Introduction

Humanitarian crises present critical threats to human health and security. Armed conflict, climate-related disasters, hunger, forced displacement, and communicable disease outbreaks are associated with the disruption and collapse of essential health services, food supply chains, water and sanitation facilities, education, and social protection programmes. Children and young people are disproportionately vulnerable to direct and indirect harms associated with crises, with nearly 1 in 5 living in or fleeing from armed conflict [1]. Child mortality data in crisis contexts are lacking [2], however the available data suggest that under 5 mortality rates in fragile and conflict-affected countries is almost three times higher than in countries not classified as fragile or conflict-affected, and nearly twice that of all countries classified as low- or middle income [3].

Few crises happen in a vacuum, however, and increasingly, crises are protracted. The result is that children face multiple forms of adversity before the crisis occurs, throughout the crisis, and often long after the acute phase has subsided. The harmful effects of early childhood adversity have been well documented in high income settings, with immediate and long-term impacts on physical health; mental health; growth; cognitive, social and emotional development; and life trajectories [4]. Interventions to protect and promote child health in FCV settings should be grounded in an understanding of the pervasive impact of adversity on children; this requires collaboration with sectors not traditionally considered to fall within the domain of public health, such as child protection and education. Additionally, the measurement of outcomes and impacts relating to severe adversity on child health and development is needed to adequately inform and monitor the safety and effectiveness of child public health interventions.

Armed conflict and humanitarian disasters often occur in settings where weak public health information systems become further strained by crisis. In most fragile, conflict-affected, and vulnerable (FCV) settings, there is limited data on the health needs of children. In the absence of routine child public health data, humanitarian actors delivering public health services in these settings often turn to survey data. The Demographic and Health Survey (DHS) and/or the Multiple Index Cluster Survey (MICS) may be used to support decision-making and prioritisation of interventions. These surveys capture public health data from only a single point in time and may be out-dated by the time they are used to inform an intervention. During humanitarian responses, the existing survey information is usually triangulated with a rapid needs assessment and the available governmental data [5, 6]. In the end, however, many humanitarian interventions are developed and delivered without having up-to-date child public health data to inform them. At the same time, nongovernmental actors working in FCV settings are actively collecting a large amount of data for internal reporting, donor reporting, and reporting to governments; so much data are collected that the United Nations Office for Coordination of Humanitarian Affairs (OCHA) established a Centre for Humanitarian Data to improve data use and data practices [7]. There is a huge opportunity for the improved measurement and use of child public health data to inform interventions and to monitor and evaluate outcomes.

The need for standardised, high-quality child public health data in FCV settings has never been greater. Recent years have seen a growth in the need for assistance to crisis-affected populations and simultaneous reduction in funding to support humanitarian response. As of 30 April 2024, the Inter-Agency coordinated funding appeals totalled US$48.2 billion, however the reported received funding is only 15% of this - the lowest level of funding in the same period since 2019, at $1.39 billion [8]. Simultaneously, monitoring and understanding the effectiveness of humanitarian aid is difficult due to a lack of standardised performance indicators disaggregated by age and gender, and in particular, lack of measurement of outcomes [9]. Increased need with simultaneously reduced funding makes it even more important to base operational decisions on sound, up-to-date, and relevant information. The effectiveness of interventions to support children in humanitarian contexts depends on trustworthy and appropriately detailed data on child health status, health risks, and services available before the crisis hit as well as during the crisis and after the acute phase has passed. These data are essential to allocate scare funds effectively, establish priorities, evaluate the safety and performance of interventions, and maintain accountability to affected populations.

Current challenges to understanding the public health needs of children in FCV settings include a lack of consensus on the definition of a child, intermittent data collection, wide variations in disaggregation of indicators by age, limitations in public health infrastructure in humanitarian settings, siloed interventions by multiple sectors engaging in humanitarian response, and limited understanding of child public health by generalist public health actors [10]. These challenges to the collection of child public health data mirror challenges in delivering appropriate and effective child public health interventions in FCV settings.

A standard set of relevant, realistically obtainable core child public health indicators in fragile, conflict-affected, and vulnerable (FCV) settings is needed to improve our understanding of children’s needs, support prioritisation of activities, inform interventions, support coordination of interventions and accountability, and track progress. This scoping review aims to make a step towards the development of a core set of child public indicators by providing an overview of the current indicator landscape and identifying overlaps, divergences, gaps, and potential for harmonisation and pooling of data.

