Mounting evidence supports the notion that high-quality childcare centers promote children’s mental health [1,2]. High-quality childcare protects and promotes the underpinnings of mental health in a phase of life when brain plasticity is most pronounced [3,4]; hence, it may prevent the onset of deviant developmental trajectories in toddlers and compensate for insufficient resources at home [5,6]. Evidence also suggests that centers of poor quality can have detrimental effects on children’s development [7,8]. Because more than 80% of 1 and 2 year olds in Norway attend childcare, and stay there for long hours [9], interventions targeting caregiver–child interactions may prove an important avenue toward universal prevention of mental health problems.

Childcare quality includes elements of both structure (eg, number of staff, staff training and education, group size, child-to-caregiver ratio) and process (eg, caregiver–child relationship, parental involvement) [10]. Process quality in terms of warm, sensitive, and stimulating relationships between caregivers and young children [11] is arguably the most important ingredient of quality as it has long-term effects on children’s mental health, well-being, and development [12]. Importantly, high process quality seems to be particularly important for children at risk [13].

New evidence shows that quality varies substantially between centers [14,15], and preliminary findings indicate that, by international comparison, quality in Norway is only moderately high [15]. Norway has failed to implement a minimum level of training for all childcare staff. As much as 60% of childcare staff have no formal qualifications on bachelor level for working with children [16]. Even some educated childcare teachers report lack of competence in taking care of children under the age of 3 [17]. This means that many toddlers spend most of their time interacting with staff with minimal or insufficient training. The need to enhance competence among childcare staff working with the youngest children is urgent. Moreover, a standardized and systematic procedure for measuring and monitoring process quality in childcare centers is lacking in Norway. This means that, unless a childcare center has participated in research, the level of quality that childcare centers provide to young children is largely unknown.

Few of the existing interventions for preventive quality improvement in childcare specifically focus on the youngest children. Furthermore, most interventions have been narrow in their scope, targeting only single topics such as conduct problems [18], anxiety [19], or language problems [20], thereby limiting their effect on children’s mental health. We will therefore break new grounds by conducting a cluster randomized controlled trial (RCT) of a newly developed comprehensive intervention for promoting mental health among toddlers through quality-enhancement in childcare, the Thrive by Three [21-23].

Thrive by Three is grounded in transactional models for development [24,25] and research showing that interactions between young children and caregivers are primary mechanisms of mental health, child development, and learning [26-29]. Thrive by Three is culturally sensitive, and is universally preventive in the sense that it includes all 1-3 year olds attending childcare. The intervention consists of 3 core elements: (1) quality building through competence enhancement for all staff regardless of level of formal education, (2) a model for peer-driven quality control and tailored supervision at classroom level in the participating centers, and (3) supervision of supervisors and mentors. Thrive by Three is modeled after the Thrive by Five, Seeds to Success, and Early Achievers quality-building and quality control measures, broadly implemented in childcare centers in the state of Washington in the United States [21-23] and piloted in a large Norwegian community sample [30].

This study has 2 primary aims: to determine the extent to which Thrive by Three (1) improves the quality of caregiver–child interactions and (2) strengthens children’s mental health, and social and cognitive development, after the intervention and at a 12-month follow-up. The secondary aims consist of determining the extent to which the intervention effects are moderated by (1) the implementation outcomes (fidelity and acceptability) and (2) child and family characteristics such as temperament, ethnicity, and socioeconomic background. In a subsample, we will also address (1) the characteristics of children with especially high cortisol levels during transition to childcare and (2) whether children in childcare units receiving the intervention have lower cortisol levels compared with controls. In another subsample we will qualitatively investigate parents’ and caregivers’ experience with the intervention.

