HPV vaccine uptake among daughters of Latinx immigrant mothers: findings from a cluster randomized controlled trial of a community-based, culturally relevant intervention

Abstract

Objective

We examined the efficacy of a culturally relevant, community-based HPV vaccination intervention among Latinx immigrant mothers with daughters aged 9–12 in Alabama.

Methods

We conducted a cluster-randomized controlled trial with “place of residence” (e.g., apartment complexes, trailer parks) as the unit of randomization that evaluated two interventions: 1) promotion of HPV vaccination and 2) promotion of healthy eating and appropriate nutrition label interpretation. Identical baseline/post/7-month follow up questionnaires were completed by all participants and both interventions consisted of four group sessions and one individual session. A total of 40 locations were randomized with 317 mother-daughter dyads enrolled in the study between May 2013 and October 2017.

Results

A total of 278 mother-daughter dyads met full eligibility and initiated the intervention/control participation. Retention rate overall was 93.2% (92.6% for the intervention arm and 93.7% for the control arm). Daughters in the intervention arm were significantly more likely to receive one, two, and three doses of HPV vaccine than daughters in the control arm p<0.001). In multivariate analyses, mothers in the intervention arm had a six times greater odds of vaccinating daughters with the first dose (OR=5.96, 95% CI: 3.38, 10.49), eight times greater odds of vaccinating daughters with the second dose (OR=8.09, 95% CI: 4.0, 16.35), and more than 16 times greater odds of completing the three-dose HPV vaccine series than mothers in the control arm after adjusting for mother’s age, time in the U.S., income, and daughter’s health insurance status (OR=16.5, 95% CI: 5.73, 47.48). Only perceived risk of their daughters’ future HPV infection remained significant as a predictor of three-dose HPV vaccination completion (OR=0.69, 95% CI: 0.23, 2.1).

Conclusions

A theory-driven, culturally-relevant intervention developed through extensive formative assessments in collaboration with community members can effectively promote HPV vaccination among 9–12 years of age daughters of Latina immigrants.

Clinical Trials Registration number

NCT04154319

1. Introduction

Human Papillomavirus (HPV) infection, the most common sexually transmitted infection in the U.S., affects ~79 million Americans at any given time, with 14 million new cases diagnosed each year [1]. Persistent infections with high-risk HPV strains is causally related to cervical, vaginal, vulvar, anal, penile, and oropharyngeal cancer [2]. Recognizing that virtually all cervical cancer cases are caused by carcinogenic HPV infections, two new approaches have emerged in prevention efforts: (1) HPV vaccination for primary HPV prevention; and (2) carcinogenic HPV detection for secondary prevention. Therefore, prevention efforts must be age-appropriate to maximize public health benefits, and primary prevention through HPV vaccination is a significant step in eliminating cervical cancer as a public health problem [3].

Since its approval in December 2014, the 9-valent vaccine has been the most widely used HPV vaccine in the U.S. and offers protection against 90% of cervical cancers as well as a large proportion of vulvar, anal and other HPV-associated cancers, and genital warts [4]. The Advisory Committee on Immunization Practices (ACIP) recommends HPV vaccination for all children and adults ages 11–26, starting as early as 9 years old, and recommends shared clinical decision-making between patients and providers aged 27–45, particularly among individuals who may be at risk for new HPV infections and have not been adequately vaccinated [5].

In the U.S., the HPV vaccination is covered by the Vaccines for Children (VFC) program for children under 19 years of age [6] and by most health insurance plans, including the plans part of the Affordable Care Act, at no cost to patients [7]. In 2018 51.1% of U.S. adolescents aged 13–17 years completed the HPV vaccination series with adolescents of Latin American descent (hereafter, the gender non-specific term Latinx is used) displaying the second highest HPV vaccination completion (56.6%) after American Indian/Alaska Natives (57.3%) and highest completion of ≥ one dose (75.5%) [8]. However, these national estimates may be misleading, particularly given the major geographic disparities in completion of the HPV vaccine series which ranges from 32.6% in Mississippi to 78.1% in Rhode Island (50.2% in Alabama) [8].

