Promoting cancer screening among churchgoing Latinas: Fe en Accion/faith in action

Abstract

Cancer screening rates among Latinas are generally low, reducing the likelihood of early cancer detection in this population. This article examines the effects of a community intervention (Fe en Accion/Faith in Action) led by community health workers (promotoras) on promoting breast, cervical and colorectal cancer screening among churchgoing Latinas. Sixteen churches were randomly assigned to a cancer screening or a physical activity intervention. We examined cancer knowledge, barriers to screening and self-reported mammography, clinical breast exam, Pap test, fecal occult blood test and sigmoidoscopy or colonoscopy at baseline and 12 months follow-up. Participants were 436 adult Latinas, with 16 promotoras conducting a cancer screening intervention at 8 out of 16 churches. The cancer screening intervention had a significant positive impact on self-reported mammography (OR = 4.64, 95% CI: 2.00–10.75) and breast exams in the last year (OR= 2.82, 95% CI: 1.41–5.57) and corresponding reductions in perceived (87.6%) barriers to breast cancer screening (P < .008). Cervical and colorectal cancer screening did not improve with the intervention. These findings suggest Fe en Accion church-based promotoras had a significant impact on promoting breast cancer screening among Latinas. Colon cancer screening promotion, however, remains a challenge.

Introduction

Although Latina breast cancer incidence and mortality rates compare favorably to non-Hispanic Whites and African Americans [1], Latinas are twice as likely to be diagnosed with and 1.5 times more likely to die from cervical cancer compared with non-Hispanic white women [2]. Survey estimates suggest that only 55.3% of Latinas in the US had a fecal occult blood test (FOBT) in the past year, and/or a flexible sigmoidoscopy in the past 5 years, and/or a colonoscopy in the past 10 years [3]. Mexican-American and Mexican immigrant women may be the least likely of any Latina subgroup to receive clinical breast exams, mammograms, or Pap test screenings [4].

Understanding and addressing barriers to cancer screening is key to increasing screening participation rates. Common barriers to screening among Latinas include lack of access to health care, lack of clinician recommendation, fear of cancer, fatalism and embarrassment stemming from language and cultural differences [4–6]. Latinas may also feel less susceptible to cancer, especially if they do not have current symptoms or a positive family history [7]. One culturally sensitive approach to addressing these barriers is through the involvement of promotoras or community health workers. Many promotora-led interventions aim to address health conditions that impact underserved communities like Latinos by improving access to health information and resources. Promotoras often speak the language of the target community and are more likely to connect with community members by providing emotional and instrumental support.

There is evidence that cancer screening programs that involve promotoras as the key agents are feasible and effective for increasing breast and cervical screening among Mexican-American women [8, 9]. Larkey and colleagues randomly assigned women who were 50 years or older to an individually or group delivered promotora-led intervention promoting cancer screening [10] Their findings suggest increases in self report Pap smear use in both groups, but the authors note that the group delivered intervention was significantly more cost effective than the individually delivered group. A recent faith-based intervention pilot study found that the cancer screening education program was feasible and acceptable among churchgoing Latinas [11]. Taken together, these findings suggest the need to continue to expand promotora-led interventions and research their effectiveness.

Churches provide excellent settings in which to promote healthy behaviors, as they are established institutions that mobilize the growing Latino community [12–14]. According to a Pew Research Center report, 55% of Latinos identified as Catholic and almost 40% of Latino Catholics report attending church at least once a week [15]. Churches offer space for meetings and PA programs, as well as volunteers eager to be trained as Lay Health Advisors or promotoras. This study evaluated self-reported cancer screening of a two-group randomized trial promoting breast, cervical and colorectal cancer screening among churchgoing Latinas through promotoras.

