Harmonizing Data Visualizations on Child Health and Well-Being to Strengthen Advocacy and Monitoring Efforts

Data visualization tools on child health have improved data accessibility but caused confusion over indicator data sources and which tools to use for specific purposes. We propose principles for generating future tools that can effectively trigger action and accountability for children everywhere.


INTRODUCTION
I n the past few decades, there has been a proliferation of country profiles and interactive web-based dashboards that graphically present data on child health and well-being.These tools are designed to make data more accessible to a wide range of audiences, facilitate simple analyses and interpretation of data, and foster greater uptake of data for decision-making as well as for accountability. 1,2However, the growing volume of data visualization tools has created confusion over which ones to use for specific purposes (e.g., global monitoring, national and subnational monitoring, program-specific monitoring, and advocacy and accountability), how they complement each other (i.e., how to use global dashboards in combination with more detailed subnational data for planning and resource allocation purposes), and where the data come from.We summarize key historical events that fostered the surge in data visualization tools on child health and well-being, discuss the benefits and challenges that have resulted from this surge, and present a proposed set of principles for producing future tools for global monitoring and crosscountry comparisons.

THE RISE OF STATISTICAL PROFILES ON CHILD HEALTH AND WELL-BEING: A BRIEF HISTORY
The increase in data visualization tools on child health is linked to advances in data analytics, improvement in data availability, and the emergence of global accountability frameworks that include goals and associated targets.Assessing progress toward these goals and targets necessitates good measurement, regular monitoring, and effective communication about the indicators used.The World Declaration on the Survival, Protection, and Development of Children and the associated plan of action for its implementation (Declaration) 3 adopted by the World Summit for Children in New York in September 1990 was the first major framework with a The World Bank Group, Global Financing Facility, Washington, DC, USA.goals and targets for children.This framework was consistent with the Convention on the Rights of the Child adopted the year before by the United Nations General Assembly 4 and the ethos of the Child Survival Revolution launched in 1982. 5he framework included 27 goals covering child survival, child development and protection, education, and nutrition.In recognition of the importance of maternal health and social determinants for child health and well-being, goals were also included on women's health, water and sanitation, and poverty. 3he establishment of these goals triggered demand for rigorous annual reviews of progress toward their achievement based on comparable, reliable data and simple visualizations that policymakers could readily understand.In response, agencies with a remit on child health, like UNICEF and the World Health Organization (WHO), began strengthening their monitoring and evaluation capacities and improving the display of data from their publicly accessible global databases.In the mid-1980s, the U.S. Agency for International Development-supported Demographic and Health Surveys (DHS) 6 were rolled out to systematically measure health outcomes for women and children and were followed by the introduction of the UNICEF-supported Multiple Indicator Cluster Surveys in the early 1990s. 7he mid-and end-decade evaluations of the 1990 Declaration resulted in a call for reaffirmation of the commitments made to children, increased investments in social services for all, and greater economic and social empowerment of the poor. 8,9These aspirations were reflected in the outcomes of the September 2000 United Nations Millennium Summit when the United Nations Millennium Declaration was adopted with a set of time-bound goals and targets that placed maternal and child survival at their core. 10Two years later, a special session of the General Assembly was held to review progress on the Declaration and renew global commitment to children's rights.The result was adoption of the World Fit for Children resolution with a list of goals, targets, and strategies for improving children's lives. 8e 8 interlinked Millennium Development Goals to be achieved by 2015 and country commitments to the World Fit for Children fueled monitoring and measurement activities on child health and sparked the establishment of several global accountability initiatives that included regular quantitative progress assessments. 5The Countdown to 2015 initiative, for example, was launched in 2003 as a rallying cry to hold the world to account for progress toward Millennium Development Goal 4 on child survival. 11In 2005, Countdown published its first global progress report and pioneered a set of 1-page country profiles for the highest child mortality countries based on a standard template and the inclusion of indicators for which robust, comparable data were available.3][14] Technical advisory groups, such as the Child Health Epidemiology Reference Group (now the Maternal and Child Health Epidemiology Estimation Group), were also formed to support measurement advancements in child health. 5ountdown was not alone in its production of statistical profiles to catalyze action.The Millennium Development Goal era ushered in a wave of similar data visualization tools produced by other maternal and child health-related initiatives, such as the Every Newborn Action Plan, 15 Family Planning 2020 (now Family Planning 2030), 16 and the African Leaders Malaria Alliance. 17United Nations agencies also began producing their own statistical snapshots to support country programming and monitoring and evaluation efforts (e.g., UNICEF's HIV 18 and immunization 19 statistical profiles, among others).
The propagation of data visualization tools on child health, including mobile-friendly applications and web-based data portals with interactive interfaces, has escalated during the Sustainable Development Goal era in part because of their popularity and advancements in technology, making them easier and cheaper to build. 20Data portals that enable users to generate graphical displays of child health data, for example, are now available for the Global Strategy for Women's, Children's, and Adolescents' Health (2016-2030) 21 and the World Bank's Global Financing Facility for Women's, Children's and Adolescents' Health. 22Many universities, research institutions, and international nongovernmental organizations also publish data visualizations to promote advocacy and accountability (e.g., the Institute for Health Metrics, 23 Save the Children's and UNICEF's co-produced Child Health Spotlights, 24 University of Oxford's Our World in Data, 25 and the Johns Hopkins University Lives Saved Tool 26 among many others).Data collection platforms, such as the Demographic and Health Surveys, Multiple Indicator Cluster Surveys, 6,7 and the DHIS2, 27 provide applications for users to create graphics.In response to increased demands for data during the first years of the COVID-19 pandemic, numerous dashboards were created to help countries monitor and address the pandemic's impact on essential health services for women and children (e.g., WHO and UNICEF's pulse survey trackers 28,29 ).

