A scoping review of mental health prevention and intervention initiatives for infants and preschoolers at risk for socio-emotional difficulties

Infant mental health (IMH) has emerged as a unique area of practice and research distinguished from child and youth sub-specialties by its advocacy for a relational practice framework with an emphasis on parents/caregivers being integral to assessment, treatment, and prevention initiatives [1]. Interest in IMH was driven initially by clinical research in the areas of childhood attachment, socialization, and development [2‐6] and by large-scale US-based social initiatives of the 1960s and 1980s, such as Head Start, that focused public attention on the merits of prevention-based programs to invest in infant well-being. Attachment theorists and proponents of pre-school programming for at-risk families [7] have helped to drive research forward and solidify the field’s knowledge-base as to the effectiveness of various programs for positive developmental trajectories. In addition, advances in assessment and treatment as well as longitudinal studies over the last 30 years have demonstrated that psychopathology which results from the combination of environmental and genetic risk factors can be reliably identified in preschoolers [1]. Therefore, infancy and early childhood are key developmental periods during which precursors to significant and lasting mental disorders may emerge, while at the same time representing key timeframes for prevention and early intervention.

The study of risk factors for early-onset mental health issues has evolved to recognize the dynamic interaction between individual genetics and temperament with the child’s environment in shaping developmental outcomes. The absence of secure parent-child attachment and the presence of adverse or traumatic experiences, especially during sensitive and/or critical periods of development are believed to be key to the onset and continuance of early childhood emotional and/or behavioral issues [8, 9]. This age group is also considered to be most at risk for experiencing abuse and/or neglect [10, 11]. Other specific risk factors associated with the presence of mental health difficulties in infants and preschool children include maternal depression, parental substance abuse, family violence, limited parental education, poverty, and neighborhood safety issues [12, 13].

Early childhood behavioral problems, including disruptive behaviors, oppositional defiance, and/or aggression, represent one of the most common referrals for mental health intervention [14, 15] with anxiety disorders being the second most frequent concern [16, 17]. Prevalence rates of preschool psychopathology vary between 7 and 28% [18] and show developmental continuity with later childhood and adolescent psychopathology [14]. In Early Childhood Education (ECE) settings, estimates suggest that as many as 30% of children require special attention within their programs due to emotional regulation or behavioral difficulties [19]. While some risk factors may be transient and not affect the child’s developmental trajectory, the cumulative interaction of environmental and genetic risk factors in some children may lead to enduring problems without intervention [1].

Research on the effectiveness of infant and preschool psychosocial interventions and prevention initiatives has grown significantly since the emergence of IMH as a distinct sub-speciality in the late 1970s. Treatment initiatives for children diagnosed with mental disorders and/or those with emergent difficulties typically involve attempts to alter the child’s deviant behavior or emotional dysregulation, and/or negative interactions within the infant-caregiver dyad, either through direct work with this dyad or indirectly through parent education programming. Regardless, in most cases, treatment for IMH issues considers the parent-child relationship as the principal mechanism for change. Although the parent-child relationship is typically the primary focus of assessment, intervention, and prevention initiatives, the functioning of the child’s broader contexts, such as family and community factors, are also viewed as important influences shaping developmental trajectories and are thus included in treatment interventions and prevention initiatives [1].

Due to the shifting influence of risk factors during this developmental period, conceptual frameworks for infant and child mental health programming must include both prevention and intervention initiatives. One such framework, commonly used in public health reporting [1, 20, 21], categorizes prevention initiatives according to the level of risk targeted by the intervention and/or the degree with which a particular health problem may be experienced by each social stratum within the population under examination. Initiatives are sub-classified as programs that are universal, selective, indicated, and direct treatment for children diagnosed with and/or experiencing mental disorders. Universal programs are offered to the broadest range of infants, preschoolers, and families, and are considered beneficial regardless of the presence of unique risk factors, problems, or need for professional intervention (e.g., state-wide/provincial subsidized high quality daycare). There is a growing body of selective prevention initiatives targeting groups deemed at-risk for future developmental outcomes. The best researched of these initiatives is the federally funded US-based Head Start programs, which provide comprehensive early childhood education, health care, nutrition, and parenting programming for groups considered at-risk due to systemic poverty. Indicated programs are provided to children demonstrating sub-clinical problems of recognizable difficulties and/or developmental issues that if left unaddressed, will likely develop into full-scale syndromes or disorders requiring professional treatment interventions [22, 23]. A further category is direct treatments by mental health professionals to children diagnosed with mental disorders as defined by mostly categorical criteria [1].

This diverse array of interventions offered across a broad spectrum of delivery models, raises several questions including: What populations are serviced by the various approaches? What are the aims of these interventions? In what settings are these interventions conducted? What research methods, including measurement tools and/or data-collection are used? To date, a large-scale mapping of the landscape of the research regarding infant and pre-school early interventions and prevention initiatives has not yet been conducted. Therefore, in order to answer these questions and determine the composition of the research landscape, a scoping review was conducted [24, 25]. A scoping review is a process of mapping key concepts, the main sources, and types of evidence available as well as gaps in a research area especially where an area is complex or not previously reviewed in a comprehensive manner [26]. Our objective therefore is to determine the scope of the mental health intervention and prevention research pertaining to infants and preschoolers (0–5 years), and their families at-risk for socio-emotional difficulties and negative developmental outcomes.

