapted questions from the National Health Interview Survey (NHIS), the NHIS cancer control supplements and the Cancer Control Needs in Multi-Ethnic Communities (CCNMEC) survey conducted in New York City in 1992. We field-tested the questionnaire by administering it to 20 community members. We incorporated their critique in the final questionnaire, hired interviewers from the population under study. The questionnaire was translated into Haitian Creole, back-translated into English by a different translator, compared translations and resolved discrepancies. The Institutional Review Board of the Boston Medical Center approved all aspects of the study.
Four trained bilingual and bicultural female interviewers conducted 25-minute interviews in the residents’ homes between June and August 1997, in English- or Haitian-Creole-based on the preference of the subject. Interviewers returned at least four times to attempt to make a contact with an eligible subject. The interviewers read a prepared informed consent form and obtained verbal consent before the interview. At the end of the survey, all participants received educational materials and information on community resources for free mammograms.