Introduction

The period from conception to the age of 3 years is characterized by the most rapid growth of mental and socioemotional capacities [47], mirroring the rapid development of brain architecture [39]. Healthy child development is an outcome of a continuous ongoing process, where the needs of the child are met, and the child acquires the skills and abilities needed to reach his or her full potential later in life [49]. Protective factors and interventions early in the developmental course have the greatest positive impact and are the most cost-effective compared to those later in life [48]. Additionally, it has been shown that the largest financial returns and the greatest effects are generated by policies focusing on delivering effective interventions to the most vulnerable populations [48].

In Slovakia, Roma living in the marginalized communities make up one of the most vulnerable groups threatened by multiple forms of disadvantage. “Marginalized communities are separated or segregated communities, excluded from mainstream social, economic, educational and cultural life [44]. The separated type refers to a Roma population concentrated in a certain part of a town or village – either inside or on the outskirts; the segregated type refers to a settlement that is remote from towns and villages or separated from them by a physical barrier [17].

Roma are the largest ethnic minority in Europe [9]. “Approximately 440 thousand Roma are living in Slovakia, of whom 53.5% live in separated or segregated communities. The rest lives scattered among the majority population [35].” People living in marginalized Roma communities are most often disadvantaged by lower education, unemployment, receiving social benefits, struggling with bills and lacking standard household facilities (sewage system, water supply, flush toilet, bathroom or shower, electricity) [31]. Such contextual characteristics can endanger healthy development in early childhood [13, 52].

The limit available evidence shows that the unfavourable conditions in early childhood are reflected in the poorer health status of Roma children [40]. This evidence shows that disparities in health begin early, with perinatal and infant mortality rates significantly higher in districts with a higher proportion of the population living in marginalized Roma communities [41]. Additionally, Roma children constitute 24.2% of all cases of sudden infant death syndrome in Slovakia [28], i.e., much more than the estimated share of Roma children in the general population, i.e., 13.3% [54]. Moreover, Roma children suffer more often from infectious diseases, injuries, poisoning, burns, respiratory diseases and chronic diseases than other children [2].

In addition to the above-mentioned contextual characteristics of the environment, poor access to health care also contributes to the poorer health status of Roma children. Barriers in access to health care, such as lack of funds for travel costs or pharmaceuticals, bad travel connections, bad previous experiences, fear or distrust [25], result in 36% lower health care use among marginalized Roma compared to the general population [34]. Therefore, Roma not only have poorer health but also use the health care system to a lesser extent, thus increasing their health disadvantage. They experience similar barriers to other health care services and to education, including early diagnosis and intervention [19, 34, 53]. In addition, Roma children younger than 3 years of age from these marginalized communities have poor access to early childhood health and educational programmes. Only a few early childhood education programmes are available for Roma, available only in a few Roma communities, and these are almost exclusively offered by non-governmental organizations [37].

The differences in health between Roma and non-Roma are well-documented, but evidence is still lacking on how to increase the chances for healthy early development of children in marginalized Roma communities [19, 53]. Therefore, the aim of our study was to assess the perceptions of professionals from different fields working with marginalized Roma communities on potential measures (i.e. interventions and policy changes) designed to improve the healthy early childhood development of children living in marginalized Roma communities and to identify priority measures on the basis of urgency and feasibility, as judged by these professionals.

Methods

Design

We used a group concept mapping (GCM) approach to structure the interventions aimed at increasing the chances for healthy early childhood development, as proposed by professionals who in their praxis are directly or indirectly involved with Roma living in marginalized communities. Group concept mapping is a participative research method comprising a mix of qualitative data collection and interpretation and quantitative data analysis [26].

The design of our study was informed by the theoretical Biodevelopmental framework for understanding the origins of disparities in learning, behaviour and health [46]. Within this framework, the foundations of healthy development and sources of early adversity are described on several levels i.e. the environment of relationships, the physical, chemical and built environments, and nutrition. This framework was elaborated by Shonkoff [46] to inform policy targeted at young children, particularly the most vulnerable ones. Based on this framework, we aimed to recruit professionals from different backgrounds with a deep understanding of the various determinants influencing healthy development and the sources of early adversities. (See Sample below.) In addition, we used the Biodevelopmental framework for structuring the data and interpreting the results.

