NET-Works:Community Preschooler Obesity Prevention

1. Trial design overview NET-Works is a two-arm, randomized controlled trial to test the efficacy of a multi-setting, multi-component intervention approach to preventing obesity among racially/ethnically diverse preschool age children. The NET-Works intervention includes four main components: 1) a pediatric primary care brief counseling intervention; 2) a home-based intervention delivered by NET-Works family connectors to support parents in making changes in the home environment and parenting practices to promote healthful eating and activity patterns; 3) community-based parenting classes designed to parallel the home-based intervention curriculum and provide social support to participating parents; and 4) linkages to neighborhood and community resources to support parents in promoting healthful eating and activity patterns for their children. Five hundred parent/child dyads were randomized to either the NET-Works intervention or a usual care comparison condition and followed for three years. Participants were assessed at baseline and annually. The primary hypothesis is that children randomized to the NET-Works intervention will have lower BMI (body mass index) at two and three years post-randomization relative to usual care comparison group children. BMI is the primary outcome across all four Childhood Obesity Prevention and Treatment Research (COPTR) trials. Recruitment for the trial began in July 2012 and was completed in December 2013, with the final three year follow-up data collected in December 2016. 2. Study setting and population The target population for NET-Works was racially/ethnically diverse preschool children and their parent or primary caregiver. To reach the intended population, NET-Works partnered with 12 primary care clinics and three managed health care systems that serve diverse populations with respect to race, ethnicity and income. Over 18 months, 500 families were recruited, enrolled and randomized to the intervention or to the usual care comparison group. Administrative databases and centralized electronic scheduling systems at the partner clinics provided data needed to target recruitment efforts on two-to-four year old children residing in certain zip code areas. The clinics had data from preventive care visits available for calculation of child body mass index (BMI) percentile to focus recruitment efforts on children who were potentially BMI-eligible for the study. 3. Recruitment Administrative databases and centralized electronic scheduling systems at three partner health care organizations provided data needed to target families of two-four year old children whose BMI percentile was at or above the 50th percentile for age and gender. Clinic liaisons obtained approval from each child's pediatrician to send a study invitation letter to the parent or primary caregiver. A list of physician-approved potentially eligible children was shared with the NET-Works recruitment team. NET-Works sent the parent or primary caregiver a study invitation letter from the pediatrician and researchers. Approximately five working days later, NET-Works recruitment staff followed up by phone to provide more information, ask additional screening questions, and assesses parent/primary caregiver's interest in the study. If the parent/primary caregiver was interested and the child appeared to be eligible, a home visit for eligibility confirmation, consent, and data collection was scheduled. The purpose of this initial home visit was to: 1) explain and discuss the study, and begin to develop rapport with parents; 2) measure the index child's height and weight to confirm eligibility; 3) obtain informed consent; and 4) begin the process of collecting baseline data. Based on our pilot study experience, we anticipated the majority of the participating parent/caregivers would be female and a significant number of Hispanic families as well as non-Hispanic White, African American, and multi-race participants would be recruited. 4. Randomization Prior to the start of recruitment and enrollment, the study statistician created 6 blocked randomization schedules, one for each age group (2, 3, 4) by gender (M, F) stratum, that would equally allocated children to the NET-Works intervention or the usual care comparison group in blocks of 10 to ensure equivalent study group size. Parent-child dyads were randomized to treatment and control conditions after completion of all baseline measures, including data eligibility requirements established by the COPTR consortium. Data eligibility requirements included valid height and weight measurement, a minimum of one weekday and one weekend day NDS-R (nutritional data survey-revised) previous day dietary recall and four days of valid accelerometry data. The study coordinator randomly assigned the participant to the condition shown in the next available slot in the stratum-appropriate pre-defined randomization schedule. The investigators and all assessment staff remained blinded to experimental assignment until after the final follow-up assessments were completed. 5. NET-Works intervention The NET-Works intervention was designed to reach into the home and family environment to influence parent behaviors and attitudes to support changes in the home that affect food availability, family meals, television viewing, and active play. Each intervention component would be situated in community settings where parents and children already live their lives. The intervention was expected to be effective in part through taking advantage of the settings that families already inhabit, and through coordination across settings to reinforce and link the messages, resources and feedback families receive about healthful food choices, physical activity, and body weight for their child. We believed the neighborhood-based model would be more likely to enhance class participation, reduce transportation needs/costs, allow the local community resources to be accessed more easily, and foster outside-of-class interpersonal connections among parents for enhanced social support. 5.1. Intervention components implementation overview The intervention curricula across each component (home visiting, parenting classes, neighborhood environment, pediatric primary care) was developed and refined during the pilot phase of the COPTR consortium. The intervention was designed to be implemented by trained research staff including a family connector who would conduct the home visiting and connector check-in calls and attend the parenting classes, and a parent educator who would conduct the community parenting classes that were based on the early childhood parenting class model and was developed in partnership with ECFE (early childhood and family education) in Minneapolis and St. Paul. The home visiting and parenting classes shared a common set of underlying curricula that included skills building around general parenting behaviors and the specific content behaviors targeted by the intervention (e.g., healthful snacks, family meals, physical activity). The family connector met parents and children in their own home, tailoring the intervention messages and strategies to best fit with the resources and motivations of parents. The family connector provided a one-on-one approach to synergize with the skills and strategies implemented in a group format in the NET-Works community parenting classes. Both home visiting and parenting classes directly connected parents to food and physical activity resources that already exist in their neighborhood ("community links"). Primary care providers bolstered these messages with parents during annual well-child visits. 