herhealedHealth Education Research1465-36480268-1153Oxford University Press10.1093/her/cyn026ORIGINAL ARTICLESMoving from theory to practice: implementing the Kin KeeperSM Cancer Prevention ModelWilliamsK. P.1*MullanP. B.2TodemD.31Obstetrics, Gynecology and Reproductive Biology, Michigan State University, East Lansing, MI 48824, USA2Medical Education, University of Michigan Medical School, Ann Arbor, MI 48109, USA3Epidemiology, Michigan State University, East Lansing, MI 48824, USA*Correspondence to: K. P. Williams. E-mail: karen.williams@ht.msu.edu420092952008242343356Published by Oxford University Press 2008.2009This paper presents the rationale and findings of a feasibility and process study of the Kin KeeperSM Cancer Prevention Intervention. An observational cohort study design was implemented with African-American women in synergistic female family relationships. Community health workers (CHWs) from two Michigan public health programs recruited women to serve as ‘kin keepers’ who in turn recruited their female family members. In total, 161 kin keepers and female family members were sampled. Trained CHWs led kin keepers and family members in learning about breast cancer. Data methods included baseline and post-training administration of a breast cancer literacy assessment, post-training focus groups and review of personal action plans. To validate the feasibility of the process, a linear mixed-effects regression with 97% power was identified and differences in pre–post scores were detected at 5% significance level. Adjusting for family random effects, breast cancer literacy scores increased for all participants recruited (P-value = 0.0004) suggesting that the process was feasible. Analysis of focus groups and action plans indicated that participants valued the instruction and planned to act upon it. This experience with kin keepers and their families offers encouragement that the theoretical model and its community-based delivery can continue to enhance scholarship dedicated to ameliorating health care disparities.IntroductionThis paper presents the rationale and findings of a feasibility study of the Kin KeeperSM cancer prevention program, which began as a theoretical model positing that African-American female family relationships are synergistic such that empowerment or self-efficacy education for individual women could engage other female family members. Assembling to learn about breast cancer prevention and early detection would empower African-American female family members [1]. In this model, community health workers (CHWs), who serve as a liaison between the family and the health care system and have a relationship of trust with one of the family members, serve as a point of access, recruiting women from their public health program catchment area to serve as ‘kin keepers’. These kin keepers, in turn, identify other female family members. The trained CHWs lead the kin keepers and their families in learning about their individual and familial risks for breast cancer. They also learn about screening guidelines, how to perform a breast self-examination (BSE), what to expect when obtaining clinical breast examinations (CBEs) and mammograms and how to share this education with their female family members. The kin keeper's home serves as the safe learning environment for this education.Thus far, approaches to cancer education, prevention and screening involved using the influences of either the health care provider or community-based programs. Although some of these programs may not have been presented as cancer literacy programs per se, they have promoted health literacy in one form or another. For example, community-based interventions render technical assistance by disseminating knowledge regarding cancer prevention and available services. Some interventions are attached to ongoing programs in the community where employment skills are being offered, fostering self-help that leads to empowerment, which is highly valued in the African-American community [2]. As it relates to the Kin KeeperSM model and specific to this study, empowerment is a process and an outcome [3]. It influences women to consider changing their behavior (a process), including the degree of effort they invest in achieving change and the long-term maintenance of behavioral change [4]. The empowerment that the women experience builds capacity (an outcome), which often extends to other areas of their lives [5]. Effective community-based interventions take a variety of forms and locations, from being housed at a local clinic to forming a partnership with a local community-based organization [6–16]. A common theme of these approaches is the use of African-American women as employees or volunteers of the clinic or program. Programs that work with African-American faith-based organizations, such as churches, are also considered community-based approaches. Researchers found that faith-based organizations and religious practices influence African-American women's decisions about cancer prevention and screening behaviors [17–19]. Some investigators have begun to expand their community-based research by designing multigenerational or life span cancer control programs involving various stakeholders, such as middle school and high school children, churches and established senior citizen programs [20].Another factor that influences cancer prevention behaviors of African-American women is the health care provider. Provider-focused approaches are of particular importance because they address issues of health care or access to health services for African-American women. Researchers have sought to investigate the influence that physicians or other health providers have on decisions by African-American women to engage in cancer prevention behaviors [15, 21–25]. Provider recommendation is probably one of the few accepted and undisputed predictors of mammography screening [26]. It is the strongest predictor of whether older African-American women will get screened [27]. Although provider recommendations are a strong predictor of behaviors related to cancer prevention [28], physicians often forget to discuss prevention in busy clinics providing primary care for chronic disease management [29]. One approach to overcoming this barrier has been to use reminders as cues to action. Such strategies have been effective in increasing breast cancer screening rates by 20% at reasonable costs [29, 30].Culture is a vital component of health communications and often serves as a barrier between African-American women and cancer screening [31–34]. ‘A group's culture influences their knowledge, attitudes, and personal practices, which affects their responses to health information’ [34]. Therefore, health literacy must be understood and addressed in the context of culture and language [35]. ‘Culture values influence the understanding that people bring to the acquisition of new knowledge and skills’ [36]. Understanding the sociocultural basis of variables that influence self-care practice and the influence of culture on language, interventions that are effective and culturally sensitive can be designed and implemented [36]. Friedell et al. [37] found that many individuals with limited literacy, rather than reading about the disease, more often obtained information about cancer from family and others who have had experiences with a late-stage diagnosis. Additionally, research studies have indicated that African-Americans tend to be more oriented towards relationship to kinship [38, 39] and possess a strong oral tradition [40, 41]. As a result, African-American women turn to their family members for guidance on cancer matters. Therefore, education of all family members is essential in order to eliminate the sharing of false information.These community-based programs and provider initiatives have met with varying degrees of success and have provided us with insights into working with these women, who have only recently gained attention as research participants [42–44]. Thus far, we have learned that we must adapt our research to the population we are studying, rather than adopt a one-size-fits-all approach. We have learned that building trusted relationships with the research team makes a difference in our ability to recruit African-American women into our studies [45]. We have also learned that tailoring interventions that integrate cultural aspects of the population being investigated are also effective. On the other hand, we have learned that even effective community-based approaches have limitations. They cast a narrow net, often leaving out women who are not connected with a community-based organization or missing women who are not plugged into the health care system through insurance and regular visits. They are often low-income women who do not have the opportunity to learn about cancer risks or low-cost or free mammograms or how to conduct a BSE. The research that has so far been brought to bear on African-American women and their cancer prevention behaviors has shown us that there is no one way to do things nor is there a quick fix for engaging this population. However, there is a lack of understanding of the role of the African-American family, specifically female family members, in influencing cancer prevention practices among other females in the family.Theoretical framework for Kin KeeperSMThe Kin KeeperSM model follows the human ecological perspective, which views an individual woman, her health and development and the health and development of her family in the context of physical, psychological, social and cultural environments [46, 47]. Environments are often described as nested and are depicted as concentric circles from proximal to distal, including individual, family and community. Environments can enhance or limit the potential for health and positive growth and development. An ecological perspective calls attention to relationships between persons and their family and community and transactions that occur among them. These transactions are dynamic and develop over time. To be successful in reducing cancer disparities and improving health-related behaviors, health promotion and prevention programs must identify and address the individual, family and community factors and the relationships among such factors that influence cultural and lifestyle practices of African-American women [48].Kin KeeperSM uses a conceptual framework based on the premise that the natural ways that African-American women communicate various health messages to females in their families (mother, daughter(s), grandmother(s), sister(s) and aunt(s)) can be used to influence them to engage in cancer prevention and screening behaviors. Building on this generational and cultural behavior, the Kin KeeperSM was developed as a women's health advocacy intervention model that naturally integrates health literacy. It uses various indigenous lay health models along with community development principles of capacity building [49]. The rationale for this model is as follows: (i) women serve as point persons for their own and their families’ health and well-being by making most of the decisions for health care; (ii) African-American female family members have the ability to influence each other to conduct regular BSE, CBE and mammograms at the appropriate time and (iii) when cancer prevention options, such as a chemoprevention trial, are presented in a non-intimidating environment, such as the home, by a trusted person, such as a CHW, who serves as a liaison between the woman and the healthcare system, African-American women will be more likely to be actively engaged in their health. The Kin KeeperSM model makes two assertions: (i) understanding individual and familial cancer risk factors will influence early detection behaviors [5] and (ii) the model helps African-American women make the connection between cancer risk factors, family cancer history and cancer detection behavior. The model helps to explain various dynamic factors—the community, the provider and the family—that influence African-American women to participate in cancer prevention and screening practices [50].The Kin KeeperSM model uses CHWs (also known as lay health advisors, community health workers, indigenous community health workers, community health advisors, lay health home visitors and health aides), who work in community settings and serve as connectors between health care consumers and providers, to promote health among medically underserved groups [51]. They promote healthy living by providing education about how to prevent disease and injury and helping community residents understand and access formal health and human service systems [52]. Using CHWs has been particularly relevant among underserved and older African-American women, who often seek advice from female family members [53, 54]. The literature is also clear that the use of CHWs is effective in connecting hard-to-reach women with health care services [53, 55–58]. Million-Underwood et al. [59] found that when African-Americans were provided with information regarding the benefits of participating in