National estimates, though, do not take into account the heterogeneity within racial/ethnic sub-populations including U.S.-born Latinx and immigrants [9]. For instance, it has been shown that Latinx parents’ length of time of the U.S. is associated with HPV vaccination uptake among their children [10]. As well, access to health care among children of Latinx immigrant parents, particularly in states where the growth of the Latinx population is relatively a new phenomenon, is influenced by a number of factors, most notably citizenship and/or legal status of the child, legal status of the parents, length of time parents have been in the U.S., financial barriers, cost and/or lack of health insurance coverage, communication barriers, and perceptions of racism or disrespect which, in turn, could influence HPV vaccination knowledge, acceptability, and, consequently, uptake [11].

Studies focusing specifically on barriers to HPV vaccination among Latinx parents have suggested that lack of awareness about the HPV vaccine generally or the link between HPV and cervical cancer, cost, and lack of provider recommendation are among the most frequently cited barriers [10, 12–14]. On the other hand, Latinx parents tend to have positive views toward vaccination, high levels of confidence in medical providers, and report high rates of intention to vaccinate their children if recommended by a health care provider [15, 16] with mothers traditionally having the responsibility for the family’s health decisions [17]. While reluctance to discuss sexual behaviours with children was a factor for many parents [18] or worry that a daughter might become sexually active after vaccination for others [19], for some, protecting their child’s health was more important than these concerns [13].

Alabama, has been considered an emerging Latinx state, with one of the fastest growing Latinx immigrant populations in the U.S. [20] and there have been a few qualitative studies conducted in order to develop efficacious interventions for HPV/cervical cancer prevention [21–23]. Results from these studies highlighted the need for a culturally-relevant, community-focused HPV education intervention focusing on Latina mothers.

The Community Preventive Services Task Force recommends four broad strategies to increase appropriate vaccination, including HPV vaccination: enhancing access to vaccination services, increasing community demand for vaccinations, using provider- or system-based interventions, and implementing a coordinated approach that combines interventions at the community level [24]. Taken together, the previously identified barriers and facilitators to HPV vaccination among Latinx immigrant parents as well as the barriers to health care access in a Latinx emerging state such as Alabama, increasing community demand may be a promising strategy in promoting HPV vaccination among children of Latinx immigrants.

Therefore, the purpose of this study was to assess the efficacy of a community-based, theory-driven, culturally relevant intervention to promote HPV vaccination uptake among daughters of Latina immigrants 9 to 12 years of age in Alabama through a cluster-randomized controlled trial. The primary outcome was completion of HPV vaccination series at 7-month follow-up through medical records/state vaccination registry review.

2. Methods

2.1. Study design

The study consisted of a cluster-randomized controlled trial that compared the promotion of HPV vaccination (HPV arm) with the promotion of healthy eating (control arm) across eight counties in Alabama with sizable Latinx populations for recruitment. Earlier work within the Latinx community and formative assessments early in the design phase indicated that healthy eating was a topic of interest and that this would make an adequate comparison arm. Because both interventions were fully implemented intervention programs, one arm was able to serve as the control for the other.

A cluster-randomized study, the unit of randomization was “place of residence” (i.e. mobile home parks, neighbourhoods, apartment complexes). The decision to randomize at the cluster level was the result of our previous work with Latinx immigrants throughout the state. In Alabama, Latinxs tend to live near other Latinxs in “clusters” or groups and tend to socialize within these clusters, in effect, isolating themselves from non-Latinxs. As well, since transportation is usually a barrier for Latinas, we felt the intervention should take place where they are located. Therefore, recruiting at the cluster level would decrease contamination across the arms. The research protocol was reviewed and approved by the University of Alabama at Birmingham Institutional Review Board. Informed consent was obtained from all participants. Daughters signed assent while mothers provided signed consent.

2.2. Theoretical framework

Two theoretical frameworks guided the preliminary formative assessments as well as intervention development and implementation, the PEN-3 model and the Health Belief Model (HBM). See Figure 1. The PEN-3 model of health behaviours has three interrelated dimensions, each with three components that make up the PEN acronym [25]. The first dimension, health education, defines the target audience (person, extended family, and neighbourhood). The second dimension consists of the educational diagnosis of a health behaviour with a focus on identifying the factors influencing the person, family, and/or community actions (perceptions, enablers, and nurturers). The third dimension is the cultural appropriateness of a health behaviour, which is integral in the development of culturally relevant interventions among racial/ethnic minorities, the components of which are positive, exotic, and negative [25, 26]. The “positive” component refers to perceptions, enablers, and nurturers that lead the target audience to engage in the health behaviour. The “exotic” component refers to practices that have no harmful health consequences and should not be changed but incorporated into the intervention. The “negative” component refers to perceptions, enablers, and nurturers that lead the target audience not to engage in the health behaviour or to engage in a harmful behaviour.