The methods for this scoping review were informed by the six-step procedure outlined by Daudt, van Mossel, and Scott [27], an extension of Arksey and O’Malley’s [24] approach and reported following PRISMA statement extension for scoping reviews [28] (Additional file 1). Our team’s “interpretation of how the consultation is achieved” [27, p. 7] was to adopt an integrated knowledge translation and exchange (iKTE) approach, which involved consulting with our knowledge users (KU) through each step of the review procedure.

For the initial step, to define the research questions, we worked closely with the project’s KU group, comprised of professionals representing social services, tertiary health care, public health, and early childhood education. Together, we refined the parameters of our scoping review by posing three research questions: (a) What intervention and/or prevention programs currently exist for children ages 0–5 at risk for mental health difficulties, and their caregivers and families? (b) What populations participated in these interventions and preventions programs? (c) How are these programs evaluated? In order to transform these questions into searchable queries consistent with research database requirements, we worked with an information scientist to define our key constructs including “at risk for mental health,” “intervention and prevention programs,” “caregivers,” and “families.” We defined “at risk for mental health” as an infant or toddler displaying deviant behavior and/or emotional patterns, interpersonal difficulties, or those demonstrating psychopathology, or diagnosed with a disorder, or exposed to known risk factors for the same, including maternal mental health issues, caregiver substance abuse, poverty, and residing in a community with safety concerns. Caregivers were defined as a parent and/or a primary caregiver responsible for providing day-to-day care and the emotional developmental needs of the child. Family was defined as including the primary caregiver/parent as well as siblings and relatives or those considered family residing in same family home and/or taking on responsibility for caregiving of the preschooler. Intervention and prevention programs were defined as any or all models, services, strategies, and/or techniques provided by professional, para-professional, or lay person purported to address, remediate, accommodate, offset, or reduce the chances of onset or continuance for mental health difficulties or disorders, behavioral or emotional deviance, or developmental issues.

In the second stage of our review, we endeavored to identify all relevant studies. Conducted in collaboration with our information scientist, we engaged a systematic search of all relevant online research databases, as well as used snowball search methods and reference tracking (i.e., checking reference lists of included sources and checking database alerts) to identify additional articles. Database searches included MEDLINE, PsycINFO, EMBASE, Educational Resources Information Centre (ERIC), Clinicaltrials.​gov, Cumulative Index to Nursing and Allied Health Literature (CINAHL), The Cochrane Library, including the Cochrane Central Register of Controlled Trials (CENTRAL), ProQuest Nursing & Allied Health Source, Latin American and Caribbean Literature on the Health Sciences (LILACS), Web of Science, and World Cat. Key search terms, including synonyms and medical subject headings (MeSH terms), were entered into these databases in systematic manner by a library scientist. Search parameters for dates of publication included all eligible studies published from inception through December 31 of 2012. Additional file 2 contains the search terms employed for Medline Ovid. Medline was searched first because it was considered the most relevant for our study’s objectives. Terms were subsequently modified as required for each search of the various databases engaged in the current review. We also conducted a search of the gray literature via the following portals: Canadian Institutes of Health Research; Centre of Excellence for Early Childhood Development; CMA Infobase; Institute for Research on Public Policy; International Network for Early Child Development; International Organization for Early Intervention; National Guidelines Clearinghouse; Offord Centre for Children at Risk; World Health Organization; World Association for Infant Mental Health. We also conducted a hand search of the table of contents for the Infant Mental Health Journal, Child Development, the Journal of the American Academy of Child and Adolescent Psychiatry, and the Journal of Child Psychology and Psychiatry for articles published between 2008 and 2012.