screening and treatment trials, they expressed a great willingness to participate. CHWs facilitate the understanding of the benefits of prevention and early detection. Like the patient navigators who act as patient advocates [60] in the Kin KeeperSM model, CHWs assist women in navigating the health care system for preventive services and provide them with additional resources.Although existing empirical research provides important lessons about cancer control outreach to African-Americans [61, 62], the continuing disparity of breast cancer provides evidence of the need for innovative cancer control models for prevention and early detection. Between 2000 and 2003, African-American women in the United States suffered a 36% higher death rate from breast cancer than Caucasian women [63, 64]. This disparity persists despite African-American women's lower breast cancer incidence rate. The best controllable predictor of a favorable outcome for breast cancer is early stage diagnosis. A major factor associated with earlier diagnosis is use of early detection screening services, including mammography and CBE. To be most effective, services like mammography must be received on a regular, recurring basis at recommended frequencies.Our study combined three forms of scholarship: theory, research integration and application [65]. We conducted an observational cohort study, using the Kin KeeperSM Cancer Prevention Intervention. In this paper, we report the results of the initial implementation of the Kin KeeperSM model and potential implications.MethodsThis study was conducted in partnership with two community-based public health programs in Michigan. In one program, each Medicaid-eligible woman deemed at risk for less favorable pregnancy outcomes was assigned a CHW through Maternal and Infant Support Services (M/ISS). Their participation in the M/ISS can range from 2 months prenatal through 12 months postpartum. The other community-based program, Village Health Worker Program (VHWP), focuses on educating and mobilizing high-risk groups to receive screening for diabetes and cardiovascular disease, using CHWs and the social networking method of hosting house parties [66]. An advantage of working with both public health programs is that women in these programs are accustomed to learning about risk and benefits related to their bodies. Therefore, they could be ready to become more empowered and increase their personal capacity by learning about early breast cancer screening, the types of screenings and the guidelines for age/risk-appropriate screenings. Michigan State University's Institutional Review Board for the study of human subjects approved this study.Research participantsIn this manuscript, we focus on the 161 kin keepers and family members recruited through the efforts of our CHWs. Of these women, 49% were aged ≥40 years and therefore met the American Cancer Society's guidelines for annual mammograms. Further descriptive statistics on the women's socio-demographic characteristics are shown in Table I.Table I.Socio-demographics of participants by public health programVariableHealth programVHWP (within-row %) [within-column %]M/ISS (within-row %) [within-column %]Total [within-column %]Age (years)    <3010 (29) [10]24 (71) [39]34 [21]***    30–3935 (75) [35]12 (26) [19]47 [29]    40–4917 (65) [17]9 (35) [15]26 [16]    ≥5037 (69) [37]17 (32) [27]54 [34]Education    College degree or higher15 (65) [15]8 (35) [13]23 [14]**    Some college28 (58) [29]20 (42) [32]48 [30] High school diploma or General Education Development50 (69) [52]22 (31) [35]72 [45]    Less than high school4 (25) [4]12 (75) [19]16 [10]Income    <$10 00015 (40) [15]23 (61) [40]38 [24]***    $10 000–$19 99943 (75) [44]14 (25) [24]57 [37]    $20 000–$39 99927 (67) [28]13 (33) [22]40 [26]    ≥$40 00013 (62) [13]8 (38) [14]21 [13]Employment status    Full time or part time38 (50) [38]38 (50) [60]76 [47]***    Unemployed or laid off19 (59) [19]13 (41) [21]32 [20]    Retired20 (91) [20]2 (9) [3]22 [14]    Stopped working due to disability4 (40) [4]6 (60) [10]10 [6]    Homemaker8 (73) [8]3 (27) [5]11 [7]    Self-employed10 (91) [10]1 (9) [2]11 [7]Marital status    Married22 (65) [23]12 (35) [19]34 [22]    Single/never married45 (52) [48]41 (48) [65]86 [55]    Separated9 (75) [10]3 (25) [5]12 [8]    Divorced11 (69) [12]5 (31) [8]16 [10]    Widowed7 (79) [7]2 (22) [3]9 [6]Health insurance type    Plan from employer49 (77) [50]15 (23) [25]64 [41]*    Plan purchased directly5 (100) [5]0 (0) [0]5 [3]    Medicare/Medicaid31 (47) [32]35 (53) [59]66 [42]    No coverage13 (59) [13]9 (41) [15]22 [14]Kin keeper or family member    Kin keeper32 (59) [32]22 (41) [37]54 [34]    Family member69 (65) [68]38 (36) [63]107 [66]Ever had self breast exam    Yes98 (67) [100]49 (33) [80]147 [93]***    No0 (0) [0]12 (100) [20]12 [8]Ever had clinical breast exam    Yes92 (65) [96]50 (35) [82]142 [90]***    No4 (27) [4]11 (73) [18]15 [10]Had clinical breast exam last year    Yes84 (69) [87]37 (31) [60]121 [76]***    No13 (34) [13]25 (66) [40]38 [24]Ever had a mammogram    Yes51 (62) [53]31 (38) [50]82 [52]    No46 (60) [47]31 (40) [50]77 [48]Had a mammogram last year    Yes45 (68) [47]21 (32) [34]66 [42]    No51 (55) [53]41 (45) [66]92 [58]Two-sided asymptotic likelihood ratio test: ***P-value <0.001, **P-value <0.01, *P-value <0.05.Community health workersNine CHWs employed by one of two public health programs—the M/ISS or VHWP—served as paid CHWs for the current study. They were selected by their respective supervisor based on their experience as CHWs, their effectiveness with their clients and their willingness to work on this project. These CHWs each had more than 8 years experience, including recruiting their clients for university-based research projects. They all completed 16 hours of Kin KeeperSM Cancer Prevention Intervention training on topics including breast and cervical cancer prevention and early detection and how to implement the Kin KeeperSM model. From her caseload of ∼30 women, each CHW was responsible for enrolling a minimum of five African-American women to serve as kin keepers. This involved arranging two home visits with family members whom the kin keeper had identified, performing two follow-ups with individual family members (mailing a 6-month postcard reminder and conducting a 12-month in-person interview), making appropriate referrals and completing follow-up documents. Although some CHWs failed to meet the recruitment goal, others recruited more than five kin keepers.Kin keeper eligibilityTo be eligible, women had to be