Materials and methods

Overview of study design and research aims

This study uses data from Fe en Accion (Faith in Action), a randomized controlled trial that assigned 16 Catholic churches frequented by Latinos to either a cancer screening or physical activity intervention from 2011 to 2014. Details of the trial design are detailed elsewhere [16]. Briefly, trained bilingual/bicultural (Spanish-/English-language, Mexican-origin) promotoras implemented intervention activities targeting multiple levels of the ecological model, including the individual- (e.g. knowledge-attitude-behavior), social- (e.g. interpersonal support and role modeling), organizational- (e.g. church space allocated for program activities, announcements at mass) and environmental- (e.g. access to health care) levels. Each participating church committed to program activities for 24 months and agreed to be randomly assigned to either condition. Baseline, 12 month and 24 month post baseline assessments were collected among 436 participants. All baseline measures were conducted prior to the start of intervention activities, with both study participants and measurement research assistants blinded to condition. Although Fe en Accion’s primary aims were to target physical activity promotion, the design provided the opportunity to concurrently examine cancer screening promotion in the attention-control group. The study was based at San Diego State University’s Institute for Behavioral and Community Health (www.ibachsd.org), a leader in Latino community health research for 3 decades [17]. All study protocols were approved by the Institutional Review Board of San Diego State University. The current trial is registered under NCT01776632.

Recruitment

Recruitment of Churches and Participants

The study recruited 16 Catholic churches in San Diego County and 436 Latinas who attended these churches. Details describing recruitment of churches and participants can be found elsewhere [16]. All recruitment efforts were conducted in Spanish or English. The following inclusion criteria were developed for the physical activity condition (primary outcome for the trial) and were used for the cancer screening condition as well: self-identified Latina between the ages of 18 and 65 years, attended the church at least 4 times a month for any reason, lived within 15-minute driving distance of the church with access to reliable transportation, identified no barriers to attend program activities, planned on attending the church for the next 24 months and had low levels of physical activity (based on self-report and accelerometer). Other church and community members were welcome to participate in intervention activities as all activities were offered free at the church or local parks and community centers. Potential participants were excluded if they had a health condition that limited their ability to be physically active (e.g. pregnancy or uncontrolled high blood pressure).

Selection and Training of Promotoras

The Intervention Coordinator approached priests and church leaders to identify eligible candidates, who were interviewed and assessed for their ability to lead groups, commitment to the 2-year project, knowledge of the church and community and leadership skills. The study hired 2 promotoras at each church, based on their experience promoting health, bilingual fluency (English/Spanish) and practice of preventive behaviors (regular doctor’s visits, physical activity, healthy diet, etc.). Selected promotoras worked 5–10 paid hours per week.

Promotoras received about 24 hours of training related to cancer screening promotion over 6 weeks (detailed elsewhere) [16]. Promotoras were assigned homework to enhance their learning, increase their knowledge of cancer screening services in their communities, review current cancer screening recommendations and develop skills as health educators.

Intervention design

Cancer Screening Intervention Components

The cancer screening intervention was based on recommendations from the National Cancer Institute (NCI) [18] and the American Cancer Society (ACS) [19]. Advisory committee members from the Moores Cancer Center at UC San Diego, the local ACS, and church leaders reviewed materials and provided feedback. The study provided NCI and ACS handouts to participants in the cancer screening condition. Screening recommendations used for the intervention were based on 2011 ACS guidelines for breast, cervical and colorectal cancer (see Supplementary Appendix). The intervention components were separated into different levels of the ecological model including individual, interpersonal and organizational/community levels.

Individual Level

Promotoras led 6-week series of classes that provided information about cancer screening recommendations and risk factors for each of the target cancers. Series were offered on a continual basis throughout the year, at minimum six times during the year. At the conclusion of each series, promotoras recruited participants for the next series. The topics for each 90–120 minute class were as follows: (i) the importance of prevention; (ii) breast cancer screening, risk factors and treatment; (iii) cervical cancer risk factors, the human papilloma virus and screening and treatment; (iv) colorectal cancer screening, risk factors and treatment; (v) skin cancer risk, prevention, treatment and screening; and (vi) a discussion of rights and responsibilities as patients (see Supplementary Appendix). Each class began with a prayer, proceeding to an overview of objectives, icebreaker activity, didactic lesson, interactive activity, question and answer and closing prayer. The four classes devoted to specific cancers reviewed the screening recommendations based on ACS and NCI guidelines, as well as risk factors for cancer, prevention tips, myths and perceived barriers to screening, and a review of anatomy of the selected area. We developed interactive activities to reinforce learning objectives. For example, the colorectal cancer session included a ‘loteria’ or ‘bingo’ style game with glues related to colorectal cancer screening. Promotoras maintained a record of participants’ attendance and contacted those who had not completed the classes, inviting them to attend.

