Procedures

Our mobile internet intervention study procedures, from a NICU exit referral to a mobile intervention, provide a unique framework for examining progression flow through the crucial junctures that mirror the Part C EI system gateways: referral, screening, assessment, and intervention [18]. After institutional review board approval, recruitment efforts focused on 3 Level 3 NICUs serving the urban poor in a Midwestern city. These NICUs were selected because they were part of a medical conglomerate with similar characteristics that included a centralized geographic location in the urban core within 2 square miles of one another, similar annual admission rates, and a racially and ethnically diverse patient population, including those who lack insurance and the ability to pay. Through an iterative process, the research team conducted a series of meetings with each NICU team to generate a referral accountability plan, which was documented by the research team and provided to the NICU team for review and revision until the NICU team confirmed that their plan was complete and accurate. Each NICU-generated plan specified the NICU personnel who would share referral information with mothers, collect cards that mothers signed indicating their interest in being contacted by the study team, send electronic referrals to the research team, and respond to a biweekly prompt to provide an electronic referral update. Electronic referral update reports included the number of eligible mother-infant dyads in the NICU during the most recently completed referral period, the number of mothers with whom referral information was discussed, the outcome of each referral discussion, barriers to referral, and identified solutions. The research team provided referral materials to each NICU, which included service provider posters with mother and infant eligibility criteria to remind and prompt providers to refer all eligible mothers, a mobile intervention study letter to be shared by providers with mothers, a mother interest card for mothers to grant permission for study team follow-up, and a script for providers to use when sharing referral materials and collecting mother interest cards. An electronic NICU referral mechanism was established for NICU service providers to connect mothers and their low-birth-weight infants to a randomized controlled trial of a mobile internet intervention.

Referral criteria included biological or adoptive mothers, living in the metropolitan area of the NICU, who spoke English and whose infants at birth weighed <2500 g, were at least 24 weeks’ gestational age, were no more than 5 months corrected gestational age at NICU exit, and who were not diagnosed with hydrocephalus, bronchopulmonary dysplasia, or beyond a grade 3 intraventricular hemorrhage. Referral criteria were established to avoid potential study burden for mothers of infants who were experiencing acute medical crises, including a high risk for NICU return or intensive care unit (ICU) entry. NICU teams were encouraged to refer all eligible mothers and infants in addition to any and all other service referrals such that all referred mothers were free to participate in existing community service referrals as usual without exclusion.

On receipt of each electronic referral, the research staff recorded the date of referral, referral source, and referral contact information into a project database. Research staff mailed consent forms to referred mothers and contacted them by phone to (1) confirm referral eligibility criteria, (2) review and discuss the consent form, and (3) determine whether mothers viewed themselves as able and willing to engage in the intervention study. Mothers who could not be reached by phone because of a disconnected number or failure to connect after at least five attempts were sent a letter encouraging mothers to contact the study team if interested in the program. Mothers’ perceived ability to participate in the study was determined on the basis of their negative responses to a brief structured interview question in which mothers were first informed of personal situations that should be prioritized over intervention study participation, such as homelessness, shelter residence, inpatient mental health or substance abuse treatment, or a major physical or mental illness requiring intensive treatment such as schizophrenia, cancer, or HIV/AIDS. Mothers were then asked whether they were experiencing one or more of these situations or any other situation that could interfere with their ability to participate in the intervention study. An affirmative response was exclusionary and met the criteria for intervention study ineligibility. For mothers who were screened eligible and agreed to participate in the study, an in-home assessment visit was scheduled. All contact attempts, the outcome of each contact attempt, the eligibility screening outcome, and the scheduled assessment date were recorded in the project database.

Informed consent was obtained at the onset of a 2-hour, in-home assessment visit. Electronic questionnaires were completed by mothers on the web via Qualtrics entry on an iPad (Apple Inc) to provide information about demographics and maternal and infant risk characteristics. The Measurement Domains and Measures Section provides a full description of the measurement domains and measures. Assessments were conducted by research assistants who had obtained at least a bachelor’s degree in education, human development, or psychology and had at least 2 years of intervention research experience conducting in-home assessments and mobile intervention protocols with mothers and infants. Assessors were trained and observed to implement the assessment protocol with fidelity before data collection. The assessment details are also provided in the Measurement Domains and Measures section.

Following assessment, mothers were randomized to 1 of 2 mobile internet interventions with identical structures. For both groups, the number of sessions and structural components of each session included (1) a web-based self-directed learning program through video-based teaching with check-in questions and provision of immediate corrective feedback, (2) an action plan outlining daily activity practice (homework) based on session content, (3) parent-recorded video and secure upload of session skill practice during interaction with her infant, and (4) a video-based coach call to coview the parent-recorded video of interaction with her infant [20]. For both intervention groups, meaningful access to a mobile internet intervention was operationalized as (1) mothers’ completion of an in-home intervention session in which mothers were fully guided and scaffolded to interact with each mobile intervention component (ie, video modeling content, review questions, action plan, video creation, and coach call) and (2) mothers’ completion of each of the above content components of the remote intervention session with on-demand scaffolding provided through messaging, phone, or video call.

At the in-home intervention orientation visit, all mothers were given an iPhone with unlimited data, text, and call plan. They were granted entry into a 12-session mobile internet intervention. Coaches used a demonstration video to introduce mothers to the mobile intervention, use the mobile phone features, and navigate through the first mobile intervention session. Coaches verbally scaffolded mothers’ use of each session component by providing the phone and materials to the mother and serving as a guide on the side when mothers navigated through the entire first session, including the coach call procedures.

After the first session with coach guidance and full scaffolding, mothers autonomously completed the second intervention session with on-demand remote coach scaffolding between and during coach calls. The demand context for the coach response included (1) questions from the mother and (2) coach electronic monitoring of mothers’ progress or nonprogress through intervention session components and feedback to celebrate mothers’ successes and address barriers to progress. We expected that this meaningful access support in the first 2 sessions would increase the probability of mothers’ continued progress in completing the remaining 10 remote sessions.