5.2. Home visiting The goal of the home visiting was to help parents develop parenting skills, set goals related to child healthful eating, active play and reduced television viewing, and provided support for sustaining and building on these behavior and home environment changes. The dose for the home visiting intervention component was one visit per month for each of the three years. The family connector built a collaborative, partnership-based goal setting process for parents to facilitate child behavior change. Facilitation of the home visits was modeled on Motivational Interviewing, an approach designed to help individuals explore and resolve ambivalence about behavior change in a non-confrontational manner. Home visiting included a goal setting process guided by the family connector, healthful action activities to create norms and skills related to the targeted behaviors, and resources for links in the community to support healthful behaviors. Family connectors also checked in with participants to review progress on goals and provide support in between home visits. Family connector check-in contacts could be conducted over the telephone or via email. Family connectors had weekly supervision sessions with a licensed mental health provider with expertise in Motivational Interviewing to review progress with each participant, problem solve regarding any challenges that arise, and ensure adherence to the intervention protocol. 5.3. Parenting classes The parenting classes were designed to work synergistically with the home visiting curriculum. The parenting class and the home visiting curriculum both included topics that addressed parenting skills development, the home food environment, healthful eating, family meals, television viewing limits, active play, and goal setting. The parenting class provided a group format where parents could share their experiences, gain support from other parents and learn from each other. The curriculum was research-based and facilitated by a trained parent educator. The dose for the parenting class component was 12 classes per year for each of the three years. The curriculum in years 2 and 3 built upon and deepen the topics introduced in year 1, and addressed parenting and developmental issues as children progress from ages 2-7 years of age. Parent educators also participated in weekly supervision sessions to ensure adherence to the intervention protocol. 5.4. Neighborhood and community initiative The purpose of the neighborhood and community intervention component was to increase access to, and use of, healthful food, physical activity, and school readiness resources available in the neighborhood communities where study participants live. Parents were directly linked with resources in their neighborhood through both the parenting classes and the family connector home visits. Community initiatives implemented with the family connector included conducting a walkability assessment of the family's neighborhood. The family connector reviewed neighborhood food and physical activity resources with the parent at each home visit, and encouraged them to use these resources to support their home and behavior changes. The community links included one school readiness community activity (e.g., a library visit) to promote developmental parenting, sustain parent interest, and provide active support for whole child development. Four community links were included in the parenting class and home visiting curriculum each year of the intervention. 5.5. Primary care The primary care provider is an influential connection for parents. In NET-Works, the primary care provider delivered key messages around parent behaviors for shaping, reinforcing and sustaining healthful child eating and physical activity behaviors and body weight. To support providers in delivering key messages to parents during well child visits, a brochure was developed in partnership with the participating clinic systems. The brochure was used by the primary care provider to converse about child BMI percentile and strategies parents could use to promote their child's healthful eating and activity patterns. The NET-Works family connector will also provided information about progress on intervention goals to the primary care provider on an annual basis to reinforce continuity and integration of intervention messages. The primary care intervention linked with the community parenting classes and home visiting to deliver consistent intervention messages over a sustained time period. 6. Intervention component integration Linkages across the intervention components were created in several ways with the family connector playing a key role in this connectivity. Specifically, in addition to conducting home visits, family connectors facilitated participant entry into parenting classes and also attended parenting classes. Attending parenting classes allowed the family connector to be knowledgeable about the parenting class content and the experience their assigned participants were having during class. This information informed the goal setting that family connectors and parents worked on during home visits. Family connectors also served as the liaison between parents and primary care providers. The home visits and parenting classes also incorporated connections to community resources for physical activity and healthy eating as described above. 7. Usual care comparison group The usual care comparison group received the primary care provider component described above, and quarterly newsletters with information about their child's general health and wellness and school readiness. 8. Participant timeline, assessments, and measures Data collection took place in the home setting at baseline, 12, 24, and 36 months by trained and certified, bilingual English and Spanish research specialists, blinded to experimental assignment. At 6, 18, and 30 months, a brief telephone survey was conducted. Participants receive a total of $50 in gift cards for each set of measurement visits (baseline and each subsequent annual assessment) and a $5 gift card for each brief telephone survey. Data collectors were trained by the investigators and, for COPTR consortium common measures, by the Research Coordinating Unit (RCU) according to standardized protocols. COPTR used a train-the-trainer model. "Master Trainers" who participated in a central training organized by the RCU were responsible for training and certifying the data collection staff at their field center. Data collectors were required to demonstrate high inter- and intra-rater reliability prior to data collection. 9. Statistical analyses The primary efficacy analysis was designed to test whether two to four year old children who were randomized to the NET-Works program have statistically and clinically lower BMI at 24 and 36 months post randomization relative to usual care comparison group children with age at randomization (2, 3, 4), sex (F, M) and BMI at baseline used as covariates. According to the study protocol, all covariates in the primary analysis would be measured at baseline prior to randomization and therefore available for observed and unobserved BMI values. 10. Data monitoring and participant safety The data safety and monitoring plan included assessing adverse events systematically at each data collection visit with direct queries for all injuries, illnesses or other medical problems requiring a visit to a medical care provider and related to participation in the study. Serious Adverse Events (SAEs) were specifically monitored. Adverse events were also be recorded and evaluated when they came to the attention of study staff between the data collection visits. An independent Data and Safety Monitoring Board was selected to review study protocols and provide oversight of recruitment and study progress, data quality and completeness, and participant safety.