Sample

Purposive sampling techniques were used to recruit professionals working with marginalized Roma communities from Slovakia from both public and non-governmental sectors of different levels of work hierarchy across these categories: health care providers, social workers, community centre workers, early childhood educators, special educators, health mediators, experts in early childhood development and policymakers. Since the expert field of early childhood education and care is relatively small in Slovakia, we approached professionals with whom we had previously built rapport or professionals recommended by other professionals. We initially addressed 79 professionals, 54 of whom agreed to take part in the study (response rate of 69.2%).

The final sample for the brainstorming step consisted of 54 participants and for the sorting/rating step 40 participants for the GCM procedure, see Procedure below. “We lost some participants between the brainstorming and sorting/rating due to the fact that some of them decided not to partake in the next steps because of their work overload. This aligns with the GCM methodology, which takes into account some losses of participants between the brainstorming and sorting-rating step without bias being likely [24]. The sample size in each step of this study was sufficient to meet the statistical requirements for obtaining valid and reliable results [24].

Procedure

We applied the general GCM procedure consisting of five steps: preparation, brainstorming, sorting and rating, analysis, and interpretation [27, 43]. Figure 1 illustrates this process and the responsibilities of both the researchers and the participants.

Fig. 1
figure 1

The Basic GCM Process (responsibility of P = participants, R = researchers) ([26], adapted)

Full size image

In the first step, the preparation phase, we formulated the research question, also known as the focus prompt: “What needs to be done to make the chances for healthy early childhood development of children from marginalized Roma communities equal with the chances of children from the majority population? The following explanations were given to the participants in order to make sure, that the focus prompt is clear to all participants: “By ‘early childhood’ we mean the period from conception to 3 years of age. By ‘healthy development’ we mean development that enables a child to reach his or her full potential in later life. This means healthy, age-appropriate growth and development of cognitive, social, emotional, language and motor skills. By ‘making chances equal’ we mean enhancing the starting positions similarly, which we know is worse compared to majority population. Roma children are disadvantaged when entering school and in their further life.”

The potential participants were identified in line with the Biodevelopmental framework [46] and contacted, and the schedule of the project was set. We decided to conduct the study online, using conference calls and the groupwisdom™ platform (https://groupwisdom.com/, a platform where each step of the GCM can be performed online). Prior to the data collection, we organized a conference call attended by the research team and the participants, where the aim of the study and the GCM procedure was explained and discussed.

In the second step, we organized the brainstorming phase using the online groupwisdom™ platform. Participants were asked to sign the consent form first and then to answer the focus prompt. In order to protect the participants from potential power-relations interplaying during the brainstorming part of the study, this phase was made anonymous. Due to the anonymous nature of this step, no demographic data was collected at that moment. This is in line with the GCM methodology, as these questions are most frequently asked during the sorting and rating phase. Participants were encouraged to generate as many statements as needed. To prepare a master list (a final list of statements) for the subsequent sorting-rating phase, we performed a synthesis of the statements and a qualitative review. We merged semantically similar statements and split those which contained more than a single concept with the aim of removing redundant and overlapping concepts and creating a reduced, concise set of statements. We further removed statements according to these criteria: statements not answering the focal question, statements not possible to handle by policy (e.g. containing rather personal issues such as family support) and statements describing a problem rather than a solution. The master list containing the final set of items was once again sent to the participants for commenting and clarification.

In the third step, the sorting-rating phase, we asked the participants to sort the statements into groups or piles of similarly themed statements and to create a descriptive label for each group based on what they view as the unifying topic or content of each group. We then asked them to rate these statements according to two selected domains of interest, i.e., urgency and feasibility, on a 7-point Likert scale (1 – not urgent/not feasible, 7 – very urgent/very feasible). Sorting and rating were performed using the groupwisdom™ platform.

In the fourth step, the analytic phase, we made the final data categorization. A quality review was performed to check if all participants followed the sorting and rating guidelines (rated at least 75% of the task or did not give negligent answers) and analysed the data using the groupwisdom™ platform. The findings of the analyses were discussed within the research team, and the resulting 4-cluster solution was chosen based on the most consistent cluster map. Cluster labels were discussed and decided on in an expert group consisting of 6 researchers, and subtopics were identified in each of the clusters.

In the interpretation phase, the final cluster solution and cluster labels proposed by the expert group were sent to the participants to gain their feedback on these results and their interpretation of the resulting maps.