African-American, aged ≥18 years and participants in one of the two community health programs. Only after giving informed consent could the public health client be recruited into the study and take on the role of a kin keeper. As a kin keeper she was asked by her CHW to (i) complete a survey identifying the African-American family lineage of her mother and father and paternal and maternal grandparents and (ii) invite adult bloodline female family members (mother, grandmothers, sisters and aunts) to two home visitations to learn about breast cancer prevention and early detection screening. This proved to be a challenge for some CHWs with clients who self-identified as African-American but whose parents or grandparents included a non-African-American. The kin keeper also agreed to assist the CHW, if necessary, in locating family members for the 12-month interviews. The kin keeper received a stipend for her participation and for recruiting her family. A total of 54 women participated as kin keepers.Family criteriaUp to five individual family members, including the kin keeper, were eligible for inclusion in the educational sessions as a family unit. At least two members were to be between ages 35 and 64 and thus eligible for free/low-cost high-risk breast cancer screening programs. For participating as a family unit, the group received a restaurant gift certificate. In addition, each research participant received a gift bag with various breast cancer learning materials. Our sample consisted of 54 families and 161 individuals (see Fig. 1).Fig. 1.Study participant recruitment and setting.Data collectionQuantitative data were collected using three instruments: a socio-demographic questionnaire, a pre/post-breast cancer literacy assessment and a personal action plan. We collected qualitative data through the use of focus groups.Pre/post-breast cancer literacy assessmentThe breast cancer literacy assessment measures a woman's functional understanding of breast cancer [67]. It is a 16-item scale administered in a pre- and post-test format. It uses a response format that combines multiple choice and true/false responses. Its three domains are (i) general cancer awareness, (ii) breast cancer knowledge and (iii) breast cancer prevention and control. The CHW reads each question aloud to the participants, thereby enabling each person to independently complete the assessment regardless of her reading level. The assessment takes about 10 minutes to complete. The instrument formative study has been reported [67]. In this study, we present findings on pre–post test scores only as validation that our implementation process is feasible and effective, as reflected by increases in test scores.Personal action planIn the Kin KeeperSM model, women have an opportunity to establish a personal action plan related to breast cancer screening. Women complete this with the assistance of the CHW and then refer back to it during the 12-month follow-up visit. The CHW keeps a copy, the participant retains a copy and a copy remains in the research office.Focus groupsFollowing the second home visit, we conducted two focus groups, one with five CHWs and the other combining kin keepers and family members in a group of six. This enabled us to elicit exploratory and confirmatory perceptions [68]. Both sets of participants were associated with the M/ISS program. Each CHW was asked to recruit two participants to be in the family focus group session. All focus group participants, including the CHWs, were given a cash stipend for their participation. Funding limitations prevented us from conducting focus groups with the VHWP that was located 160 miles from the M/ISS program.We asked the kin keepers/family members questions focusing on their opinion of the overall program and about the unique features of this program, such as the CHW reading all the materials at the home visits, the personal action plans and their experiences discussing their goals with other members in their family. In the CHWs’ focus group session, we asked them about their perception of how their families received/liked the intervention, what they learned from the research project and things they would change. Each focus group session lasted ∼1 hour. An observer took notes while a professional transcriber recorded all responses and comments. To protect confidentiality, each participant was identified numerically rather than by name in the transcriptions.ImplementationThe Kin KeeperSM educational session took place during family home visits. The CHW and kin keeper together scheduled the first family visit. The second family visit was scheduled by the group during the first visit, when all were in attendance at the kin keeper's home. Both visits were scheduled for evenings or weekends. The second visit occurred 1–3 weeks after the first one. Each visit lasted ∼1.5 hours. Before and after each visit, all family members completed the breast cancer literacy assessment.During the first visit, after each family member completed the assessment, the CHW instructed the family on the benefits of CBE and mammograms, what to expect in the clinic and individual and familial risks. Participants learned how to perform a BSE with the use of breast models. Each person then completed the assessment again. At the second visit, after the assessment, the CHW reinforced key information presented during the first visit and cleared up any misunderstandings. Women were able to practice their BSE technique. At the end of this visit, the CHW assisted individuals in developing a personal action plan.Statistical methodsTo describe the participants, we performed descriptive analyses and used chi-square tests. To establish the feasibility of the Kin KeeperSM Cancer Prevention Intervention and the implementation process, we calculated participant drop-out rates between the first and second home visits. We also analyzed participants’ understanding of key concepts and screening and prevention practices through their responses to the breast cancer literacy assessment. To validate the feasibility of the Kin KeeperSM Cancer Prevention Intervention, we performed a paired t-test of the pre/post-test scores, then controlled for covariates using a random-effects mixed model with the difference in pre- and post-test scores as the dependent variable. Explanatory variables included in the model were the public health program (M/ISS or VHWP) and the woman's kin keeper status (i.e. kin keeper or other family member). Interactions between these explanatory variables were also assessed.Responses from