The third phase of the review, study selection, began with two reviewers independently screening titles and abstracts against specified inclusion and exclusion criteria to determine suitability for inclusion in the study. Duplicate citations were removed automatically from our RefWorks library. During the next stage of screening, the same reviewers independently screened the full text articles deemed relevant in the first stage of citation screening. Disagreements regarding study inclusion during the title/abstract or full text review processes were resolved by a senior member of the research team (N.C.) if consensus could not be achieved through discussion. This process was required for 20 articles during the title/abstract review phase for which consensus was not achieved, which represents an inter-rater reliability for inclusion of 98.4%. Consistent with our published scoping review research protocol [29], we considered for inclusion any/all studies demonstrating a clearly articulated and research methodology, including quantitative (i.e., randomized controlled trials, quasi-experimental designs, single group pre-post with and without follow-up, case-control), qualitative (case studies, phenomenology, ethnography, grounded theory), or mixed methodologies. We excluded review articles or studies presenting filtered information (i.e., systematic reviews, scoping reviews, evidence syntheses, narrative reviews, qualitative syntheses). Qualitative studies were also included where rigorous qualitative methodologies were clearly reported, including those adhering to a case study methodology. However, papers offering clinical scenarios, vignettes, case descriptions, or clinical examples that failed to report a research methods consistent with qualitative case study methods were excluded. Specific a priori inclusion criteria required that all included studies were (a) primarily related to the examination of and/or reporting on psychosocial interventions involving at-risk preschooler; (b) parents demonstrating risk factors or adversity (e.g., depression, substance abuse), or the parent-child relationship demonstrating risk factors or adversity (e.g., negative parenting practices, abuse); (c) a reliable assessment of baseline mental health diagnosis or risk factors and at least one child-related mental health outcome; (d) a child sample/population at risk for socio-emotional difficulties (e.g., anxiety, depression, aggression); and (e) a child sample/population 0–5 years of age, or a sample/population of parents of children 0–5 years of age, or the sample/population being reported upon is the long-term follow-up of a population that was 0–5 years of age in the original study. Studies were excluded if they (a) primarily examined autism (autism spectrum disorder) or developmental disorders including intellectual or emotional impairment (i.e., mental retardation, severe fetal alcohol syndrome), language/communication disorders (i.e., receptive, expressive disorders); (b) studies focusing primarily on assessment with no intervention component and no outcome measures; and (c) generic parenting groups not targeting change in the parent-child relationship and/or the child’s current or future socio-emotional/developmental outcomes.

In the fourth stage of the review, we charted data, first by developing an extraction tool used in combination with an Excel spreadsheet to systematically record pertinent information for included studies. Categories included in the data extraction tool and spreadsheet were as follows: (a) publication year, (b) author, (c) location of study (country), (d) setting for study (i.e., clinic, community, home), (e) population targeted, (f) age of participants, (g) type of intervention, (h) outcome measures used, (i) study outcomes, (j) duration of intervention, (k) number of participants, (l) cost of program, (m) professional background of treatment provider, (n) methodology for study, and (o) whether the study was a replication. Consistent with Levac and colleagues’ [25] suggestion, two senior members (N.C. and A.M.) of the research team applied the extraction tool to the same five studies to determine the inter-agreement of our processes. Following this process, two reviewers used this tool to independently extract the data from each study included in this scoping review with discussions at regular intervals to ensure coding accuracy. As an extra degree of inter-agreement, a senior member of the research team (A.M.) reviewed 10% of completed extractions and compared these to the original full text articles to further verify the inter-agreement of our extraction process. No concerns were found with the extraction process.

The fifth stage of the review, known as the collating, summarizing, and reporting phase, was the most intensive stage of the study where we analyzed the data, reported the results, and applied meaning to these results [25]. To efficiently and effectively conduct the analysis, we examined the data-extraction chart corresponding to each study included in the review. Two members of the research team (A.M. & R.C.) analyzed this material to identify key themes and/or commonalities and differences between and across studies.

The sixth stage, the consultation process, involved an iKTE process whereby project researchers collaborated closely with project KUs in order to ensure the relevance and usefulness of the final product. Although this is described as the final stage in the review process, it actually occurred throughout each stage. We collaborated with our KU partners through face-to-face meetings, teleconferences, and email document exchange to refine the research questions, finalize terms for the search strategy, identify important data extraction elements, and interpret findings.

We identified 533 potential studies from 1233 title and abstracts after the first round of screening. Full text article review in the second round of screening resulted in a total of 162 included articles for the final analysis. Figure 1 shows the study selection procedure in PRISMA format. A table presenting key extracted data from all 162 articles can be found in an online appendix (Additional file 3), which also includes key extracted data from 23 studies not referred to elsewhere in this manuscript.

The primary strength of this scoping review is its comprehensive nature, including 162 articles following the full-text assessment of 532 studies; the only review of its type ever conducted in relation the 0–5 population. Due to its breadth, we were able to achieve the goals of the scoping review methodology as articulated by Arksey and O’Malley [24]. That is to examine and report on the extent, range, and nature of mental health-related interventions and prevention initiatives for children 0–5 and their families. We comprehensively examined trends in publication dates, geographic locations, institutional settings of the research, research methodologies employed, level of the intervention within a public health framework, key therapeutic mechanisms underpinning the interventions, and the outcome measures utilized.

Through this review, we were able identify key gaps in the early years mental health intervention literature, including the need for future research from non-Western countries, better definitions of risk factors and associated outcomes, and the role of fathers’ involvement in IMH initiatives. This scoping review was able to examine and describe the intervention literature within the parameters of an accepted model of public health [1, 20, 21] with an overlapping “nested” view of prevention and intervention. As such, both the current scoping review and the public health framework chosen are heuristic concepts which reflect the dual realities of prevention and intervention continua in the early years, but the model is more descriptive rather than predictive. More research on risk factors, therapeutic mechanisms, and outcomes is needed to separate out children and families with differing trajectories or “developmental assets” in order to match need to risk level and build more comprehensive “ecologically valid” intervention models.