Interpersonal Level

Promotoras conducted up to two motivational interviewing (MI) calls [20] over the course of the 1-year intervention. For each call, promotoras were given one month to complete the MI call with all participants at their assigned church. Promotoras had 8 attempts to contact each participant for the call. They used a modified version of a previously developed guide to evaluate barriers to cancer screening, fears about cancer, values related to health and spirituality, culminating in solutions to overcome barriers to screening [21, 22]. Participants reported on their most recent cancer screenings and made goals for age-appropriate screening appointments. Promotoras accompanied participants to cancer screening appointments as needed, providing social support and comfort. At the end of each call, promotoras helped participants establish goals related to cancer screening, including scheduling appointments, learning which screening tests are covered by insurance, and other topics.

Organizational and Environmental Level

Cancer screening classes were advertised in the church bulletin and promotoras made announcements during mass. In addition, churches allocated space for cancer screening classes. During training, promotoras gathered information about local clinics and the services they provided, which they shared with the participants. They also completed Affordable Care Act workshops.

Measures

Demographic measures included age, household income, education, marital status, number of children in the home, employment status, occupation, country of birth, years living in the US and health insurance status at 12-month follow-up. Other survey measures included health conditions, religiosity, depressive symptoms, perceived stress, sleep, cancer knowledge and screening behaviors and perceived barriers to screening. The Bidimensional Acculturation Scale (BAS) for Hispanics assessed language use, linguistic proficiency and electronic media use [23]. Acculturation scores were estimated for the 2 cultural domains assessed by the scale (Hispanic and non-Hispanic) with scores < 2.5 on the non-Hispanic domain being indicative of low levels of acculturation. The 2010 BRFSS survey assessed cancer screening behaviors for breast, cervical and colorectal cancers [24]. If the participant responded ‘YES’ to ever completing a mammogram, clinical breast exam, Pap test, fecal occult blood stool test and colonoscopy or sigmoidoscopy, they were asked how long ago the last screening was completed. Response options included 5 or more years ago, within the past 5 years (but more than 3 years ago), within the past 3 years (but more than 2 years ago), within the past 2 years (but more than 1 year ago) and within the past year (anytime less than 12 months ago). Because sigmoidoscopy or colonoscopy screening was assessed in the same item and each exam has its own recommended screening interval (i.e. every 5 years for sigmoidoscopy and every 10 years for colonoscopy), we only assessed whether or not the participant ‘ever’ had either test done.

We used the Esperanza y Vida cancer knowledge questionnaire [25] to assess participants’ knowledge of breast and cervical cancer (15 items). Three true/false items from the 2005 Health Information National Trends Survey (HINTS 2) [26] were used to assess colorectal cancer knowledge. Items included ‘People with colon cancer would have pain or other symptoms prior to being diagnosed,’ ‘There’s not much you can do to lower your chances of getting colon cancer,’ and ‘There are ways to slow down or disrupt the development of colon cancer.’ ‘I don’t know’ responses were treated as missing. Knowledge scores were based on the percent of correct responses for breast cancer (6 items), cervical cancer (9 items) and colorectal cancer (3 items).

Perceived barriers to screening for breast, cervical and colorectal cancers were assessed with items obtained from the 1990 Tampa survey on beliefs about breast cancer screening (9 items from the 29-item survey), Perceived Barriers to Getting Screened for Pap test screening (all 10 items), Perceived Benefits of CRC Screening and Barriers to Undergoing CRC Screening (1 item from the 27-item survey), and the Colon Cancer Knowledge and Attitudes scale (6 items from the 29-item survey) [27–30]. Respondents reported on their level of agreement to a series of statements related to potential barriers to getting screened (e.g. ‘it is too expensive to have a Pap test’). Responses were on a 5-point Likert scale, where 1 = ‘strongly disagree’ and 5 = ‘strongly agree’. Response options for positive statements were reverse coded (e.g. ‘would change your diet to do FOBT’). Mean scores were computed for each cancer with higher scores being indicative of greater perceived barriers to screening.

