Population‐based approaches for monitoring the nurturing care environment for early childhood development: A scoping review

Abstract Selecting indicators to monitor nurturing care (NC) environments that support decision‐making and guide the implementation of integrated early childhood development (ECD) programmes has become a priority globally. Several population‐based approaches have been attempted to create a set of indicators or a composite index methodology to measure the NC environment using existing secondary data. However, they have not been systematized. Our scoping review aimed to analyse the population‐based approaches for monitoring the domains of the NC (e.g. good health, adequate nutrition, responsive caregiving, security and safety, and opportunities for early learning). ECD experts, peer‐reviewed, and grey literature were systematically searched with no year or language restrictions. Data extraction used a standard predefined protocol. Thirty‐two population‐based approaches were identified. Most approaches were composed of a set of indicators (53.1%, n = 17) versus composite indexes (46.9%, n = 15) and had the country as their unit of analysis (68.8%, n = 22). Twenty‐seven approaches were applied in middle‐income countries (84.4%) and thirteen in low‐income countries (40.6%). Four approaches were guided by the NC framework (12.5%), and 56.3% (n = 18) did not include any indicator representing responsive caregiving. NC indicators (n = 867) were sorted into 100 groups of indicators. Twenty of the 32 approaches had some kind of methodological validation (62.5%). We identified six methodological challenges to build a population‐based approach. Standardized methods for selecting and validating indicators, and coordinated efforts to share findings/data with stakeholders should be prioritized. Given the great variability in methods and indicators used to measure NC environments, valid approaches should be flexible to work well across different contexts.


| INTRODUCTION
Early childhood is a phase characterized by intense neurogenesis and brain plasticity in response to nurturing care (NC) (Clark et al., 2020;World Health Organization et al., 2018). The NC framework calls for safe, secure, and stimulating environments where children have opportunities to learn and interact with caregivers that are emotionally supportive, sensitive, and responsive to their developmental and physiological needs (Black et al., 2017;Pérez-Escamilla & Segura-Pérez, 2020;Richter et al., 2017;World Health Organization et al., 2018). However, that is not a reality for millions of children worldwide. In low-and middle-income countries, 250 million children are at risk of not reaching their full potential due to a variety of unfavourable conditions that threaten early childhood development (ECD), such as extreme poverty, hunger, and violence (Black et al., 2017;Clark et al., 2020;Lu et al., 2016;Richter et al., 2017;World Health Organization et al., 2018). In this context, investing in ECD services, programmes, and policies is one of the most cost-effective mechanisms to support human development and provide an enabling environment for NC (Clark et al., 2020;World Health Organization et al., 2018).
Recently, a major effort to operationalize NC for ECD outlined five strategic actions (World Health Organization et al., 2019). Specifically, the fourth strategic action calls for countries to develop mechanisms to monitor activities that support NC across the five domains (good health, adequate nutrition, responsive caregiving, security and safety, and opportunities for early learning) at the individual, population, and programing/system level (World Health Organization et al., 2019). Building on this call, monitoring NC framework and its components have become a priority globally to support decision-making, advocacy, and tracking progress on different governance levels (Black et al., 2017;Clark et al., 2020;Lu et al., 2020;Pérez-Escamilla & Segura-Pérez, 2020;Richter et al., 2017;World Health Organization et al., 2018).
Several studies have attempted to develop population-based approaches for monitoring NC over the past few years. In this manuscript, NC population-based approaches are defined as a set of existing indicators or a composite index methodology that uses secondary data to measure, characterize, classify, and evaluate NC environment(s) for ECD Köhler, 2016;UNICEF & Countdown to 2030. The Countdown to 2030 early childhood country profile, the State of Babies in the United States, and the Early Childhood Friendly Municipal Index (IMAPI) in Brazil are examples of initiatives that measure NC at the countries, state, and municipal level, respectively Keating, Cole, & Schaffner, 2020;UNICEF & Countdown to 2030.
However, to our knowledge, no systematic mapping of these NC population-based approaches has been conducted.
Therefore, the aim of this scoping review is to analyse the population-based approaches for monitoring the domains of the NC framework.

| Eligibility criteria
We included all types of study designs, articles, reports, and websites that presented population-based approaches which evaluated the NC environment for ECD, assessing more than one domain of the NC framework (good health, adequate nutrition, responsive caregiving, security and safety, and opportunities for early learning) (World Health Organization et al., 2018). There was no restriction regarding language and publication period, and the search included both published and unpublished (grey literature) materials. We excluded studies that (1) were unavailable to retrieve (webpage not found n = 9; unavailable website n = 5; unavailable full text n = 9), (2) were not related with ECD, (3) evaluated only children over five years old or children with specific characteristics (e.g., preterm, low birth weight, atypical, with pathologies, twins, foster children), (4) did not assess policy environments related to early childhood (i.e., studies evaluating neonatal ICU; orphanage; natural disasters; exposure to chemicals, metals, toxins, drugs, and alcohol; and which only evaluate indicators related to the home and family), (5) did not evaluate the NC environment for ECD, and (6) did not use indicators from secondary databases.