Statistical analysis and reporting

As data handling and data analysis occur simultaneously in GCM, this paragraph discusses them both in their sequence as applied, with some initial data handling already presented in the previously described steps 2 and 4. First, we performed a descriptive analysis to describe the background characteristic of our sorting-rating sample.

Second, we categorized using cluster analysis the perceptions of professionals from different fields working with marginalized Roma communities regarding potential measures to improve the healthy early childhood development of Roma children. In that analysis we clustered the generated statements into a cluster point map using a multidimensional scaling and hierarchical cluster analysis [27]. In this map, a point represents a single statement, the distance between the points indicates how often participants sorted particular statements into the same group, and the size of the cluster indicates the degree to which its various contributing statements are related. The expert group computed the stress index for a varying number of clusters (3–13, i.e. the highest and the lowest desired number of clusters, as sorted by participants) and used bridging/anchoring analysis and spanning analysis. The first analysis shows the relationship of the statement to its location on the map; the latter one visualises the statement’s strength of connection to every other item on the map [26]. The expert group finally proposed a 4-cluster solution.

Third, to identify priority clusters according to urgency and feasibility, we generated cluster rating maps, where more layers indicate more urgency and feasibility, respectively. To identify separate priority measures, we produced a Go-Zone map, i.e., an X–Y graph which compares items across two rating criteria and is divided into quadrants above and below the mean value of each rating variable [8]. We checked the model fit using the stress-index i.e., the degree to which the distances on the map are discrepant from the values in the input similarity matrix; a high stress-index value indicates a greater discrepancy (i.e., the map does not represent the input data well) [27, 29]. Within a GCM context, stress-index values should range from 0.10 to 0.35, with lower values indicating a better fit [20]. All analyses were performed using the groupwisdom™ software (https://groupwisdom.com/).

Results

Background characteristics of the sample

The sorting-rating sample, consisted of 31 women and 9 men, with a mean age of 42.5 years (minimum = 27, maximum = 59; years of age were recorded in rounded numbers). Out of all the participants, 33 had completed university education (82.5%), 6 had completed secondary education (15.0%) and 1 had completed elementary education (2.5%). Direct contact with marginalized Roma in their work was reported by 34 participants (85.0%); 6 participants (15.0%) did not have direct contact with marginalized Roma in their work. These 6 participants were from higher levels of the work hierarchy and despite not working directly with the Roma, their expertise was relevant to the research question (e.g. they focus on policy making targeted at inclusive policies). The mean period spent in the current work position was 11.2 years (minimum = 0.5 years, maximum = 30.0 years). Participants who have worked in their current position only for a short time had previous experience working in related fields.

Final cluster solution of proposed measures

Participants generated 178 proposals for measures in the brainstorming phase. After review and synthesis of the statements, we created a master list containing a final set of 90 proposed measures. Participants sorted the 90 proposed measures into 3 to 13 groups. The expert group consisting of 6 researchers chose and approved the final 4-cluster solution. The expert group agreed upon the final cluster labels and topics identified within each cluster, which were as follows.

Cluster 1 represents Public resources for instrumental support and contains two topics, Financial and institutional frameworks and Tools for instrumental support. This cluster contained measures reducing barriers in access to services, ensuring funding to programmes targeting children living in marginalized Roma communities, and providing basic equipment and adequate nutrition. An example of a measure from Cluster one is “Guarantee the legal right to access early childhood care services for all children at risk (health, social).” Cluster 2 represents Enhancement of living conditions and contains Access to income, Access to housing and Access to basic infrastructure. In the second cluster, basic needs, such as housing, a healthy living environment and employment for parents, were targeted. An example of a measure from this cluster is “Negotiate with mayors about enabling access to drinking water, heat, garbage collection and so on.” Cluster 3 represents Quality and accessibility of health care and contains topics related to Prenatal care, Perinatal care, Postnatal care, Paediatric care, Reproductive health and Field health care. An example of a measure from Cluster 3 is “Implement a system of field paediatric and nursing care.” Cluster 4 represents Community interventions focused on the transfer of cultural capital and contains answers to the questions Who should be educated, By whom, Where should this education take place, How should the education occur and What should be the content of the education? An example of a measure from this cluster is “To support the establishment of maternity centres with a library, toys, sports opportunities for future mothers but also for mothers with children.”