Analyses

Descriptives

Means and frequencies were obtained for select baseline demographic variables, health insurance status (at 12-month follow-up), screening status and scores for perceived barriers to screening and cancer knowledge. Bivariate mixed or generalized mixed effects models, adjusted for clustering effects of the churches, were used to compare the conditions on each variable.

Intervention effects

Screening

For each screening outcome, generalized linear mixed models were fitted with binomial error distributions (logistic models) to obtain odds ratios (OR) and 95% confidence intervals (CI), adjusting for clustering effects of the churches. For screenings recommended annually (mammogram and FOBT), models were also adjusted for baseline screening status, health insurance status at 12-month follow-up and baseline scores for perceived barriers to screening and cancer knowledge for the relevant cancer. Models were omitted for Pap test and sigmoidoscopy/colonoscopy screening, which have longer screening intervals, due to reduced statistical power from small sample sizes (at baseline, only 53 women reported not having had a Pap test and 79 reported not having had a colonoscopy/sigmoidoscopy).

Analyses were performed among those who met each tests’ recommended age for screening at baseline. For example, women aged 40–65 years at baseline comprised the analytical sample for mammograms; those who turned 40 at 12-month follow-up were not included in the analysis for mammograms. For CRC screening, we were limited in our ability to assess whether participants met at least one of the screening recommendations (i.e. FOBT, colonoscopy, or sigmoidoscopy) in large part because we assessed colonoscopy and sigmoidoscopy screening in the same item. Thus, we decided to include all participants aged 50–65 years in separate models for FOBT and colonoscopy/sigmoidoscopy screening.

In addition, because the recommended screening interval for clinical breast exams is different for women aged 21–39 years compared with those aged 40 years and older, separate models were performed for the subset of women aged 21–39 years that were due for clinical breast exams at baseline (n = 22) (since adherent women were not necessarily due for breast exams at 12-month follow-up) and the subset of women aged 40–65 years regardless of their adherence status to clinical breast exams at baseline. The model for clinical breast exams for women aged 21–39 years was adjusted for health insurance status at 12-month follow-up and baseline scores for perceived barriers to breast cancer screening and cancer knowledge, while the model for those aged 40–65 also included baseline screening status for breast exams.

Perceived Barriers to Screening

Mixed effects models were used to assess intervention effects on perceived barriers to screening, adjusting for baseline perceived barriers to screening score (for the relevant cancer), age, marital status and clustering effects of the churches.

Cancer Knowledge

Mixed effects models were fitted for breast and cervical cancer knowledge scores at 12-month follow-up among all respondents (i.e. no age criteria). To account for the skewed distribution of the colorectal cancer knowledge scores, we created a binary variable that categorized participant knowledge as ‘at least 1 correct response (out of 3)’ or ‘none correct.’ We used a generalized mixed effects model with a binomial error distribution (logistic model) to obtain the OR and 95% CI of getting at least one correct response, adjusting for clustering effects of the churches. Models were adjusted for baseline knowledge score (for the relevant cancer), age, education and clustering effects of the churches.

Motivational Interviewing Effects

We obtained frequencies for motivational interviewing (MI) calls for participants in the cancer screening condition (n = 219). The total number of MI calls received was used as a continuous variable (range: 0–2). Due to the small sample sizes of women in the cancer screening condition that were due for a Pap test (n = 22) or sigmoidoscopy/colonoscopy (n = 37) at baseline, we omitted analyses of MI calls on these screening outcomes.

Statistical significance was defined as P < .05. All statistical analyses were performed in SAS version 9.4 (SAS Institute Inc., Cary, North Carolina).