The Thrive by Three intervention was piloted in a large community sample in 2 cohorts of 25/49 public (51%) and 24/49 private (49%) childcare centers from August 2016 to June 2018. A total of 49 centers and 243 groups of children aged 1-5 participated. The design was a simple pre–post design. The childcare centers were observed with Classroom Assessment Scoring System (CLASS; toddler and Pre-K), before the intervention started (pre) and after 10 months (post). Childcare staff filled out electronic questionnaires on background factors of the centers, the childcare classrooms, and themselves (ethnicity, education, staff turnover, attitudes toward the intervention). The pilot intervention differed from the current version of Thrive by Three intervention in 2 ways:

For toddler groups (ages 1-3) we found a small but significant increase in quality from pre to post in the Emotional and behavioral support domain (mean 5.97 [SD 0.41] vs mean 6.08 [SD 0.38]; t₁₀₃=–2.45; P=.02) but not in the Engaged support for learning domain (mean 3.24 [SD 0.76] vs mean 3.44 [0.77]; t₁₀₃=–1,86; P=.07). For the Pre-k groups (ages 3-5), there was a significant increase in quality from pre to post measures across all 3 domains: Emotional support (mean 6.06 [SD 0.56] vs mean 6.20 [SD 0.43]; t₁₂₇=–2.56; P=.01), Classroom organization (mean 5.27 [SD 0.76] vs mean 5.55 [SD 0.66]; t₁₂₇=–3.38; P=.001), and Instructional support (mean 2.67 [SD 0.84] vs mean 3.02 [SD 0.98]; t₁₂₇=–3.40; P=.001). Moreover, qualitative interviews with center leaders, head teachers, and staff from the first cohort (N=53) revealed great satisfaction with the quality-building framework and the holistic approach to quality building. Overall, they reported feeling more competent in their day-to-day interactions with the children and highlighted that the intervention had given all staff on different levels shared knowledge and a shared language to talk about quality with colleagues. There were, however, also several challenges. Most importantly, staff needed more time than predicted to fully comprehend the CLASS framework and content of the specific dimensions. The supervisors did not receive enough support and reported lacking a manual/written material to help structure the supervision throughout the intervention period.

This study is a clustered RCT delivered in childcare centers with toddler groups in 2 regions in Norway: southern and central. We selected a random sample using SPSS (IBM). The stages of enrollment, intervention, and assessment are presented in Table 1. A more detailed display of the different assessment time points is presented in Table 2.

Childcare settings with toddler groups in the Oslo region and Central Norway are eligible for participation. The participating children must be aged between 12 and 35 months. Childcare center eligibility is determined by region of the country and municipality. The primary study sites are in southeastern and central parts of Norway. The recruitment period started in February 2017 and lasted until August 2018. First, municipalities were recruited during the spring of 2017. Then, childcare centers within each municipality were recruited during spring and fall of 2017. Childcare staff consented during the fall of 2017. Children were recruited from fall of 2017 until August 2018. The long recruitment period of children is due to the fact that children under the age of 2 typically enter childcare centers for the first time in August, the year they turn 1. The lists of eligible children were therefore not complete until August 2018. Because of a high staff turnover, new staff members are invited to consent to participate throughout the year.

The intervention is offered during a complete year cycle in the participating centers. The control group is offered the intervention the following year cycle. Childcare quality is measured before, during, and after the intervention (3 time points). Parents and staff fill out questionnaires before and after the intervention. In addition, participating toddler groups and children will be measured 1 year after the intervention to identify possible long-term effects.

Children in the participating toddler groups are recruited via a written invitation to their parents given to them at the childcare center. Childcare staff keep in touch with the parents throughout the study period. The project team has no direct contact with parents. All information is given through the childcare center.

All childcare centers in the participating municipalities were allowed to participate as long as they had at least one toddler group, or groups with children aged 12–36 months. After the municipalities had consented to participate in the study, childcare centers within each municipality were recruited. Childcare centers volunteered to participate. All parents of children in the toddler groups of the participating childcare centers were invited to participate.

Childcare centers with no children under the age of 36 months were not eligible for the study.