Fig 1.

Theoretical Framework that guided study process included PEN-3 Model of health behaviors, which has three interrelated dimensions, each with three components, and the Health Belief Model, which posits that health behaviors depend on individual’s perceptions of vulnerability to a disease or condition.

Although the PEN-3 model takes into account cultural sensitivity and appropriateness in data collection and intervention development/implementation, in our previous work we have found that intrapersonal factors are strongly associated with engagement in healthy behaviours. However, the “perceptions” component of the PEN-3 does not truly capture these variables to be able to inform intervention development [21, 27], which are addressed in the Health Belief Model (HBM). The HBM is based on the premise that changes to health behaviours hinge on individuals’ beliefs that: 1) they are vulnerable to a disease or condition (e.g. exposure to HPV infection), 2) the health condition could have serious consequences (HPV may lead to cervical cancer/death), 3) a particular action will reduce the vulnerability to the threat (receive a vaccination), 4) the benefits outweigh the costs, and 5) they can perform the action to prevent the disease or condition (self-efficacy) [28, 29]. Table 1 outlines the preliminary findings that informed intervention development based on these theoretical frameworks.

Table 1.

Summary of Preliminary Findings Based on Theoretical Framework (PEN-3 and HBM^*)

Theoretical Constructs	Preliminary Findings
PEN-3
Perceptions	Positive
	• Some knowledge that cervical cancer can be sexually transmitted • Belief that vaccinations (in general) are good health habits and necessary for children • Perceived responsibility as the caretaker in terms of the health of the family • Belief they have control over their own health and interest to engage in preventive practices • Receptive to health education/Doubt they might be misinformed
	Negative
	• Lack of knowledge regarding HPV, HPV infection, cervical cancer, screening, and HPV vaccine • Concerns regarding HPV vaccine safety and side effects • Concerns of sending a message to daughters that they can have sex since they got the vaccine
Enablers	Positive
	• Trust in lay health educators and UAB – welcome in their homes Negative • Lack of health insurance, lack of transportation, cost, language barrier • Lack of knowledge of where to obtain the vaccine • Lack of recommendation by providers • Differences in health care system from home country • Fear of going to the doctor due to legal status
Nurturers	Positive
	• Strong alliance with other Latinas and desire to help each other • Knowing other mothers who vaccinated their daughters strong motivation to vaccinate their own • Getting information from other Latinas may be motivating factor
	Negative
	• Reluctance to talk to spouse about vaccinating daughters • Unsure how to talk to other family members about HPV vaccination • Not sure how to talk to daughters • Concerns regarding parenting skills – conflict between parenting style in home country and in U.S.
HBM
Perceived	Negative
Susceptibility	• Belief they are not at risk for cervical cancer (e.g. because they clean themselves after intercourse). Consequently, daughters not at risk • Preventive care is not a priority
Perceived	Positive
Severity	• Understanding that cervical cancer is a deadly disease
Perceived	Negative
Barriers	• Structural (language, lack of transportation, lack of health insurance, do not know where to go) • Reluctance to talk to a health care professional about getting their daughters vaccinated for a STI
Perceived	Positive
Benefits	• Belief that vaccination in general are good • Belief that communication with partners can be helpful in getting daughters vaccinated • Belief that they will prevent a serious disease such as cervical cancer
	Negative • Concerns over vaccine safety, side effects, and cost
Self-efficacy	Positive • 68.1% are willing to vaccinate their daughters between the ages of 9 and 12 • Desire to communicate better with their daughters
	Negative • Lack of communication skills with partners and daughters regarding getting daughter vaccinated

NOTE: HBM=Health Belief Model; HPV=human papilloma virus; STI=sexually transmitted infection