Results

Table I presents baseline demographics, cancer screening rates, perceived barriers to screening and cancer knowledge scores among women in the cancer screening (n = 219) and physical activity (n = 217) conditions. Over two thirds of the overall sample (N = 436) were 40–65 years of age. The majority of women lived in households with total incomes of less than $2000 per month and had less than a high school education, indicating a generally low socioeconomic status. Slightly more than three fourths were married or living as married. Approximately 91% of participants were born in Mexico and nearly half had been living in the United States for fewer than 20 years. This was reflected in the relatively low acculturation level to the Anglo/US culture. More than 80% of the women were either overweight or obese. In addition, less than half of the sample reported having any form of health insurance (e.g. private insurance or Medicaid/MediCal). The cancer screening and physical activity conditions did not differ significantly on these demographic variables.

Table I.

Baseline demographics, cancer screening rates and barriers to screening and cancer knowledge scores among churchgoing Latinas, stratified by study condition. Fe en Accion, San Diego, CA, 2011–2014

			Condition
	Total (N = 436)		Cancer screening (n = 219)		Physical activity (n = 217)		P values d
Variable	%		%		%
Demographics
Age (y)
18–39	31.9		33.8		30.0
40–65	68.1		66.2		70.1		0.51
Household income < $2000/month a	58.3		60.5		56.1		0.37
< High school completed	54.8		56.0		53.7		0.87
Married or living as married	77.3		76.5		78.1		0.71
Employed	65.8		63.0		68.7		0.21
Born in Mexico	90.8		92.6		88.9		0.29
< 20 years living in the US	46.6		49.8		43.3		0.28
Low acculturation levels b	67.8		70.7		64.9		0.46
Overweight/obese	83.2		82.6		83.8		0.73
Has health insurance c	48.0		42.7		53.3		0.19
	ne	%	ne	%	ne	%	P values d
Had screening test (recommended age)
Breast cancer
Mammogram, in last year (40–65 y)	295	47.8	144	43.8	151	51.7	0.37
Clinical breast exam, in last 3 years (21–39 y) or in last year (40–65 y)	427	52.2	215	47.0	212	57.6	0.11
Cervical cancer
Pap test, in last 3 years (21–64 y) or ever (65 y)	427	87.6	216	89.8	211	85.3	0.18
Colorectal cancer (50–65 y)
Blood stool test, in last year	120	14.2	59	15.3	61	13.1	0.74
Sigmoidoscopy or colonoscopy, ever	120	34.2	59	37.3	61	31.2	0.48
	mean (SD)		mean (SD)		mean (SD)		P values d
Perceived barriers to screeningf
Breast cancer (1–5 point rating)	1.9 (0.8)		2.0 (0.8)		1.9 (0.8)		0.32
Cervical cancer (1–5 point rating)	2.2 (0.7)		2.1 (0.7)		2.2 (0.7)		0.51
Colorectal cancer (1–5 point rating)	2.5 (0.7)		2.5 (0.7)		2.5 (0.6)		0.79
Cancer knowledge (% correct)
Breast cancer (out of 6 items)	48.9 (22.4)		48.9 (22.1)		48.9 (22.8)		0.95
Cervical cancer (out of 9 items)	44.8 (18.3)		46.2 (18.7)		43.3 (17.9)		0.35
Colorectal cancer (out of 3 items)	29.4 (31.1)		28.6 (30.7)		30.3 (31.6)		0.83

^aMissing n = 31 (cancer screening = 19; physical activity= 12).

^bDefined as scores ˂ 2.5 on the non-Hispanic domain. Missing n = 17 (cancer screening = 11; physical activity= 6).

^cHealth insurance status at 12-month follow-up. Missing n = 69 (cancer screening = 34; physical activity = 35).

^dAdjusted for clustering effects of churches.

^eSample size used in the denominator.

^fRespondents reported their level of agreement with a series of potential barriers to screening statements, with 1 = strongly disagree; 5 = strongly agree.

With respect to breast cancer screening, fewer than half of women aged 40–65 years reported having had a mammography in the last year and just over half of those aged 21–65 years were up to date on recommended clinical breast exams [18] (Table I). For cervical cancer screening, nearly 90% of women reported being up to date with their Pap test. Regarding colorectal cancer screening, only 14% reported having had a blood stool test in the last year and about a third reported ever having had a colonoscopy or sigmoidoscopy. Screening behaviors at baseline did not differ significantly between the study conditions.

