An increasing body of empirical evidence suggests that early intervention has positive outcomes for parents of children with neurodevelopmental disabilities (Guralnick 2017). Early intervention leads to an increase in developmental, social and functional outcomes for children (Dunst 2007; Guralnick 2017). Furthermore, there are numerous psychosocial benefits for parents, including an increase in parental empowerment, a decrease in parental stress and the improvement of parental self-efficacy (PSE) levels (Barlow, Coren & Stewart-Brown 2002). Consequently, lack of access to early intervention has been proposed as one explanation for why low- and middle-income (LAMI) countries have fallen short of effectively addressing Millennium Development Goals relating to child health (Samuels, Slemming & Balton 2012). In addition, many LAMI countries lack sufficiently skilled health practitioners to initiate and sustain such early interventions (Einfeld et al. 2012; Samuels et al. 2012).

Parents have an important role to play in a child’s psychosocial development (Kagan 1999). Consequently, a number of parenting interventions for families of young children with neurodevelopmental disabilities have been designed and evaluated globally over the past few decades (Kaminski et al. 2008; Salas & Cannon-Bowers 2001). These interventions are designed to improve a parent’s ability to successfully parent their children, through training, support or education, and the main goal is to influence the parent’s psychosocial well-being (Mejia, Calam & Sanders 2012). The majority of these programmes consist of skills training, parent education, parent support and/or parent coaching, and as a result they are said to be focused on the provision of knowledge (parent support) or techniques (parent-mediated intervention) (Bearss et al. 2015). The primary aims of these interventions are to reduce the impact of the challenges faced by the family of children with disabilities through teaching parents new knowledge and skills to reduce the child’s behavioural, emotional and developmental difficulties (Reichow et al. 2013). The methods of delivery of such training may include large seminar delivery, small group programmes and individual coaching sessions. The formats include telephone-assisted programmes, face-to-face programmes, self-directed programmes and online parenting programmes. The effectiveness of these programmes is not solely reliant on the delivery methods utilised, or content taught, but rather on the types of activities that are incorporated into the programmes (Kaminski et al. 2008; Woods et al. 2011). According to Kaminski et al. (2008), intervention teaching methods that included practising new skills with their own child and role play demonstrated the greatest effect size. Through these types of teaching activities, parents are taught intervention techniques that can be incorporated into their daily routines. This makes the impact of the intervention more sustainable compared to clinician-implemented interventions (Sanders & Kirby 2012; Strauss et al. 2013).

Researchers with a focus on the psychosocial development of children with developmental disorders indicate that PSE may have an important role to play in the development of a child (Coleman & Karraker 2003; Jones & Prinz 2005; Kendall & Bloomfield 2005; Montigny & Lacharité 2005). The PSE construct is primarily grounded in Bandura’s social-cognitive theory and has been defined as the belief in one’s own abilities to arrange and carry out tasks or actions to yield a specific achievement (Bandura 1977; 1989; 1997; Bandura & Walters 1977). A high level of PSE will cause parents to think and act in ways that will optimise the developmental outcomes of their children (Reichow et al. 2013). In other words, parents who face numerous stressors, but have high levels of PSE, are still able to facilitate positive developmental experiences for their children (Elder 1995). Consequently, developers of parenting interventions have paid considerable attention to the mechanisms whereby PSE beliefs can be enhanced (Bloomfield & Kendall 2007; Hudson et al. 2003; Jones & Prinz 2005; Sanders & Woolley 2005).

There are four primary methods in which self-efficacy can be modified (Bandura 1989). These methods serve to either enhance or decrease perceived levels of PSE. The first and most important method is that of enactive mastery (personal) experience. This results from prior accomplishment in certain activities. Enhancing PSE levels is thus achieved by allowing parents to experience success in situations that they previously found challenging (Bandura 1977). A second, likely method for improving personal self-efficacy is through the use of vicarious experiences. The individuals learn by observing challenging activities carried out by competent models, allowing them to re-evaluate their own mastery capabilities in relation to similar challenges they would encounter. It is especially useful when individuals see themselves as being similar to the observed model (Bandura 1997). Thus, having group discussions with other parents facing similar challenges, or watching videos or live parent models carrying out challenging tasks, are activities that may enhance PSE levels. A third mechanism to improve self-efficacy beliefs is the use of verbal and social persuasion, whereby others provide informed verbal feedback of an individual’s capabilities pertaining to a certain task (Bandura 1997; Woods et al. 2011). Encouragement from others is believed to be useful in improving self-efficacy and skill, whereas discouragement has the opposite effect (Bandura 1986). Within parenting programmes, feedback or coaching from the interventionists may provide this source of modification. The fourth way self-efficacy beliefs can be modified is through emotional and physiological arousal. Parents may experience stressful physiological responses that include increased stress, anxiety and/or fatigue, which make it harder to experience success (Bandura 1986). Therefore, reducing negative emotional arousal to subjective fears (through increased knowledge or skills, or access to necessary formal and informal support) would subsequently enhance performance and improve perceived self-efficacy (Bandura 1986). Figure 1 is a visual representation of common intervention activities and how they may influence self-efficacy beliefs. However, as programmes typically aim to decrease stress as an outcome for their intervention, based on the cumulative benefits of the other activities rather than the inclusion of a specific intervention activity (like mindfulness), this modifier is not included in Figure 1.