^*Structural perceived barriers overlap with negative enablers

2.3. Intervention and Control Arms

Both study arms were delivered by trained Lay Health Educators (LHEs) in Spanish. Each intervention had a different LHE assigned to avoid cross-intervention contamination. Intervention contact hours with LHEs were identical in both arms to control for contact time. Each study arm included an orientation where participants provided informed consent, program was described, and baseline assessments completed. The HPV arm consisted of four group sessions and one individual session. Each group session focused on specific topics with the first session introducing the program, the second discussing HPV and cervical cancer, the third on HPV vaccination and how to talk about HPV with partners and daughters, and the fourth on the importance of communication and self-responsibility. The individual session was a home visit, occurring between the third and fourth group sessions. At that time, the LHE met with the mothers in their homes to review course material and to talk about individual mother/daughter issues in related to communication and/or HPV vaccination. The control arm consisted of four group sessions and one individual session focusing on healthy eating and interpretation of nutrition labels. See Table 2 for detailed outline of both interventions.

Table 2.

Outlines for Intervention Sessions – HPV Vaccination and Healthy Eating

HPV Vaccination Arm	Healthy Eating Arm
Group Session 1: Introduction to Program Understand the goals of the program Commit to participating in the program Discuss Culture and Gender pride Understand importance of familismo Review parenting skills Discuss communication skills	Group Session 1: Introduction to Healthy Eating Understand the goals of the program Commit to participating in the program Learn the importance of variety in fruit and vegetable consumption Learn about Portion Sizes Introduction to nutrition labels Introduction of the Food Diary Receive lunch box with dividers for personal use
Group Session 2: HPV/Cervical Cancer Understand what cervical cancer is Understand what a Pap smear is and the importance of having one for early detection of cervical cancer Understand the steps involved in problem solving Demonstrate ability to problem solve to reduce barriers to healthcare access for self and friends	Group Session 2: Healthy Shopping and Eating Out Recognize how to identify foods high in sugar, salt, and fat on nutrition labels Learn tips on choosing healthy foods when eating out Understand how to plan a healthy grocery list Review strategies to decrease fried food intake Discuss food diaries
Group Session 3: HPV Vaccination Learn the importance of HPV vaccination Discuss misgivings about giving the vaccination to daughters Learn how to become comfortable discussing HPV with partners and daughters	Group Session 3: Guide for a Healthy Pantry Discuss priorities for changes toward healthy eating Demonstrate understanding of nutrition labels using pantry items from home Participate in interactive activities
Individual Session: Home Visit Review knowledge about HPV vaccination and cervical cancer prevention Describe individual mother/daughter issues in relation to communication and/or HPV vaccination Demonstrate understanding of importance of good communication skills with children and partner Assess/assure mother’s self-efficacy in the context of perceived barriers and facilitators to HPV vaccination	Individual Session: Home Visit Review items in home pantry Demonstrate understanding of nutrition labels Explain plan to incorporate healthy eating changes for family diet
Group Session 4: Communication and Importance of Self-responsibility and Good Communication Articulate the role body language plays in the listening process Demonstrate ways to be a good listener Indicate ways to build trust and enhance reliance on others Participate in a cooking activity	Group Session 4: Healthy Cooking Review healthy eating habits Summarize what was learned in the program Prepare a healthy meal using nutrition labels and strategies learned in the program

During the formative assessment phase of the study, we found that mothers were very reluctant to discuss sexuality with their daughters and suggested that different interventions should be developed for mothers with daughters between the ages of 13 and 17. Given the early sexual initiation among Latinas, we believe that our greatest investment was to promote vaccination among girls between ages 9–12. At the time of study commencement in 2013, HPV vaccination was approved for use in both boys and girls, but during the formative assessments prior to the grant submission in 2011, routine vaccination of boys had yet to be recommended by the ACIP and it was not reimbursed by the Vaccines for Children (VFC) program. Hence, the focus of this intervention on girls between the ages of 9 and 12.

2.4. Recruitment and randomization

Ensuring enough participants for both arms at each study site, sequential randomization was utilized. A recruitment strategy successful with prior studies was implemented. Potential areas with a high percentage of Latinx residents were identified by conducting a door-to-door census. During that brief interaction, eligibility and interest in participating was determined. Eligible participants had the following criteria: 1) be 18 years of age or older, 2) have at least one daughter between 9–12 years of age who had not had the HPV vaccine, 3) be a Latina immigrant, and 4) reside in that particular location. Once two viable locations were found, a third party (not involved in the study) randomized each location to the two arms.