Regarding perceived barriers to screening, the mean score was lowest for breast cancer screening (mean = 1.9) and highest for colorectal cancer screening (mean = 2.5) (Table I). Mean scores for cancer knowledge were lowest for colorectal cancer (29%) and highest for breast cancer (49%). None of the scores were significantly different between study conditions at baseline.

Figure 1 shows screening rates at baseline and 12-month follow-up. The cancer screening group improved breast cancer screening substantially: the proportion of women meeting guidelines increased from 44% to 61% for mammograms, and from 47% to 63% for breast exams. In contrast, breast cancer screening (both mammograms and breast exams) worsened over 12 months in the physical activity group. Cervical cancer screening rates were high at baseline for both groups (90% for the cancer screening group; 85% for the physical activity group); rates remained high at 12-month follow-up, increasing slightly in the physical activity group. Colorectal cancer screening rates increased in both conditions for FOBTs (from 15% to 25% in the cancer screening condition and from 13% to 20% in the physical activity condition), and colonoscopy/sigmoidoscopy screening (from 37% to 53% in the cancer screening condition and from 31% to 40% in the physical activity condition).

Fig. 1.

Rates of cancer screening for specific tests among recommended age groups of churchgoing Latinas at baseline and 12-month follow-up, stratified by study condition. Fe en Accion, San Diego, CA, 2011–2014.

Mixed effects analyses showed that at 12-month follow-up, mammogram and breast exam rates (among women aged 40–65 years) were significantly higher in the cancer screening condition compared with the physical activity condition, even after adjusting for potential confounders such as baseline screening status and health insurance (Table II). In addition, scores for perceived barriers to breast cancer screening were significantly lower in the cancer screening group compared with the physical activity group (difference in adjusted means = −0.32, P = 0.008). No significant effects were found from the Motivational Interviewing calls from the promotoras on the participant cancer screening outcomes.

Table II.

Mixed effects models to evaluate cancer screening outcomes among churchgoing Latinas in a cancer screening intervention compared with those in the physical activity condition. Fe en Accion, San Diego, CA, 2011–2014

Outcomes	nf	OR	95% CI	P values
Had screening test (recommended age)
Breast cancer
Mammogram, in last year (40–65 y) a	195	4.64	2.00–10.75	0.0004
Clinical breast exam b
In last 3 years (21–39 y)	33	4.68	0.52–42.26	0.16
In last year (40–65 y)	195	2.81	1.41–5.57	0.003
Colorectal cancer
Blood stool test, in last year (50–65 y) a	86	1.14	0.26–4.92	0.86

Outcomes	nf	CS adj. mean	SE	PA adj. mean	SE	Diff in adjusted means (CS-PA)	P values
Perceived barriers to screening
Breast cancer c	281	1.91	0.09	2.24	0.09	−0.32	0.008
Cervical cancer c	274	2.12	0.07	2.31	0.08	−0.18	0.07
Colorectal cancer c	285	2.92	0.06	3.02	0.06	−0.10	0.20
Cancer knowledge
Breast cancer d	287	54.67	3.29	47.74	3.29	6.94	0.14
Cervical cancer d	286	47.66	2.09	43.65	2.09	4.01	0.18
	nf		OR		95% CI		P values
Colorectal cancer e	288		1. 23		0.75–2.03		0.41

CS, Cancer screening; PA, Physical activity.

^aModels were adjusted for baseline screening status, health insurance status at 12-month follow-up and baseline perceived barriers to screening and cancer knowledge scores.

^bModel for the 21–39 age group was adjusted for health insurance status at 12-month follow up and baseline perceived barriers to screening and cancer knowledge scores. Model for the 40–65 age group was adjusted for the same covariates plus baseline breast exam screening status.

^cModels were adjusted for baseline perceived barriers score, age and marital status.

^dModels were adjusted for baseline knowledge score, age and education.

^eModel was adjusted for baseline knowledge score, age and education.

^fObservations used in the model.