members of the same family are likely to be correlated because two individuals within the same family will have responses that are more similar than two individuals in different families. Ignoring this dependence is apt to yield incorrect standard errors [69]. To address this, we used the principle that similarity in the responses within a family may be conceptualized as due to unobserved random family-level variables. Therefore, a family random-effects term was added to the mean model. This adjusts for heterogeneity of families, which can be due to unmeasured family-level random effects (predispositions).A simple linear mixed-effects model (also referred to as the intercept-only model) to analyze the differenced breast cancer literacy scores denoted as yijklfor a woman l in the family i, participating in the program j and of kin keeper status k can be formulated as yijkl = δ+Fi + εijkl, where δ is the mean score difference, Fi is the random family effect and εijkl is the measurement error. It is assumed that the random effect Fi and the measurement error εijkl are independent and distributed as central normal distributions with variances σF2 and σ2, respectively.The fixed-effects parameter δ captures the overall change over time across all levels of the kin keeper status and the health program. If the null hypothesis δ = 0 is rejected at 5% significance level, we can then conclude that there is some evidence from the data of a difference between the pre- and post-intervention scores with adjustment for the family random effects. The model can also be extended to assess whether the kin keeper status or the health program affects the change. The model is formulated as yijkl = δ0 + βj + γ k + Fi + εijkl. Tests on βs and γs assess the effect of the kin keeper status and the health program on the outcome difference. This model can also examine the interaction between explanatory variables. The estimates of these mixed models are used to compute the posterior power by simulation to show that the study has adequate power to detect differences observed in the data.Qualitative analysisThe focus group transcripts and observation notes were discussed, analyzed and validated by consensus in accordance with grounded qualitative theory [70]. Investigators independently read transcripts and extracted key comments associated with knowledge and experiences with breast cancer prevention and early detection using pragmatic and semantic content analysis [70]. The latter focuses on the meaning of the statement, the former on why something was said. Observer notes were compared with the categories for investigators’ independent coding. Using the grounded theory of analytic process provided a context for understanding the practicality of the Kin KeeperSM model.ResultsResults of the bivariate analyses using the chi-square tests showed that age, education and income were significantly different between women from the two health program groups (see Table I). Similar findings are listed in Table I concerning their self-reported breast examination behaviors.To assess feasibility of the Kin KeeperSM Cancer Prevention Intervention, we calculated the overall participant drop-out rate and found a very low rate of 3%. The drop-out rate for women recruited through the M/ISS health program was 5%, while that for women recruited through the VHWP was 2%. The low drop-out rates suggest that the Kin KeeperSM Cancer Intervention and the implementation process are feasible and capable of retaining participants over time. Results of the two-sample t-tests for the pre/post-test scores show that there was a significant increase in test scores. Detailed results by health program and by kin keeper/family member categories are shown in Table II. The increase in scores across the board suggests that our process was indeed feasible and effective for different categories of women.Table II.Individuals' matched (paired t-test) pre/post-projects breast cancer literacy scoresPaired differencesMean difference in percentage scoreStandard deviationStandard error of the mean95% CI upper95% CI lowerdfP-valueVHWPa (n = 100)4.311.71.1711.9896.63699<0.001M/ISS (n = 60)15.316.12.08211.14619.47959<0.001Kin keepers (n = 54)10.015.22.0745.79514.11353<0.001Family members (n = 106)7. 714.11.3724.94410.386105<0.001Overall (N = 160)8.414.51.1476.17310.702159<0.001CI, confidence interval.aVHWP—Diabetes Cardiovascular ProgramTo validate the result that the Kin KeeperSM Cancer Intervention was feasible, we assessed changes in cancer literacy scores using random-effects mixed models. While results from these models could be construed as assessing program intervention effect, here we use them solely to validate the feasibility of the Kin KeeperSM model. If the women's cancer literacy scores did not increase (while controlling for fixed and random effects), then we would have inferred that the practicality and feasibility of the Kin KeeperSM model were questionable. Results are shown in Table III. As predicted, no significant differences were detected between the cancer literacy scores of kin keepers and family members, nor were there any significant interactions between kin keeper status and community-based public health programs. Controlling for random effects of between-subject correlations, the cumulative improvement in mean scores was 5.9% following the implementation of the Kin KeeperSM model. The post-implementation breast cancer literacy scores are significantly higher than the initial scores at 5% significance level (P-value = 0.0004) when family random effects are accounted for. In the main-effects model, the increase in cancer literacy scores is significant (P-value = 0.0377). While all participating women increased their breast cancer literacy scores over time, women in the M/ISS program appeared to increase their scores more compared with women in the VHWP program. Interaction between the community-based public health program variable and kin keeper status was assessed and found to be non-significant at the 5% level (P-value = 0.6667). The increase in scores across the board suggests that our finding of the Kin KeeperSM Cancer Prevention Intervention as feasible was valid.Table III.Random-effects models: parameter estimates and inferencesEffectsIntercept-only modelMain-effects modelMain-effects with interactionsEstimateStandard errorP-valueEstimateStandard errorP-valueEstimateStandard errorP-valueIntercept0.0590.0150.00040.0590.0150.00040.0940.0270.0009Health program (VHWP)—−0.0650.0370.0377−0.0710.0340.0397Kin keeper status (kin