Each group’s assigned LHE contacted the potential participants identified as eligible from the door-to-door census and invited them and their daughters to attend an orientation session. During the orientation session, participants completed the consenting documents and baseline assessments, administered by data collectors who had no part in the interventions.

2.5. Assessments

Both arms used identical assessments. Mothers completed a 91-item questionnaire administered via a structured interview at baseline and self-administered questionnaires, with 23 items, were completed by daughters. The daughters’ questionnaires were limited to demographics, healthy eating, and nutrition label interpretation. Mothers’ interviews were more extensive as they also included questions related to health insurance status, where they go for medical care, cervical cancer screening, HPV awareness/knowledge, and reasons they might or might not want their daughters vaccinated against HPV.

Participants also completed a post-test within two weeks of the last group session. Seven months after the post-test, a follow-up questionnaire was completed. Mothers received 20 USD in cash for each completed questionnaire; as theirs were shorter, daughters were given 10 USD for each completed questionnaire. HPV vaccination was confirmed through medical records and/or state vaccination registry.

2.6. Sample size and power calculation

The power calculations were based on a test of proportions. We based the power calculations on one of our previous studies that indicated that 68% of Latina immigrants with daughters between the ages of 9 and 12 years indicated willingness to vaccinate their daughters. Using a two-sided two-group continuity corrected χ² test of equal proportions and α=0.05 unadjusted estimates revealed that we determined we would have 80% power to detect an absolute difference of 20% with 76 subjects per arm when we compare compliance rates of HPV vaccination between the two groups by assuming we observe 68% compliance rate in the control group. We assumed that the intervention condition could further persuade those mothers who initially responded ‘maybe’ to vaccinating their daughters (24.6%) in our previous study. Calculating the inflation factor to account for clustering, and adjusting the previous power calculations, we observed that maintaining the 80% power with an intra-class correlation of 0.1, a sample size of 130 subjects (17 groups with 8 participants per group on average) were required per intervention arm. Further inflating this number for possible non-compliance or dropout (estimated at approximately 20%) we determined we would need 20 groups of 8 participants per intervention arm (N=320). As sample size of the final data is 278 participants (136 HPV and 142 Control arms) with 9% dropout rate, the statistical analysis was performed with larger statistical power than was estimated.

2.7. Statistical analysis

Demographic analyses compared participants characteristics between study arms at baseline and 7-month follow up with chi-square statistics for significance. Unadjusted and adjusted multivariate logistic regression models estimated differences between arms of perceived risks associated with cervical cancer and HPV vaccination for daughters completing the first, second and third doses. An adjusted model controlled for mother’s age, time in the U.S., income, and daughter’s health insurance status. As we found there was no significant clustering effect in HPV vaccination, no further account for clustering effect was performed when the adjusted model was estimated. At the time of the study commencement (2013), the quadrivalent HPV vaccine series consisted of three doses [30] and the current two dose 9-valent vaccine in use did not become exclusive until 2017 [31]. Consequently, the primary outcome was daughters’ completion of three HPV vaccine doses through medical records/state vaccination registry review. Due to the timing of vaccine doses, the third dose for some having received second doses were not feasible with follow-up assessments occurring at seven months. Therefore, the vaccination confirmation for third dose was extended to 13 months after the end of the intervention.

3. Results

3.1. Enrolment and retention

Three hundred seventeen mother-daughter dyads attended an orientation meeting and completed baseline questionnaires between May 2013 and October 2017. See Figure 2 for CONSORT Flow Diagram. Of these, three mothers were born in the U.S. and another 21 were found to have vaccinated their daughters prior to the orientation meeting (based on medical records review conducted at the end of the study) rendering them ineligible for analyses. Of those remaining, 15 did not attend any of the sessions after the orientation leaving 278 participants (HPV arm=136; control arm=142) that commenced the intervention. Seven months after the last session, 259 participated in the follow-up interview questionnaire (HPV arm=126; control arm=133). The overall retention rate was 93.2% (92.6% for the HPV arm and 93.7% for the control arm). The sample size was smaller than determined in the power analysis; however, with the higher than estimated retention rates for both arms at 7-month follow up, sample size was determined to be adequate for the primary outcome. With 278 participants (136 HPV and 142 control arms) with 9% dropout rate. The trial ended after results from the 7-month follow up for the original 40 clusters were completed.