Discussion

This study showed the effectiveness of a promotora-led cancer screening promotion intervention for Spanish-speaking Latinas in urban San Diego County. This intervention was conducted in churches, which facilitated access to physical resources and social networks of these Roman Catholic facilities and congregations. The intervention was particularly effective at promoting breast cancer screening, with anecdotal evidence indicating that several participants discovered lumps that they had not known about before the intervention. Self-reported cervical cancer screening did not change substantially from baseline to posttest nor were there substantial differences between the groups. However, baseline rates approached 90%, making it difficult to improve over these levels. This rate was substantially higher than the national cervical screening rate for Latinas (73.6% in 2010) [31] or the rate reported for Mexican immigrant women living in California in 2001 for 10 or more years (81%) [32]. Conversely, colorectal cancer screening rates remained low. Although changes in the intervention group were in the positive direction, there were no significant differences between the cancer screening group and the physical activity group.

This study’s limitations included reliance on self-reported screening rates, as confirmation of actual screening behavior was beyond the scope of our research. In this population, self-report validity may be attenuated by comprehension or a social desirability bias [33]. Second, our sigmoidoscopy/colonoscopy data are not directly comparable to those of other surveys and studies as the wording of our items was different. Because we combined the sigmoidoscopy and colonoscopy tests in the same item, we could not determine compliance to each test separately, and were unable to come up with an overall evaluation of compliance to colorectal cancer screening (i.e. compliance with colonoscopy, sigmoidoscopy, or FOBT). Moreover, the small sample size in some of the comparison group and wide confidence intervals likely limited our ability to detect changes between groups. Finally, our data are derived from a largely Mexican immigrant sample of women living within 30 miles of the California/Mexico border and may not be generalizable to Latinas from other backgrounds or living in other regions.

Among the several strengths of the study was that it targeted multiple cancer sites, as contrasted to studies that target a single site or type of cancer (e.g. breast cancer). Second, the intervention was a part of a randomized community trial conducted in churches. Faith-based organizations are an important and promising setting for promoting preventive behaviors among Latinas [13, 34]. Given the central role of churches within the Latino community and their overall commitment to the well-being of their members, places of worship setting (i.e. individual Roman Catholic churches) comprised the ideal setting for the current study. Churches are ideal settings for implementing health promotion programs as they are organizations that have space and facilities to implement these programs while providing culturally appropriate comfortable surroundings for participants, especially when led by promotoras or other fellow members of the congregation. Future health behavior change efforts should continue to collaborate with places of worship of all denominations to accomplish public health goals.

One of this study’s most notable strengths relates to the design of randomized community trials. As in our previous trials conducted at SDSU IBACH [35, 36], the Fe en Accion study employed a study design in which both arms of the trial received an active and important intervention. The main objective of Fe en Accion was to investigate the promotion of physical activity among Latinas. However, instead of a no-treatment or waitlist control, the comparison group consisted of the cancer screening intervention presented in this paper. Both groups could thus benefit substantially from their participation rather than simply providing data in a control group. This ‘win-win’ scenario benefitted not only the community, but cancer prevention science as well, contributing to the broader mission of NIH/NCI and public health in general.

In summary, the composite results have several implications. First, it is possible to ‘move the needle’ to increase the number of Spanish-speaking Latina immigrants participating in breast cancer screening through a promotora-led face-to-face intervention. Additional randomized community trials elaborating on the strengths and limitations of promotora-led interventions in this and other areas are warranted [8, 37]. While it is important to continue to promote cervical cancer screening, this data indicate that levels of cervical cancer screening are already approaching the Healthy People 2020 goal of 100%, at least in this population [38]. Our sample reported low rates of health insurance, which is an important factor for promoting cancer screening. Any attempts to promote cancer screening should also be paired with improved access to screening and healthcare. Finally, much work is yet to be done in terms of promotion of colorectal cancer screening via blood stool tests, colonoscopies and sigmoidoscopies in this population.

Supplementary data

Supplementary data are available at HEAL online.

Funding

This work was supported by the National Cancer Institute of the National Institutes of Health [R01CA138894 and F31CA206334-01].

Conflict of interest statement

None declared.