keeper)—0.0220.0200.26130.0120.0320.7132Program × Kin keeper status (kin keeper and VHWP)——0.0180.0410.6667Family random-effect variance (σ2F)0.00800.0027—0.00730.0026—0.0070.003—Error term variance (σ2)0.0140.0020—0.0140.0019—0.0140.002—Power analysisWith use of the intercept-only model, yijkl = δ + Fi + εijkl, there is a power of 97% to detect the difference of 0.059 on the log-scale between the pre- and post-breast cancer scores at the 5% level of significance. The statistical package SAS version 9.1 (SAS Institute, Inc., Cary, NC) was used for fitting mixed-effects models (PROC MIXED) and calculating the non-centrality parameter and the related numbers of degrees of freedom for the power calculation. A P-value of <0.05 was selected to denote statistical significance.Qualitative dataAs was anticipated with the Kin KeeperSM model, women did share their breast cancer screening goals with their family. Examples from the focus group responses corroborate this element of the design. KF103: ‘My sisters and I discussed our goals and that [BSE] was one of our goals:… to check more often since the three of us had a bout with a lump in the breast, … check ourselves much more often than what we did in the past.’ Women in the study particularly appreciated instruction in how to properly perform a BSE. CHW 104: ‘Once we got past the myths of mammograms and all that type of stuff, the education played its part. They were amazed. A lot of them were surprised that they were doing their BSE the wrong way. So they enjoyed that.’ CHW 105: ‘All my families really enjoyed it. We had a ball with all the information and especially the hands-on stuff’.Although women readily included both BSE and CBE in their actions plans, their responses to open-ended questions about the frequency of these examinations suggested their uncertainty about the required frequency as well as the need for a revised form to make it easier to use. Most women (88%) indicated they would conduct a monthly BSE. This health behavior is not dependent on health care access. However, even with two home visits and a cancer literacy score of ≥75 for all participating women, 10% of women who cited conducting a BSE as their goal were confused about how often they should perform a BSE. For example, a few indicated ‘daily’ and two indicated ‘over a 100 times a month’. American Cancer Society guidelines call for 80 of the 161 women to have mammograms annually due to age (≥40 years). Even though 82 women had had a mammogram at some time, only 66 had mammograms in the past year and only 54 were aged ≥40 years, implying that 67.5% of women aged ≥40 years received needed mammograms. The issue of frequency is also raised by the fact that 10% of the 93 women who included mammography in their personal actions plan were confused about how often they should get a mammogram.DiscussionTo teach women about the salient aspects of breast cancer prevention and early detection, we implemented an innovative family-focused intervention, Kin KeeperSM, which builds on the generational and cultural behavior of African-American women [71]. We theorized that the existing ways in which African-American women communicate with other females in their family could be used to empower the family unit to engage in cancer prevention and screening behavior. Using the ability of CHWs to access vulnerable African-American women in their communities, we demonstrated that a dynamic recruitment protocol worked to reach other women in the family who were not connected to the public health program. The successful track record that CHWs have had in public health in building trust with women who have been marginalized and connecting them to health services [72, 73] is why they are an integral part of the Kin KeeperSM model. Women of the same family were willing to attend two meetings to discuss breast health and wellness and their individual and familial risks, as well as learn about screening guidelines. The family setting was a safe one for family discussions about personal goals and familial risks and a source of screening motivation. The low drop-out rates suggest that the Kin KeeperSM Cancer Intervention and the implementation process are feasible and capable of retaining participants over time.The breast cancer literacy assessment baseline scores, compared with post-intervention assessment scores, showed that women increased their functional understanding of breast cancer prevention and early detection. In their personal action plans, most women set attainable goals for themselves. Data from the focus group sessions suggested that women saw value in learning the proper techniques of BSE. The increase in scores across the board suggests that our process was indeed feasible and effective for different categories of women.Currently, the predominant current approaches for cancer education, prevention and screening have focused on using the influences of either the health care provider [26–30, 74] or community-based programs [6–13, 15, 19]. As mentioned, these approaches cast a narrow net, often leaving out women without connections to a community-based organization or missing women without connections to the health care system through insurance or regular visits. Low-income women who do not have the opportunity to learn about cancer risks, low-cost screening services or how to conduct a BSE may be left particularly vulnerable by traditional outreach approaches. By design, the Kin KeeperSM accounts for the diversity that exists within African-American families, such as income and access to health care information [71]. Working with two extremely different public health programs, we demonstrated the implementation of the life span approach and its appeal in a familial setting. As we look toward the expanded utility of Kin KeeperSM as a model for education and recruitment of women into clinical trials and genetic studies, family lineage will be important to identify for familial cancer registries.We observed that some CHWs were challenged by our kin keeper eligibility criteria, which required that parents and grandparents be African-American and that at least two family members who live in the defined community be between ages 35 and 64 years. As a result, data collection was extended by 3 months. Other than redefining the community to include the county, which is considered part of our public health programs’ service area, we retained the criteria. We deliberately focused on the African-American family rather than multiracial families that include African-Americans to gain more