THE NEED FOR STREAMLINING AND HARMONIZING STATISTICAL PROFILES ON CHILD HEALTH
The abundance of statistical dashboards on child health has helped put data into the hands of decision-makers and civil society actors and has made data more digestible and interpretable to nontechnical audiences.Their promotion through webinars and electronic communications has also helped keep children visible on the global landscape and further promoted the democratization of data (e.g., webinars hosted by the Health Data Collaborative and information shared through the Partnership for Maternal, Newborn, and Child Health monthly e-blasts).However, the sheer volume of data visualization tools has resulted in some discordance, redundancy, and frequent overlap in the information presented.There are many country profiles generated at the global level that include the same or similar child health-related indicators yet define or label them differently, use different data sources for them, update them at different time points, or display them in inconsistent ways, creating confusion.As an example, the Table presents variations in the display of care-seeking indicators for childhood illnesses on 3 global child health-related dashboards.These variations can result in confusion over the latest coverage values for these indicators, given differences in timelines for updating and how they should be interpreted and monitored.For example, a user may not be able to determine if care-seeking for symptoms of acute respiratory infection, as shown on the Countdown profile, is the same or a different indicator from care-seeking behaviors for pneumonia as presented on the Child Health Spotlight.

A WAY FORWARD: OPPORTUNITIES AND RECOMMENDATIONS FOR HARMONIZING DATA VISUALIZATIONS FOR CHILD HEALTH
In recognition of these problems and their contributory role, WHO and UNICEF began working closely together in recent years to improve coordination and streamlining of their data visualization tools on women's, children's, and adolescents' health.For example, through the Children in All Policies 2030 initiative, WHO and UNICEF led an extensive consultative process to create a simple dashboard in support of the Convention on the Rights of the Child. 30This dashboard was launched in May 2022 simultaneously on WHO, UNICEF, and Children in All Policies 2030 websites, and a plan was established to continue jointly developing and updating the back-end databases and the front-end data display.Both agencies are also putting procedures into place to ensure their websites are consistent in terms of data sources used, timing of updates, and how data on women's, children's, and adolescents' health are visualized.
To help users navigate and understand the complementarity of existing data visualization tools on women's, children's, and adolescents' health, UNICEF, in collaboration with WHO, created a webpage to compile key data visualization tools and links to other resources, including related topic-specific data visualization tools, global initiatives, journal articles, blogs, and reports. 31his webpage will need continuous curation to keep pace with new data visualization tools and the changing evidence base.
The following proposed set of good practices for producing data visualization tools, particularly for global monitoring, were developed by the WHO-UNICEF co-convened Child Health Accountability Tracking (CHAT) Technical Advisory Group. 32A function of CHAT is to help harmonize indicators and information on child health.This set was developed based on review of data visualization theory and experiences to date with the design and dissemination of child health and well-being dashboards, including the example of WHO and UNICEF working together on the Children in All Policies 2030 dashboards. 1,2,20,33Specify what need the tool will fulfill (e.g., will it provide information on a specific topic in child health, or will it enable progress assessments across a set of child health issues, etc.).