| Assessment of population-based approaches characteristics and data extraction
A pretest of the standardized protocol to extract data was done by two reviewers (JP and SEACK). First, the approaches were classified into two groups: (1) Set of indicators, which are a group of individual NC indicators (measures generated from observed facts that can evaluate performance, positions, and changes across time if evaluated regularly) that did not have a single overall summary measure and (2) composite indexes, which are a set of individual NC indicators combined into a single overall summary measure (Köhler, 2016; Organization for Economic Co-operation and Development, 2008  To disseminate the approaches, less than a half had websites (25.0%, n = 8), presented rankings (43.8%, n = 14), or maps (21.9%, n = 7). To share the population-based approaches with the end users (e.g., decision makers), less than one fourth (21.9%, n = 7) had a dynamic type of data communication, and the type of user interaction design was similar between composite indexes and set of indicators.
Twenty of the 32 approaches included were validated (62.5%). While less than one third (29.4%, n = 5) of the set of indicators have been validated, 80% (n = 12) of the composite indexes have been validated (               Organization et al., 2018). However, the majority of approaches mapped in our scoping review did not have disaggregation by geographic location or an equity approach. A shift to more local level monitoring approaches is expected as we move forward, especially in LMIC given the need to understand where the vulnerabilities are.
Therefore, we recommend that approaches include indicator(s) that can be disaggregated by geographic location, sociodemographic (i.e., gender and ethnicity), and economic characteristics (i.e., social status), to allow the identification of the most vulnerable families (Clark et al., 2020;World Health Organization et al., 2018).
In addition, decision-makers must understand how NC environ- (IMAPI) reported in this supplement issue  For this approach, it was reported that association tests between the set of indicators or composite index were conducted with other measures, indicators, and/or instruments. c For this approach, it was reported that the identification of indicators was based on a participatory process involving experts on ECD and stakeholders.
There are trade-offs between population-based approaches composed of a set of indicators (i.e., a group of individual NC indicators without a single overall summary measure) and those composite indexes (i.e., NC indicators combined into a single overall summary measure); thus, the choice of one over the other should be based on the goals and intended use (Köhler, 2016; Organization for Economic T A B L E 4 Pathways to move forward with the methodological challenges of building and using an approach to monitor Nurturing Care environment for Early Childhood Development

Methodological challenges Gap Pathways to move forward
Lack of clarity about the definition of nurturing care indicators There is no international consensus on the best indicators to measure the nurturing care environment, posing a challenge for researchers and stakeholders on the best indicators to adequately measure the complexity of the environment.
• Agree on a common approach to select indicators that are relevant and suitable to different contexts (e.g., guidance on participatory approaches). • Develop a guidance on basic indicators that needs to be prioritized and others that can be adapted to each sociocultural and political context. • The Nurturing Care Framework is a good theoretical framework for the construction of this common base.
Data quality and availability Critical aspects are related to timeliness and completeness of data, comparability between units of analysis (countries/ states/municipalities), use of proxy measures for some indicators, self-report survey data, and data accuracy. This aspect was the most reported limitation and challenge when building an approach.
• Prioritize standardized data collected from routine information systems and complement with data from recurrent surveys. • Use data from internationally recognized institutions.
Equity approach Lack of equity lens when building a nurturing care monitoring system (i.e., consideration of gender, ethnicity, social class, and other social characteristics on the indicators and/or approaches).
• Consider multiple layers of equity aspects when selecting nurturing care indicators. • Advocate for data collection that provides disaggregated data on these aspects.

Validity of approaches
Validation is essential to ensure that the assessment approach is truthful and accurate in measuring the nurturing care.
Assessing construct validity was identified as the greatest challenge in this regard.
• Develop processes that allows building the approach in a participatory way (government, researchers, international institutions). • Define other indicators and/or composite indexes related to child development outputs and inputs to test the correlation with the results of the approach.