insight into the social, economic and heterogeneity among African-Americans that could facilitate development of more tailored screening services.CHWs and program staff alike also cited the design of the personal action plans as something that needed to be revamped. A routine protocol meeting with the CHWs and the research coordinator revealed that the formatting was confusing, leading some women to complete the section on ‘future goals’ and ‘accomplished goals’ in the same setting. In addition, some of the older women were so moved by the intervention that they set goals that clearly were not in line with any guidelines, such as performing a BSE ‘100 times in a month’.LimitationsCommunity-based research methods caution against generalizations or one-size-fits-all recommendations, which could be viewed as a limitation. Moreover, our study employs a single-arm pre/post-test without a control group; therefore, our results are somewhat limited because they are based on simple measurement. Nevertheless, we have employed multiple techniques and analyses to validate the feasibility of our process and found corroborating results. Despite the limitations and our challenges, our experience has demonstrated the public health utility of the Kin KeeperSM model and the feasibility of applying it in a real-world environment. In our continuing study, we will track breast cancer screening behaviors, along with the impact of time on the assessment scores. We will conduct a larger investigation to study the actual effects of the model measured statistically. In further analysis, we anticipate linking the personal action plan goals with the breast cancer literacy scores.FundingThe Susan G. Komen for the Cure (POP06000280).We would like thank all the study participants, the CHWs and supervisors Ms Christie Peck, Ms Yolanda Hill-Ashford and Ms Tedra L. Jackson for coordinating this project. We thank Mr Athur Mabiso for assisting in preparing this manuscript. Thank you to the Susan G. Komen for the Cure for their funding of this project.1.WilliamsKPKin Keeper: A Family-Focused Cancer Prevention Model for African-American Women2007vol. 15. Binghamton, NYHaworth Press;2913052.NeighborsHWThe help-seeking behavior of black Americans. A summary of findings from the National Survey of Black AmericansJ Natl Med Assoc19888091009123.GutierrezLLewisEEmpowering Women of ColorNew York, NY. Columbia University Press; 19994.JemmottLSJemmottJBIIIIncreasing condom-use intentions among sexually active black adolescent womenNurs Res199241527395.WilliamsKPSheppardVCapacity building: a strategy to help narrow the health disparity for African American womenCent Res Afr Am Women J2002349526.YoungRFWallerJBJrSmithermanHA breast cancer education and on-site screening intervention for unscreened African American womenJ Cancer Educ200217423167.CardinVAGrimesRMJiangZDLow-income minority women at risk for cervical cancer: a process to improve adherence to follow-up recommendationsPublic Health Rep20011166608168.EarpJAEngEO'MalleyMSIncreasing use of mammography among older, rural African American women: results from a community trialAm J Public Health2002924646549.KlassenACSmithALMeissnerHIIf we gave away mammograms, who would get them? A neighborhood evaluation of a no-cost breast cancer screening programPrev Med2002341132110.DanigelisNLAshleyJAWordenJKTwo community outreach strategies to increase breast cancer screening among low-income womenJ Cancer Educ200116155811.BaldwinDA model for describing low-income African American women's participation in breast and cervical cancer early detection and screeningANS Adv Nurs Sci1996192274212.EngEThe save our sisters project. A social network strategy for reaching rural black womenCancer1993723 Suppl.1071713.BrownsteinJNChealNAckermannSPBreast and cervical cancer screening in minority populations: a model for using lay health educatorsJ Cancer Educ199274321614.RobersonNLComparison of breast screening outcomes from a cancer control intervention for African-American and white women in western New YorkJ Assoc Acad Minor Phys199782293315.GreenePGSmithDEHullettSCancer prevention in rural primary care. An academic-practice partnershipAm J Prev Med1999163 Suppl.586216.PaskettEDMastenKBPhillipsKCReceptivity of a worksite breast cancer screening education programHealth Educ Res1999145667417.ErwinDOCancer education takes on a spiritual focus for the African American faith communityJ Cancer Educ200217146918.HoltCLClarkEMKreuterMWSpiritual health locus of control and breast cancer beliefs among urban African American womenHealth Psychol2003223294919.PaskettEDTatumCMD'AgostinoRJrCommunity-based interventions to improve breast and cervical cancer screening: results of the Forsyth County Cancer Screening (FoCaS) ProjectCancer Epidemiol Biomarkers Prev199985453920.LoweJIBargFKNormanSAn urban intergenerational program for cancer control educationJ Cancer Educ1997124233921.SiminoffLAZhangAColabianchiNFactors that predict the referral of breast cancer patients onto clinical trials by their surgeons and medical oncologistsJ Clin Oncol200018612031122.EllisPMButowPNTattersallMHRandomized clinical trials in oncology: understanding and attitudes predict willingness to participateJ Clin Oncol2001191535546123.FinnRSurveys identify barriers to participation in clinical trialsJ Natl Cancer Inst200092191556824.TaylorKMMargoleseRGSoskolneCLPhysicians’ reasons for not entering eligible patients in a randomized clinical trial of surgery for breast cancerN Engl J Med1984310211363725.TaylorKMKelnerMInterpreting physician participation in randomized clinical trials: the Physician Orientation ProfileJ Health Soc Behav198728438940026.O'MalleyMSEarpJAHawleySTThe association of race/ethnicity, socioeconomic status, and physician recommendation for mammography: who gets the message about breast cancer screening?Am J Public Health2001911495427.LongEBreast cancer in African-American women. Review of the literatureCancer Nurs199316112428.RutledgeDNBarsevickAKnobfMTBreast cancer detection: knowledge, attitudes, and behaviors of women from PennsylvaniaOncol Nurs Forum200128610324029.YabroffKRMandelblattJSInterventions targeted toward patients to increase mammography useCancer Epidemiol Biomarkers Prev1999897495730.BirdJAMcPheeSJJenkinsCThree strategies to promote cancer screening. How feasible is wide-scale implementation?Med Care1990281110051231.Hoffman-GoetzLMillsSLCultural barriers to cancer screening among African American women: a critical review of the qualitative literatureWomen's Health: Research on Gender, Behavior and