The stress index was 0.1916, which suggests a strong fit between the cluster map and the data, as the value should range from 0.10 to 0.35, with lower values indicating a better fit [20]. The final cluster solution is shown in Fig. 2. All proposed measures and their groups can be seen in the 15.

Fig. 2
figure 2

Cluster point map – final 4-cluster solution regarding the proposed measures. Note: Each point represents a measure suggested by participants. The distance between the points indicates how often participants sorted particular measures into the same group (a smaller distance from the centre of a cluster means that more participants placed those particular measures into the same group). The size of the cluster indicates the degree to which its various contributing measures are related

Full size image

Priority measures based on urgency and feasibility

Regarding urgency and feasibility of the clusters, participants considered Cluster 2, i.e. Enhancement of living conditions as the most urgent. Cluster 3, i.e. Quality and accessibility of health care was considered to be the least urgent. In terms of feasibility, the results were the opposite. Cluster 2, which was considered to be the most urgent cluster, was rated as the least feasible one. Cluster 4 had the highest feasibility, i.e. Community interventions focused on the transfer of cultural capital, despite being rated as one of the least urgent. The urgency and feasibility of the various clusters as rated by participants are shown in Fig. 3.

Fig. 3
figure 3

Cluster rating maps regarding urgency and feasibility of the proposed measures. Note: More layers indicate more urgency and feasibility, respectively. The clusters with fewer layers were rated as less urgent (upper map) or less feasible (lower map). The clusters with more layers were rated as more urgent (upper map) or more feasible (lower map)

Full size image

Regarding urgency and feasibility of separate measures, 27 of the 90 proposed measures were rated as highly urgent and highly feasible, i.e., according to the respondents they should be implemented with a priority in order to equalize the chances of Roma children having healthy development. The identified priority measures according to urgency and feasibility are shown in the Go-Zone map (Fig. 4). All four clusters were represented in these separate measures.

Fig. 4
figure 4

Urgency and feasibility of the proposed measures combined into a “Go-Zone map”. Note: The priority measures rated as the most urgent and most feasible are placed in the green sector in the upper right corner. The colour of a point represents the cluster to which the measure belongs: red—Cluster 1, lime—Cluster 2, orange—Cluster 3, dark green—Cluster 4

Full size image

Most priority measures belong to Cluster 4 (Community interventions focused on the transfer of cultural capital). Only one of the priority measures belongs to Cluster 1 (Public resources for instrumental support). Participants considered the items “Education and raising awareness of adolescents about reproductive health in schools in cooperation with Roma health mediators” (item 64) and “Education and raising awareness of adolescents about reproductive health within community centres” (item 65) to be the most urgent and feasible. In contrast, the least urgent and feasible was to “Build a wide network of breast milk banks” (item 77).

Discussion

We assessed the perceptions of professionals from different fields working with marginalized Roma communities and identified 90 measures they considered to be needed for healthy early childhood development in children from marginalized Roma communities. These measures were grouped into four clusters, and rated by urgency and feasibility as perceived by participants. Participants identified 27 measures as having priority, i.e., being highly urgent and feasible.

We identified four clusters of measures which mostly align with the sectoral objectives of the EU Roma strategic framework for equality, inclusion and participation [15]. The objectives of this framework regard Housing, Employment, Education, and Health. In the clusters that we identified, Housing and Employment are combined into Enhancement of living conditions, Education is the main topic of Community interventions focused on the transfer of cultural capital, and issues of Health are addressed within Quality and accessibility of health care. Compared to the framework objectives, we identified one additional cluster, i.e., Public resources for instrumental support, which includes measures overarching the other three clusters from a policy perspective, enabling access to and strengthening the capacities of the system of care. The grouping of the overarching measures into a separate cluster and not integrated into respective clusters may be interpreted as the participants perceiving the current lack of intersectoral strategies and collaboration in Slovakia as a barrier to the implementation of measures included in other clusters. Slovakia lacks an integrated system of education and care for children aged 0–6 [34]. Such an overarching objective may also add to the EU framework.