Determine if the tool will add value.This decision may involve reviewing existing tools and assessing whether a new tool will provide needed and actionable complementary information.The continuation of the mapping work started by UNICEF and the World Health Organization on data visualization tools for women's, children's, and adolescents' health, with additional information on the intended purpose and target audience for each tool, is important for facilitating this step.
3. The purpose and target audience of any new data visualization tool for child health should be defined at the outset to avoid duplication and to clarify its added value (Box 1).
4. Data visualization tools should be developed through an inclusive and consultative exercise.Some options include the creation of an advisory group composed of a diverse representation of actors, a wide consultative process done through an open web-based platform, or a tailored delphi-like process that solicits inputs from all key stakeholders.Country actors and civil society organizations should be engaged throughout the entire development process to ensure any new data visualization tools on child health meet their decision-making or advocacy-related needs.
Technical and data visualization experts should also be involved in all steps so that the tool is based on the latest scientific evidence and graphical techniques.
5. The development of a data visualization tool should be iterative, with a lead institution or individual managing and building consensus throughout the process (Box 2).This consensus-building process involves engaging with all partners identified in principle 4.
6. Indicators selected for inclusion on data visualization tools should ideally be based on standard definitions and recommended data sources, as well as be readily available through existing data collection platforms.They should be as consistent as possible with indicator recommendations from the following 3 technical advisory groups established to promote indicator harmonization and standardization: Mother and Newborn Information for Tracking Outcomes and Results, 35 CHAT, 36 and the Global Action for Measurement of Adolescent Health. 37though these good practices are geared for the development of global, regional, or national level tools that allow for comparisons across countries and time, most are applicable to the development of subnational data visualization tools based on local data sources as well.If these practices are followed, the usefulness and quality of data visualization tools that can spur action and drive accountability for child health and well-being could be substantially improved.
bJohns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA.Health Services Unit, KEMRI-Wellcome Trust Research Programme, Nairobi, Kenya.
c World Health Organization, Geneva, Switzerland.d McGill University, Montreal, Canada.e f Maternal and Child Health Division, icddr,b, Dhaka, Bangladesh.g Department of Pediatrics and Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD, USA; RTI International, Research Triangle Park, NC, USA.h University Federal Fluminense, Rio de Janeiro, Brazil.i Center for Birth Cohort Studies, University of Yamanashi, Yamanashi, Japan.Global Health: Science and Practice 2023 | Volume 11 | Number 6

TABLE .
Examples of Inconsistencies in Key Global Data Visualization Tools for Child Health and Well-Being, Examination of Care-Seeking for Childhood Illnesses Indicators Checklist for Defining the Purpose of a Data Visualization Tool for Child Health and Well-BeingIdentify the target audience (e.g., local decision-makers, civil society organizations, health care professionals, policymakers, donors or other development partners, regional organizations, etc.).