Policy1997318320132.PhillipsJMBreast cancer and African American women: moving beyond fear, fatalism, and silenceOncol Nurs Forum19992661001733.PoweBDJohnsonAFatalism as a barrier to cancer screening among African-Americans: Philosophical perspectives1995vol. 34New York, NYSpringer11912634.GuidryJJFaganPWalkerVCultural sensitivity and readability of breast and prostate printed cancer education materials targeting African AmericansJ Natl Med Assoc1998903165935.Institute of MedicineHealth Literacy: A Prescription to End Confusion2004Washington, DCNational Academic Press36.SkellyAHSamuel-HodgeCElasyTDevelopment and testing of culturally sensitive instruments for African American women with type 2 diabetesDiabetes Educ200026576974776–737.FriedellGHRubioAWagnerWDWhat providers should know about community cancer controlCancer Pract1997563677438.GuidryJJMatthews-JuarezPCopelandVABarriers to breast cancer control for African-American women: the interdependence of culture and psychosocial issuesCancer2003971 Suppl.3182339.StewartPHallCJBowieSLWho Is Kin? Family Definition and African American FamiliesAfrican American Behavior in the Social Environment: New Perpectives2007vol. 15Binghamton, NYHaworth Press16318140.Williams-BrownSBaldwinDMBakosAStorytelling as a method to teach African American women breast health informationJ Cancer Educ200217422723041.Banks-WallaceJATalk That Talk: Storytelling and Analysis Rooted in African American Oral TraditionQualitative Health Research200212341042642.KillienMBigbyJAChampionVInvolving minority and underrepresented women in clinical trials: the National Centers of Excellence in Women's HealthJ Womens Health Gend Based Med200091010617043.SwansonGMWardAmyJRecruiting minorities into clinical trials: toward a participant-friendly systemJ Natl Cancer Inst1995872317475944.McCarthyCRHistorical background of clinical trials involving women and minoritiesAcad Med1994699695845.JacksonFMConsiderations for community-based research with African American womenAm J Public Health2002924561446.BronfenbrennerUEcology of the family as a context for human development: research perspectivesDev Psychol1986227234247.ButterfossFDMorrowALRosenthalJCINCH: an urban coalition for empowerment and action. Consortium for the Immunization of Norfolk's ChildrenHealth Educ Behav19982522122548.Sanders-PhillipsKCorrelates of health promotion behaviors in low-income Black women and LatinasAm J Prev Med1996126450849.WilliamsKPThe Intersection of Family, Community and Provider in Breast Cancer Prevention: A Conceptual FrameworkBlack Families4th ednThousand Oaks, CASage Publications200750.WilliamsMVBakerDHonigEInadequate literacy is a barrier to asthma knowledge and selfcareAm Coll Chest Phys199811410081551.WitmerASeiferSDFinocchioLCommunity health workers: integral members of the health care work forceAm J Public Health1995858 Pt 11055852.ZhuKHBernardLJPayne-WilksKRecruiting elderly African-American women in cancer prevention and control studies: a multifaceted approach and its effectivenessJ Natl Med Assoc20009241697553.EarpJAViadroCIVincusAALay health advisors: a strategy for getting the word out about breast cancerHealth Educ Behav19972444325154.TessaroIEngESmithJBreast cancer screening in older African-American women: qualitative research findingsAm J Health Promot1994842869255.RomanLALindsayJKMooreJSCommunity health workers: examining the helper therapy principlePublic Health Nurs1999162879556.WilkinsonDBraithwaiteRTaylorSEIndigenous community health workers in the 1960′s and beyondHealth Issues in the Black Community1992San Francisco, CAJossey Bass Publishers2556657.WilliamsGAAbbottRRTaylorDKUsing focus group methodology to develop breast cancer screening programs that recruit African American womenJ Commun Health1997221455658.WilliamsKPSheppardVBHinesRDTrust's Impact on Recruiting for Chemoprevention Trials: Issues of trust's in the recruitment of African American women into breast cancer chemoprevention trialsInt J Cancer Prev20041213714359.Millon-UnderwoodSSandersEDavisMDeterminants of participation in state-of-the-art cancer prevention, early detection/screening, and treatment trials among African-AmericansCancer Nurs1993161253360.FreemanHPMuthBJKernerJFExpanding access to cancer screening and clinical follow-up among the medically underservedCancer Pract199531193061.VernonSWMeissnerHIMillerSMThe role of behavioral science in cancer prevention research: planning the next steps in the collaborative processAm Assoc Cancer Res200615413562.PasickRJHiattRAStewartSCommunity-Based Cancer Screening for Underserved Women: Design and Baseline Findings from the Breast and Cervical Cancer Intervention Study2001 vol. 33. Academic Press19020363.American Cancer SocietyCancer Facts & Figures for African Americans 2005–20062005AtlantaAmerican Cancer Society64.American Cancer SocietyCancer Facts & Figures for African Americans 2007–20082007AtlantaAmerican Cancer Society65.WallersteinNDuranBMinklerMWallersteinNThe conceptual, historical, and practice roots of community based participatory research and related participatory traditionsCommunity-Based Participatory Research for Health2003San Francisco, CAJossey-Bass275266.SchulzAJIsraelBParkerEALockettMLHillYRRochelleWMinklerMWallersteinNEngaging women in community based participatory research for healthCommunity-Based Participatory Research for Health2003San Francisco, CAJossey-Bass29331567.WilliamsKPReckaseMRivera-VasquezOToward the development of cancer literacy assessment toolsMich J Public Health200821213168.DenzinNLincolnYCollecting and Interpreting Qualitative Materials20032nd ednThousand Oaks, CASage Publications69.MolenberghsGVerbekeGLinear Mixed Models for Longitudinal Data2000New YorkSpringer-Verlag70.JanisILasswellHDLeitesNCThe problem of validating content analysisLanguage of Politics: Studies in Qualitative Semantics1965Cambridge, MAMIT Press558571.WilliamsKPMcAdooHKin KeepersSM Breast Cancer Prevention for African American WomenBlack Families20074th ednThousand Oaks, CASage Publications72.RoMJTreadwellHMNorthridgeMCommunity Health Workers and Community Voices: Promoting Good Health Atlanta, GA. National Center for Primary Care Morehouse School of Medicine; October 200373.SheppardVBWilliamsKPRichardsonJTWomen's priorities for lay health home visitors: implications for eliminating health disparities among underserved womenJ Health Soc Policy2005183193574.MandelblattJSYabroffKREffectiveness of interventions designed to increase mammography use: a meta-analysis of provider-targeted strategiesCancer Epidemiol Biomarkers Prev19998975967