Weighing of indicators (only for composite indexes)
Whether or not to weigh indicators of composite index is not a consensus. It may also be due to the lack of a solid empirical or statistical basis in the construction of the composite approaches so far.
• Starting from a strong theoretical base that enables the understanding of causal relationships. • Conduct a participatory process with experts to identify the best way to weigh the indicators. • Test different statistical methods of weighing method and define the best fit based on clear and transparent criteria Sharing of nurturing care data to users Type of resources used to communicate data may limit the possibilities for understanding and analysing the results.
• Conduct small meetings with key stakeholders to consult with them about the best forms to provide data for their decision making. • Develop dynamic resources for data communication, such as interactive websites, maps, etc. • Provide a manual with clear steps on how the approach was built, which data were used, and the aspects of validity can help increasing the use of data. Co-operation and Development, 2008 Development, 2008). Therefore, approaches with composite indexes must be well-designed, following carefully evidence-based methodologies . The indicators are often standardized to allow comparisons among them and aggregations (Köhler, 2016 Delivering NC data to those who will use them in an accurate and comprehensive manner can be challenging. To be effective, data should be presented to decision makers in creative and user-friendly formats, such as graphs, dashboards, score cards, rankings, and maps (Köhler, 2016;Organization for Economic Co-operation and Development, 2008;World Health Organization, 2019;World Health Organization et al., 2018). The dynamic type of communication is likely to be a more useful option because it allows stakeholders to interact with data according to their interests and goals. IMAPI is an example of a population-based approach with dynamic communication. It has a user-friendly format, presenting on the website a page for each municipality, a ranking for the general index and for each domain of the NC framework, it also allows the exploration of data at the regional, state, and municipal scale . We recommend that researchers not only focus on sharing data but also on how to improve the understanding, interest, and interpretation by stakeholders, professionals, and the overall population (Köhler, 2016; Organization for Economic Co-operation and Development, 2008;Sarkies et al., 2017;World Health Organization et al., 2018). To make the message clearer, researchers may use knowledge-brokers and other materials (e.g., country profile, maps) and avoid the use of technical language (Baker et al., 2018;Sarkies et al., 2017). They should also make resources available that support transformation, given that an adequate presentation of data to its users may be an important step to support decision-making (Sarkies et al., 2017). In addition to the presentation, the approaches must be valid and reliable to support high quality decision-making processes. Stakeholders, including government, researchers, civil society organizations, and international institutions, should participate in the development of approaches to monitor the NC, guaranteeing that the data addresses the key decision-making related questions. Unfortunately, despite its importance, only half of the approaches have been validated; thus, efforts should be made to validate the existing approaches guaranteeing to stakeholders that they can trust the information provided to guide decision-making.
We acknowledge that our scoping review has some limitations.
As other literature reviews, the publication and selection bias may affect the results. However, to decrease the risk of this type of bias, we registered the protocol a priori and included in the search strategy grey literature and consultation with ECD experts. Despite these limitations, the findings of this scoping review are robust and elucidate a gap in the literature by mapping the characteristics of populationbased approaches for monitoring the NC environment for ECD. More studies are necessary to validate the population-based approaches and address how the information provided by them can actually promote evidence-based decision-making. Thus, we recommend that future studies apply sound implementation science and research methods to design, implement, scale up, and evaluate strategies with the potential to improve the NC environment for ECD, for a better future for children. In addition, it is important that future studies verify whether and how the approaches are used by stakeholders and the quality with which they are being implemented. Qualitative studies with programme planners, policymakers, government officials, and researchers, as well as with civil society organizations and the general population will be key to evaluate implementation quality aspects of diverse NC environments monitoring systems.
Based on the results of this scoping review, we recommend that (1) a global monitoring system with a set of core NC indicators should be created, allowing comparisons across countries and throughout time (World Health Organization, 2019); (2) disaggregated data should be available, facilitating the identification of the most vulnerable groups according to geographic location, sociodemographic, and economic characteristics (Clark et al., 2020;World Health Organization et al., 2018); (3) approaches should focus on a continuous monitoring of NC indicators, allowing stakeholders to identify barriers for integrated ECD programmes and services and to promptly address them to maintain adequate levels of coverage and quality over time; (4) approaches should be systematically constructed, validated, and interpreted, based on a careful selection of robust, relevant, reliable, and valid indicators (Köhler, 2016; Organization for Economic Co-operation and Development, 2008); (5) authors should provide specific and transparent reporting of the methodology followed (Köhler, 2016) so that others can replicate their findings; (6) countries should address the NC data gaps identified in this scoping review, especially for responsive caregiving indicators ; and (7) approaches should be presented and disseminated in a userfriendly dynamic format, increasing their comprehension and use by the stakeholders and the overall population (Köhler, 2016; Organization for Economic Co-operation and Development, 2008).
In summary, the existing approaches vary greatly in their methodology for selecting indicators or composite indexes to measure the NC environment. Standardized methodologies to select sensitive indicators are needed, and they must consider data availability, data quality, and level of disaggregation. User-friendly formats to report and disseminate valid approaches to measure NC environment(s) are necessary to support decision-making to advance ECD globally.