Fig 2.

CONSORT Flow Diagram

3.2. Intervention Results

Results from descriptive, bivariate, and multivariate analyses are in Tables 3–5. Of 278 participants at baseline, mothers averaged 35 years of age, had completed just under nine years of education, had been in the U.S. less than 13 years, and had an income of just under 20,000 USD/year. Most worked at least part-time, were married or living with a partner, and very few had health insurance, while most daughters had health insurance coverage. There were no significant differences between the intervention and control groups with regard to demographics with two exceptions: mothers in the HPV arm had lived in Alabama a shorter length of time and had higher monthly incomes compared to mothers in the healthy eating arm. (Table 3).

Table 3.

Participant Characteristics by Study Arm (N=278)

	HPV Vaccination Arm (n=136)	Control Arm (n=142)	p
Age1 (years)
Mother	35.4 (5.9)	34.8 (5.1)	0.302
Daughter	9.8 (0.9)	9.8 (1.0)	0.813
Education1 (years) – Mother	8.7 (3.5)	8.9 (2.5)	0.852
Time in U.S.1 (months) – Mother	151.1 (54.8)	150.4 (50.2)	0.914
Time in Alabama1 (months) – Mother	120.3 (54.5)	134.4 (50.3)	0.026
Monthly Income1 ((USD)	1,775.34 (1068.4)	1,549.64 (553.2)	0.029
Marital Status - Mother
Single %	2.9	8.5	0.083
Married/Living Together %	94.9	87.3
Separated/Divorced %	2.2	4.2
Employment Status - Mother
Full-time %	14.0	22.5	0.293
Part-time %	33.1	29.6
Homemaker %	48.5	45.1
Unemployed %	4.4	2.8
Health Insurance Coverage, including Medicare/Medicaid
Mother (% with)	8.1	12.0	0.282
Daughter (% with)	78.5	76.8	0.726

¹Mean (SD)

Table 5.

Multivariate Associations – Perceived Risks by Daughter’s Dose of HPV Vaccination at follow-up^*

	Completed 1st HPV Vaccination Dose			Complete 2nd HPV Vaccination Dose			Completed 3rd HPV Vaccination Dose
Unadjusteda	OR	95% CI		OR	95% CI		OR	95% CI
HPV vs. Control	5.96	3.38	10.49	8.09	4.0	16.35	16.5	5.73	47.48
Perceived risk for cervical cancer
Yes vs. No	1.85	0.94	3.65	2.28	1.04	4.96	1.64	0.69	3.88
DK/NS vs. No	1.32	0.68	2.56	1.51	0.69	3.3	1.38	0.59	3.23
Worry about getting infected with HPV
Yes vs. No	0.85	0.46	1.58	1.14	0.57	2.3	1.02	0.47	2.23
DK/NS vs. No	0.7	0.27	1.8	0.65	0.21	2.03	0.7	0.2	2.44
Perceived risk for HPV infection
Yes vs. No	1.26	0.6	2.61	1.0	0.44	2.28	0.56	0.2	1.56
DK/NS vs. No	0.71	0.41	1.23	0.77	0.42	1.41	0.64	0.33	1.27
Worry daughter would be infected with HPV in future
Yes vs. No	1.46	0.38	5.63	0.92	0.24	3.58	1.05	0.22	5.02
DK/NS vs. No	0.7	0.14	3.45	0.2	0.03	1.4	0.16	0.01	1.99
Perceived risk for daughter’s HPV infection in future
Yes vs. No	0.81	0.31	2.09	0.47	0.18	1.24	0.41	0.15	1.12
DK/NS vs. No	0.61	0.22	1.66	0.36	0.13	1.03	0.24	0.08	0.75
Adjustedb
Intervention (HPV) vs. Control	5.96	3.32	10.72	7.52	3.62	15.62	15.82	5.26	47.6
Perceived risk for cervical cancer
Yes vs. No				1.25	0.53	2.93
DK/NS vs. No				1.03	0.44	2.44
Perceived risk for daughter’s HPV infection in future
Yes vs. No							0.69	0.23	2.1
DK/NS vs. No							0.76	0.2	2.82

^*N=259 participants at follow-up. Intent to treat HPV arm=126, control arm=133.