Participants perceived the cluster concerning Enhancement of living conditions as the most urgent, which aligns with the materialist framework of how social determinants shape health and health outcomes [7]. Within the materialist framework, living conditions are reflected in three key mechanisms influencing health and health outcomes: (1) experience of the material living conditions, (2) psychosocial stress caused by these conditions, and (3) adoption of health-threatening or health-supporting behaviours [7]. The particular measures in the mentioned cluster address living conditions, housing and employment of parents. Their importance is supported by research showing that housing instability and food insecurity do indeed negatively affect health and development in early childhood, with long-term effects in later life [10, 30, 45]. The proposed measures in this cluster thus require action from ministries, such as the Ministry of Labour, Social Affairs and Family, and from municipalities, with some of them also requiring inter-sectoral cooperation.

The cluster concerning Quality and accessibility of health care was considered to be the least urgent compared to the other clusters. This seems to contrast the relatively poor quality and accessibility of health care in Slovakia, which are among the lowest of all countries of European Union [16]. Moreover, Roma in Slovakia face significant barriers in access to health care [4, 5]. Hence, an explanation for the low urgency of this cluster may simply be that other clusters contained more urgent issues, or that the interviewed professionals perceive not to need additional policy measures, as they can address this issue themselves. This definitely requires further study.

The cluster of measures regarding community interventions focused on the transfer of cultural capital was rated as the most feasible, reflecting the widespread use of community interventions in marginalized communities in Slovakia and across Europe [1]. Currently, several early childhood education and care (ECEC) projects are being realized by NGOs in Slovakia and have been successful on a local level but would need to be scaled up to the national level [55]. Some of the interventions and measures proposed in this cluster may be considered to be an extension of services provided on a community level (such as Project OMAMA, People in Need Slovakia, Healthy Communities) [1]. An explanation may be that participants viewed these measures as highly feasible since such interventions already exist, have been shown to be effective and have received public approval. Evidently, this supports further research and development to implement adequate Roma-centred community interventions.

The cluster concerning enhancement of living conditions was considered as the least feasible despite being rated as the most urgent, which aligns with the fact that improving the living conditions of marginalized Roma communities in Slovakia has been identified by the EU as a main area requiring EU funding. The spending rate of these EU funds has remained low, possibly due to the resistance of municipalities to use the corresponding schemes [14]. The lack of political will to enhance the living conditions of Roma [42] reflects the public discourse related to poor people’s responsibility for bettering their own living conditions. In this discourse, Roma are labelled as less deserving of public support and consciously abusing the social system [12]. This suggests our participants do not see the political will of municipalities to allocate funds and services for marginalized Roma and to enhance their living conditions. As the measures in this cluster align with the rights guaranteed by Convention on the Rights of the Child [51], which Slovakia signed in 1991, reluctance to implement such measures may be considered a violation of children’s rights.

We also found a significant discrepancy between urgency and feasibility in the cluster related to community interventions focused on the transfer of cultural capital, which was rated as the most feasible but also as one of the least urgent. One of the explanations for the low urgency of this cluster could be that many of the proposed measures already exist and just need to be scaled up. The low appreciation for urgency may also be due to the fact that this cluster contained measures that transfer the responsibility for healthy development onto the parents, rather than making it a public policy issue. Moreover, some of the proposed measures can be viewed as paternalistic, as they do not account for participation and involvement of the Roma community in the design or content of the proposed interventions. Paternalistic measures fail to foster autonomy in the recipient, and therefore such measures replicate the power relations that have contributed to the need for these services in the first place [6]. Participatory measures, on the other hand, promote physical, mental and social health and reduce inequities in health [18]. The general population views Roma as unchangeable and incapable of making wise choices [36]. As a consequence, Roma as a target of policies are perceived as needing directive and authoritarian means and as not capable of participating in enhancing their own conditions. Since many of the measures proposed in this cluster already exist to some extent, participants could feel that the measures in this cluster are not as urgent as those in other clusters.

Within the priority measures, a variety of measures belonging to each of the clusters were included. Measures with the highest urgency and feasibility targeted planning parenthood, on the one hand, and scaling up existing projects on the other hand (See measures 65, 64, 67, 68, 22, 75, and 90 in the 15). The first five of the above-listed measures focus on planning parenthood, in particular on education and awareness about reproductive health, readiness for future parenthood, prenatal development, and the availability of contraception. Roma women get pregnant for the first time earlier than non-Roma women and have a higher number of pregnancies per woman [50]. Unintended pregnancy leads to health-compromising behaviours in pregnancy [32], increases the risk of low birthweight [21] and can lead to negative outcomes for child health and development [11]. Presently, education on sexual, relational and reproductive health, that would meet the international human rights standards, is not available in Slovakia [33]. The high priority given to these measures by our participants suggests that participants view the issues of unwanted and mistimed pregnancies as pressing and the solutions as attainable. Other measures rated as the most urgent and feasible focused on scaling up already existing successful small-scale projects regarding early childhood interventions (Project OMAMA [1];) and access to housing via microloans for self-help construction [38]. These regard examples of good practices in enhancing the living conditions of marginalized Roma [3]. The high priority of these measures reflects the fact that our participants believe in the feasibility of interventions generally viewed as effective and successful.