This project represents a shared effort and close collaboration between the RBUP (East/South), The Master’s Program for Preschool Leadership at the Norwegian Business School (BI), The National Network for Infant Mental Health in Oslo, Regional Centre for Child and Youth Mental Health and Child Welfare (RKBU) Central Norway at the Norwegian University of Science and Technology (NTNU), and The Psychology Department at NTNU.

Professor Robert Pianta (University of Virginia), who led the team that developed CLASS, is a consultant for the research group. The national research team has extensive collaboration with Assistant Professor Gail E. Joseph and her associates. She is the director of the Early Childhood and Family Studies program at the University of Washington and the Principal Investigator and Director of the Childcare Quality and Early Learning Center for Research and Professional Development [49], which is the institute currently investigating effects of the broadband implementation of the quality building and quality control of childcare centers in Washington state. She is also coprincipal investigator and codirector of the National Center for Quality Teaching and Learning.

The data sets or other material used during this study are available from the principal investigator (Elisabet Solheim Buøen) on reasonable request.

Ethics approval for the study was given by Regional Committees for Medical and Health Research [40], reference number 2017/430 [49]. We anticipate low risk of harm for participating in the intervention since the work on employee process quality is based on literature and previous research that have shown to be positive for children’s development. In addition, education and guidance for employees is not likely to be harmful for children’s development. Consent was given written by municipalities, childcare centers, and parents. Childcare centers volunteered to participate. All parents of children in the toddler groups of the participating childcare centers were invited to participate and gave a written consent. Consent was given by childcare staff throughout the year.

As of August 2020, in total, 1531 children and 769 staff from 187 toddler groups have been recruited. Due to turnover, staff recruitment will be ongoing until August 2020. As of January 2020, the intervention group has been working with Thrive by Three for 1.5 years. Data collection at T1, T2, and T3 from both the intervention and control groups has been completed and T4 will be completed in August 2020. We expect to find an increase in childcare quality and better mental health and development outcomes for children in the intervention group. The reporting of these results will be made in accordance with the CONSORT guidelines [45]. The trial began recruitment in February 2017 and finished in August 2018. Data collection will finish in August 2020.

Changes to the project are made in the standard operating procedure. These changes are recorded and maintained by the principal investigator for RBUP East and South. Changes which are merely procedural but may impact the experience of the participants in the study are reported to the Regional Committees for Medical and Health Research ethics for approval.

Study participants are provided anonymous ID. A study key with the participants’ name and ID is stored in a separate, encrypted file on an internal server at RBUP East and South.

Dissemination will be done though scientific publications in well-regarded journals for child mental health and development as well as preventive science journals. Project newsletters and press releases, as well as 2 PhD theses will also be used to disseminate our results. Planned scientific publications include primary outcomes, secondary outcomes, fidelity, and the intervention and implementation. Publication of the data will be in accordance with CONSORT guidelines. The project team has adopted the Vancouver protocol for determination of authorship of scientific publications.

This study is an important contribution to the field regarding the quality of childcare settings in Norway. The results of the study will provide an indication as to whether or not the model for quality building, Thrive by Three, is effective to promote childcare quality within Norwegian childcare toddler groups. The study will also provide information about the effects of Thrive by Three on child development, mental health, and well-being. Finally, the results will provide insight into the implementation of a system for quality building in childcare centers and as such be of great value for Norwegian society, on individual, family, community, and national level.

As with any study, there may be potential limitations and challenges. One of the challenges we anticipate is high dropout from the control group. To counteract this, we are working to motivate the control group by offering them the intervention 1 year later as well as a cash incentive. Another limitation we anticipate is that because the intervention only lasts 1 year, the long-term effects of the intervention are not completely followed up. We do, however, aim to follow the children in the intervention group to see if the positive effects do persist even after the intervention.

The results of this study will offer a greater understanding of how high-quality childcare can be achieved and the effects of such care on children’s mental health. The study will present significant contributions to the field and present policymakers and the greater Norwegian society with important evidence on quality in childcare.