^aUnadjusted statistically significant (p<0.05) OR in bold.

^bOR adjusted by mother’s age, time in US, income, and daughter’s health insurance status. No demographic significant. Significant HPV predictors only shown in adjusted multivariate logistic models.

There were significant differences between intervention and control arms in mothers’ perceptions of risk associated with cervical cancer (22.8% and 11.4% respectively) and their own risk of HPV infection (10.3% and 5.7% respectively) at baseline. However, there was an increase in perceived susceptibility to cervical cancer (44.8%) and HPV infection (25%) at 7-month follow-up among mothers in the intervention arm as compared to mothers in the control arm (23.9% for cervical cancer and 3.3% for HPV infection). There were also significant differences between the two arms with regard to mother’s worry of future HPV infection with these differences being greater at 7-month follow-up (Table 4).

Table 4.

Bivariate Analyses Baseline v. Follow-up Perceptions by Study Arm

	HPV Arm		Control Arm
	Baseline (n=136)	Follow-up (n=126)	Baseline (n=142)	Follow-up (n=133)	p1	p2
Perceived risk of cervical cancer (mother)
% Yes	22.8	44.8	11.4	23.9
% No	22.1	16.2	22.9	32.4	0.038	<0.001
% DK/NS	55.1	39.0	65.7	43.7
Perceived risk of HPV infection (mother)
% Yes	10.3	25.0	5.7	3.5
% No	59.6	43.4	49.3	59.3	0.027	<0.001
% DK/NS	30.1	31.6	45.0	47.2
Perceived risk of daughter’s future HPV infection
% Yes	59.6	71.3	40.1	52.1
% No	9.6	11.8	4.2	2.8	<0.001	<0.001
% DK/NS	30.9	16.9	55.6	45.1
Worry about future HPV infection (mother)
% Yes	68.4	77.2	62.4	59.9
% No	22.8	16.2	17.7	24.6	0.029	0.005
% DK/NS	8.8	6.6	19.9	15.5
Worry about daughter’s future HPV infection
% Yes	83.1	89.7	82.4	81.7	0.288	<0.001
% No	5.1	6.6	2.1	1.4
% DK/NS	11.8	3.7	15.5	16.9
Completed 1st dose of HPV vaccine (daughter)
% Yes	-	52.2	-	15.5	-	<0.001
% No	-	47.8	-	84.5
Completed 2nd dose of HPV vaccine (daughter)
% Yes	-	40.4	-	7.8	-	<0.001
% No	-	59.6	-	92.2
Completed 3rd dose of HPV vaccine (daughter)
% Yes	-	32.3	-	2.8	-	<0.001
% No	-	67.7	-	97.2

DK/NS=Don’t know/Not sure; HPV=human papillomavirus

¹p-value between arms at baseline

²p-value between arms at 7-month follow-up

With regard to perceived susceptibility to HPV infection among their daughters, there were also significant differences between the two arms at baseline (59.6% among mothers in the intervention arm and 40.1% among mothers in the control arm). At 7-month follow-up, 71.3% of mothers in the intervention arm indicated that their daughters could be at risk for future HPV infection as compared to 52.1% of mothers in the control arm. There were no significant differences between the two arms at baseline with regard to mother’s worry about daughter’s future HPV infection, but significant differences were observed between the two arms at 7-month follow-up (Table 4).

For primary outcome, daughters in the intervention arm were significantly more like to receive one, two, and three doses of HPV vaccine than daughters in the control arm (Table 4). In the multivariate analysis, mothers in the HPV arm had a six times greater odds of vaccinating their daughters with the first dose, eight times greater odds of vaccinating their daughters with the second dose, and just over 16 times greater odds of completing the three-dose HPV vaccine series than mothers in the control arm after adjusting for mother’s age, time in the U.S., income, and daughter’s health insurance status. Only perceived risk of their daughters’ future HPV infection remained significant as a predictor of three-dose HPV vaccination completion (Table 5).