Strengths and limitations

The main strength of this study regards the size and quality of the sample, because of its variety, including the viewpoints of diverse stakeholders from practice, research and policy. Stakeholders proposed a set of relevant and acceptable measures, which may help reduce inequalities in the early childhood of children from marginalized Roma communities. Our study also has some limitations, the first being that we lost 14 out the initial 54 participants in the sorting-rating phase of the study. However, this is in accordance with the GCM methodology, which takes into account some losses of participants between the brainstorming and sorting-rating step without bias being likely [24]. Second, the GCM methodology may be prone to social desirability. We reduced the likelihood of this bias by anonymizing the brainstorming phase, which also helped to reduce the potential effects of power relations between participants. Third, the GCM methodology might be prone to subjectivity, typical for qualitative methods, as researchers interact with the data generated by the participants whose selection was purposive. However, we tried to eliminate this as much as possible by using a participatory approach and discussing each step of our study with the participants, implying that in all steps decisions were made by several people jointly. Finally, the interpretations and implications of our finding should be taken with caution, as they were not discussed with representatives of the target population, showing a need for confirmation in future research.

Implications

Our finding that measures on living conditions are rated as most urgent but least feasible show that this topic evidently deserves further attention regarding the development of measures and their evaluation. The availability of EU funding could facilitate such actions, but evidently the various barriers for use of these funds require further study. Implementing an international monitoring protocol that would collect data both on compliance with the CRC as well as on the developmental outcomes was previously proposed [22]. In the case of violation of children’s rights, the United Nations could utilize legal and political channels to make claims of rights violation.

We identified four clusters and a number of measures that should have priority and which require further interpretation jointly with the target population. The target population could also complete the sorting-rating phase with the same set of proposed measures. Alternatively, this study could be replicated with a different set of stakeholders.

Our findings imply that several policy and practice measures are needed in particular to improve the healthy early childhood development of children living in marginalized Roma communities. The implementation of such measures requires a further involvement of appropriate stakeholders. The participants further identified a high need for coordination and cooperation in addressing the inequalities in early childhood, and hence policymakers should make an effort to approach this matter from a complex perspective. Since Slovakia lacks an integrated system of education and care for children aged 0–6 [34], the creation of such a system is highly needed. Many proposed measures highlight the need for participation. Currently, marginalized Roma hardly participate in policymaking and self-governance [23] and should thus be invited to participate in formulating needs and creating and running programmes aimed at helping Roma. Moreover, our findings show barriers in access to and the capacity of services that should be targeted.

The individual measures that were given the highest priority by the participants highlight the need to address the inability of young Roma to plan their parenthood, so access to methods of contraception should be ensured, as well as access to quality sexual education. Participants gave high priority to measures that would only scale up already existing interventions [1], which means that measures should build upon existing good-practice interventions and programmes.

The prepared summary report of this study in Slovak language will be disseminated among the participants as well as relevant policy makers, to open the discussion about proposed measures and to push for policy change even in those measures, that the participants viewed as not feasible.

Conclusion

Participants proposed a set of relevant measures that would help equalize the chances of children from marginalized Roma communities for healthy early development, which are in line with European strategic framework for equality, inclusion and participation [15] and which reflect the most pressing issues. Participants viewed the enhancement of living conditions as the most urgent and community interventions focused on the transfer of cultural capital as the most feasible. Discrepancies between urgency and feasibility show significant barriers in the implementation of the proposed measures; they reflect the public discourse and the lack of political will to address the problems of marginalized Roma. The most urgent and feasible measures regarded planning parenthood or were a scaling up of already existing programmes. The proposed need for a supra-sectoral strategy as a separate measure and the variability of proposed measures confirmed that reducing inequalities in early childhood development is a complex issue and needs to be addressed through cross-sectoral cooperation and coordinated efforts.