To our knowledge, there have been no systematic reviews of randomised controlled trials (RCTs) to assess the effects of parent training interventions on PSE for parents with young children that have neurodevelopmental disabilities. Through a systematic review of existing studies the primary objective was to assess the immediate change in PSE levels following parent training programmes for parents in the intervention arms of the included studies. The secondary objectives were to compare the change in PSE levels:

Ethics approval is not required for this study, given that systematic reviews draw on secondary publicly available data from published studies.

This systematic review found evidence for parent training programmes being effective in enhancing parental PSE levels. This finding was statistically significant and thus we are able to conclude that PSE is a robust parent outcome measure to evaluate the effectiveness of parenting programmes. Parental self-efficacy levels had a significant increase and large effect size (d = 0.60) for parents of children younger than 5 years of age, irrespective of the children’s diagnosis in the studies. Thus, data suggest that training parents of younger children are more beneficial in improving PSE outcomes than training initiated after the child is 5 years of age. The authors think that this may be because the skills taught to parents of younger children are based on developmental principles and consequently have a more direct impact on the developmental outcomes of children than skills taught to older parents. Parents who can see the positive impact that their newly acquired skill has on child outcomes would potentially be more likely to increase their belief (PSE) that they are able to provide the support that their child needs. These findings corroborate the increasing body of empirical evidence documenting the beneficial effects of early intervention on both parents’ and children’s outcomes (Guralnick 2017). These findings correspond to an earlier model that shows that heightened levels of PSE lead to subsequent heightened levels of success in the child (Ardelt & Eccles 2001). Thus, parenting programmes that increase PSE levels may also indirectly promote positive child outcomes (Ardelt & Eccles 2001; Coleman & Karraker 2003).

Parent training programmes were shown to be effective irrespective of whether they were administered by psychologists or other healthcare professionals. This finding may be of particular relevance in certain developing country contexts that do not have well-established professional training programmes for medical and allied health professionals and consequently may graduate a limited number of healthcare professionals on an annual basis. Task shifting has been suggested as a way to maximise access to interventions in contexts where there is a scarcity of trained professionals (Flisher et al. 2010; Rahman et al. 2008). In addition, there is an emerging body of evidence to suggest that alternative cadre professionals, such as rehabilitation care workers or community-based carers, are also able to effectively deliver parent training programmes (Flisher et al. 2010; Rahman et al. 2008; Reichow et al. 2013).

Finally, we wish to discuss the substantial amount of heterogeneity for the primary outcome measure. We employed the random-effects model throughout the analyses to account for this; however, in this meta-analysis, heterogeneity was particularly affected by one study. When removing the study by Whittingham et al. (2009), heterogeneity decreased considerably (the I² value decreased from 74% to 52%) and the effect size increased. Heterogeneity in this study may also have been attributable to the high risk of performance and detection bias present in this study. Alternatively, we propose that the high number of children with Asperger’s syndrome (12 out of 29) included in this study compared to the other included studies may have affected the heterogeneity. Characteristics of children with Asperger’s syndrome include relatively strong language abilities and milder difficulties with social interaction relative to children with a diagnosis of autism.

We used the risk of bias tool as per PRISMA recommendation (Moher et al. 2009). Areas of bias that were underreported included performance bias, detection and attrition bias, including allocation concealment. Authors should pay attention to how they report participant selection and randomisation procedures, as well as how they report incomplete outcome data. These biases should be carefully considered in the design and implementation of future RCTs involving parent training programmes.

While there have been systematic reviews supporting the effectiveness of parent training programmes for parents of children with neurodevelopmental disorders, such as Skotarczak and Lee (2015) as well as Tellegen and Sanders (2013), this review is the first to evaluate the effect these interventions have in changing the PSE levels. No language limitations were set and articles not written in English were translated and included if they met the inclusion criteria. Furthermore, when investigating parent training programmes we chose to include all forms of parent training, rather than selecting specific programmes as other systematic reviews, such as Tellegen and Sanders (2013), have previously done. It is interesting to note that non-licensed interventions were ineffective in enhancing PSE levels. The authors postulate that licenced interventions have undergone a more rigorous development process than non-licensed interventions. This may result in stronger theoretical underpinning relating to both the development of the content and intervention activities, as well as a more detailed process of stakeholder engagement. This finding creates an interesting tension for researchers in LAMI settings interested in designing parent training programmes. Licensed interventions may be better at enhancing PSE, but they are not always contextually relevant and may need to be adapted to be socially acceptable in developing country contexts.

As researchers within an African context, we recognise the need to pilot the efficacy of parenting interventions to change PSE levels in a LAMI context since, by the middle of this century, 40% of the world’s population of children will live in Africa (You et al. 2014). It is well known that Africa, as a continent, has limited access to resources and services to promote the health and development of its children. Therefore, it is important to consider how to reach the families of children with neurodevelopmental disabilities in these resource-constrained contexts. Consequently, we suggest that future research builds on this evidence base, which indicates that parents can be effectively trained by psychologists and allied health practitioners, by examining the effects of parent training provided by alternative cadre professionals.