3.3. Satisfaction/Intervention Feedback

When mothers were asked about their favourite program component, the most frequent responses were “everything” (40%), learning about HPV vaccination (16%), and learning about health in general (14%). These responses were consistent with their answers regarding the most important things they learned in the program: importance of HPV vaccination as a preventive strategy (37%), and how to prevent cervical cancer (31%). When asked about their least favourite components, 81% of mothers indicated “nothing.”

4. Discussion

Our findings demonstrate that a community-based, theory-driven, culturally relevant intervention was efficacious in promoting HPV vaccination uptake among daughters of Latina immigrants 9 to 12 years of age in Alabama. Additionally, the obtained results provided valuable insights into perceptions of cervical cancer risk and HPV vaccination of Latina immigrant mothers in the Deep South. Consistent with the Community Preventive Task Force recommendations [24], most interventions to promote HPV vaccination have focused either on providers/health care systems (improve/maximize delivery) [32–35] or parents/community (increase demand) [36–38]. Given that provider recommendation has been shown to be one of the strongest predictors of HPV vaccination [39, 40], most of the interventions to date have focused on providers/health care systems with mixed results [32–35].There have been a few studies targeting parents and children/adolescents and results have been promising. However, most of these studies used quasi-experimental designs or were conducted in clinical settings rather than in the community which tend to be biased toward parents who may frequently use the health care system [36–38, 41].

To our knowledge, this is the first randomized control trial to test an intervention to promote HPV vaccine uptake among Latina immigrant mothers at the community level. Several aspects of the intervention design contributed to its success. First and foremost, the intervention was implemented where participants reside and in their native language by Lay Health Educators whose first language was Spanish, which addressed the two most frequent barriers experienced by Latino immigrants in participating in health education program (transportation and language barriers). The community-based component using a group session format provided an opportunity for participants, who might feel lonely and isolated even within their enclaves, to engage and interact with community members they might not otherwise meet. Contrary to the assumption that residents of immigrant communities experience strong social connections [42], more current research suggests that Latinx immigrants tend to experience high levels of isolation and few social ties [43], possibly because of residential instability [44], resulting in lower levels of social cohesion, particularly when living in highly concentrated immigrant communities [45, 46]. This can lead to difficulties in accessing community resources that promote health and wellbeing [47].

Because Latinx immigrants typically live in homogenous enclaves in Alabama, considering dynamics related to shared beliefs and values was important in the intervention design. The need to belong and form relationships within a community with whom one can ethnically identify can lead individuals to purposely limit exposure to those from different cultural backgrounds [48]. This phenomenon motivated the incorporation of familismo to encourage participants to connect with the subject matter in a meaningful way. Family-first is an important component of Latinx culture and results from focus groups conducted during the formative period of the study indicated that it would need to be included in every phase of the study, from the topics of gender pride and cultural misgivings of communicating about sexual behaviours to the inclusion of daughters only in the final group session.

It was evident that perceived susceptibility is a strong determinant in the decision of whether or not to complete the HPV vaccination series among daughters of Latina immigrants, which is consistent with previous findings in the literature, including our own work in cervical cancer screening [21–23, 49]. Future studies are needed to better understand the “don’t know/not sure” responses when asked about perceived susceptibility to cervical cancer and HPV infection given that approximately 1/3 of participants indicated this response in the baseline questionnaire, which, in turn, can inform more targeted interventions.

This study has limitations that should be acknowledged. Although the intervention was developed based on extensive formative assessments, it was clear that the intervention format and content would be different between younger and older youth. Also, given the more recent ACIP recommendation to vaccinate boys, this study has the limitation of focusing only on girls. Despite its limitations, we believe that the study has several strengths. First, the cluster-randomized trial implemented at the community level adds rigorousness to the study design. Second, the primary outcome was based on medical records/vaccination registry rather than self-report. Third, we were able to test two interventions by having a theory-driven, culturally relevant intervention to promote healthy eating as the control group. There were no observed harms to any participant.

5. Conclusions

Theory-driven, culturally relevant community programmes can lead to positive changes in preventive behaviours when it comes to HPV vaccination and cervical cancer awareness. Using focus groups during the formative assessment phase is an important first step in incorporating concepts that are important to homogenous participants from target populations in order to effectively promote preventive behaviours such as HPV vaccine uptake and, ultimately, improve health outcomes.