Recruitment and retention of low-income minority women in a behavioral intervention to reduce smoking, depression, and intimate partner violence during pregnancy

Statistics and Epidemiology Unit, RTI International, 6110 Executive Blvd., Suite 902, Rockville, MD 20852-3903, USA.
BMC Public Health (Impact Factor: 2.32). 02/2007; 7(1):233. DOI: 10.1186/1471-2458-7-233
Source: PubMed

ABSTRACT Researchers have frequently encountered difficulties in the recruitment and retention of minorities resulting in their under-representation in clinical trials. This report describes the successful strategies of recruitment and retention of African Americans and Latinos in a randomized clinical trial to reduce smoking, depression and intimate partner violence during pregnancy. Socio-demographic characteristics and risk profiles of retained vs. non-retained women and lost to follow-up vs. dropped-out women are presented. In addition, subgroups of pregnant women who are less (more) likely to be retained are identified.
Pregnant African American women and Latinas who were Washington, DC residents, aged 18 years or more, and of 28 weeks gestational age or less were recruited at six prenatal care clinics. Potentially eligible women were screened for socio-demographic eligibility and the presence of the selected behavioral and psychological risks using an Audio Computer-Assisted Self-Interview. Eligible women who consented to participate completed a baseline telephone evaluation after which they were enrolled in the study and randomly assigned to either the intervention or the usual care group.
Of the 1,398 eligible women, 1,191 (85%) agreed to participate in the study. Of the 1,191 women agreeing to participate, 1,070 completed the baseline evaluation and were enrolled in the study and randomized, for a recruitment rate of 90%. Of those enrolled, 1,044 were African American women. A total of 849 women completed the study, for a retention rate of 79%. Five percent dropped out and 12% were lost-to-follow up. Women retained in the study and those not retained were not statistically different with regard to socio-demographic characteristics and the targeted risks. Retention strategies included financial and other incentives, regular updates of contact information which was tracked and monitored by a computerized data management system available to all project staff, and attention to cultural competence with implementation of study procedures by appropriately selected, trained, and supervised staff. Single, less educated, alcohol and drug users, non-working, and non-WIC women represent minority women with expected low retention rates.
We conclude that with targeted recruitment and retention strategies, minority women will participate at high rates in behavioral clinical trials. We also found that women who drop out are different from women who are lost to follow-up, and require different strategies to optimize their completion of the study.

Download full-text


Available from: Michele Kiely, Apr 30, 2014
1 Follower
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Domestic violence (DV) against women during pregnancy affects many women and unborn infants worldwide. Pregnancy presents a window of opportunity for health care providers to identify DV and provide appropriate intervention. The aim of this systematic review was to appraise the effectiveness of DV screening and interventions for women identified for DV through screening in pregnancy. The Cochrane Library, EMBASE, MEDLINE, and PsycINFO were searched from January 1995 to November 2009 to identify potentially relevant studies. Studies using any comparative methodology from both national and international arenas were included but had to be in the English language. Nine studies (13 references) met the inclusion criteria, five for screening and four for interventions. Of the five screening studies, the identification of DV was significantly higher compared to studies that used a nonstandardized screen or no screen at all. There was also evidence that recurrent screening throughout the pregnancy further increased identification rates. There was some evidence that interventions for pregnant women who had experienced DV reduced the amount of violence experienced by these women, but the evidence is very limited by the small number of randomized studies with small participant numbers. Further research is required to establish the most effective interventions for women who are identified at risk of DV during pregnancy.
    Trauma Violence & Abuse 10/2010; 11(4):190-201. DOI:10.1177/1524838010378298 · 3.27 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We apply Wijsman's theorem for the ratio of densities of the maximal invariant to signal detection applications. The method does not require the explicit use of a maximal invariant statistic or its density to derive the uniformly most powerful invariant (UMPI) test. We describe its use for common classes of detection problems and illustrate its use in examples. The analytic form of the representation provides new insight into the relationship between the UMPI and generalized likelihood ratio test.
    Signals, Systems and Computers, 2002. Conference Record of the Thirty-Sixth Asilomar Conference on; 12/2002
  • [Show abstract] [Hide abstract]
    ABSTRACT: Unintended pregnancy is associated with risk behaviors and increased morbidity or mortality for mothers and infants, but a woman's feelings about pregnancy may be more predictive of risk and health outcomes than her intentions. A sample of 1,044 black women who were at increased risk were enrolled at prenatal care clinics in the District of Columbia in 2001-2003. Bivariate and multivariate analyses assessed associations between pregnancy intentions or level of happiness about being pregnant and multiple psychosocial and behavioral risk factors, and identified correlates of happiness to be pregnant. Pregnancy intentions and happiness were strongly associated, but happiness was the better predictor of risk. Unhappy women had higher odds than happy women of smoking, being depressed, experiencing intimate partner violence, drinking and using illicit drugs (odds ratios, 1.7-2.6). The odds of being happy were reduced among women who had other children or a child younger than two, who were single or did not have a current partner, who had had more than one sexual partner in the past year and who reported that the baby's father did not want the pregnancy (0.3-0.6). In contrast, the odds of being happy were elevated among women who had better coping strategies (1.03), who had not used birth control at conception (1.6) and who had 1-2 household members, rather than five or more (2.1). Additional psychosocial screening for happiness about being pregnant and for partner characteristics, particularly the father's desire to have this child, may help improve prenatal care services and prevent adverse health outcomes.
    Perspectives on Sexual and Reproductive Health 12/2007; 39(4):194-205. DOI:10.1363/3919407 · 1.41 Impact Factor