Methods

Search strategy

This review characterises child public health indicators that are routinely collected, required, or recommended for use in FCV settings (Table 1). We use the term “humanitarian” to refer to crisis settings where external support is required to meet the needs of the affected population. The terms “crisis” and “humanitarian” are used interchangeably. The term “FCV” includes a broader range of settings, including pre- and post-crisis as well as protracted crises, for which external support may or may not be required to meet the population’s needs. For the purpose of this review, child public health indicators were defined as summary measures that captured information on different attributes of children’s health. We defined four child public health indicator domains for this review: health status, healthcare, social determinants of health, and health behaviours (See Table 1 for definitions and Table 2 for domains). These domains are based on the public health practice areas of health protection, health improvement, and health service delivery, considering the types of interventions conducted in FCV settings and the data required for their planning, monitoring, and evaluation. In order to identify indicators that are feasible to collect, required, and recommended for use in FCV settings, we collected lists of (i) indicators that are routinely collected by nongovernmental agencies providing public health services in FCV settings; (ii) required indicators from a sample of donors; and (iii) indicators recommended for routine use by technical agencies, technical advisory groups, partnerships, and academics working in public health in FCV settings.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. List of definitions.

https://doi.org/10.1371/journal.pgph.0003843.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Child public health indicator domains.

https://doi.org/10.1371/journal.pgph.0003843.t002

There were two rounds of data collection in this review: 1) a search of the peer-reviewed and grey literature, and 2) collection of routinely collected, required, and recommended child public health indicators from operational agencies delivering child public health services in FCV settings, donors, and groups or organisations providing technical guidance for child public health programming, respectively.

Peer-reviewed and grey literature database searches

Database searches were developed by AS with support from AK and an experienced librarian at the Leeds Teaching Hospitals NHS Trust and undertaken 14-24 March 2024. AS conducted searches in Embase and MEDLINE from 1^st January 2013 to 24^th March 2023, with English language restrictions. The searches included variations of terms for: (1) fragile, conflict-affected, and vulnerable settings; (2) child; (3) public health; (4) indicators. The final search strategy for EMBASE is available in S1 Text. Identified records were exported into the online reference manager, Rayyan [17]. Duplicates were removed and the records were screened by AS and AK.

A grey literature search was run by AS in DuckDuckGo [18] and through screening the websites of major agencies supporting child public health activities in FCV settings. The DuckDuckGo search included the terms ‘Child public health indicators humanitarian' and expanded to include ‘all regions', ‘anytime’, in a ‘moderate safe search’. Organisational websites searched included the World Health Organisation (WHO), WHO Global Health Cluster (GHC), United Nations Office for the Coordination of Humanitarian Affairs (OCHA), United Nations Children’s Fund (UNICEF), United Nations High Commission for Refugees (UNHCR), Sphere Handbook, United States Agency for International Development (USAID), European Civil Protection and Humanitarian Aid Operations (ECHO), the United Kingdom Foreign, Commonwealth and Development Office (FCDO), Alliance for Child Protection in Humanitarian Action, Inter-agency Network for Education in Emergencies, Healthy Newborn Network, Inter-Agency Working Group on Reproductive Health in Crises, Interagency Gender Working Group, and ReliefWeb.

Reference lists of identified records from the database and grey literature searches were scanned and relevant references were reviewed to identify additional documents containing routinely collected, required, or recommended child public health indicators for FCV settings. AS and AK independently screened all records obtained by the database and grey literature searches. Disagreements were resolved by discussion. Records from the peer-reviewed literature and grey literature searches were included for extraction of indicators if they contained at least one indicator that met the inclusion criteria (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Inclusion and exclusion eligibility criteria.

https://doi.org/10.1371/journal.pgph.0003843.t003

Results

The database searches identified 662 records. After removing duplicates, the titles and abstracts of 504 records were screened for eligibility. Seven records were screened in full text, of which one report contained indicators that were relevant for inclusion in the review. A further 135 further records were identified from grey literature searches, snowball searches, and indicator lists obtained from a sample of donors, operational agencies, and technical experts, of which 23 records contained indicators relevant for the review. The final dataset contained indicators from 24 records: three peer-reviewed publications, nine indicator lists from 8 global technical organisations or partnerships (OCHA, UNHCR, Sphere, Alliance for Child Protection in Humanitarian Action, Inter-Agency Working Group on Reproductive Health in Crises, Healthy Newborn Network, Inter-agency Network for Education in Emergencies and the Interagency Gender Working Group), three indicator lists from two donor agencies (Bureau for Humanitarian Assistance (USAID BHA) and ECHO), and nine indicator lists from five operational agencies (Save the Children, MSF, ICRC, IRC, and UNICEF). The indicators from the included records were extracted and screened to identify recommended, required, and routinely collected child public health indicators in FCV settings, respectively (Fig 1 and S4 Text).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram.

https://doi.org/10.1371/journal.pgph.0003843.g001

A total of 2784 indicators were extracted and screened, of which 668 met inclusion criteria (Table 4). The included indicators were categorised into (i) routinely collected; (ii) required by donors; and (iii) recommended for routine collection (see Table 1. Definitions, and Table 5. Included indicators by category) Four of the six donor agencies that were approached did not require specific indicators for reporting. The complete list of included indicators is available in supplementary materials (S1 Data).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Indicators screened and included, by category.

https://doi.org/10.1371/journal.pgph.0003843.t004

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Included indicators by category.

https://doi.org/10.1371/journal.pgph.0003843.t005

Table 6 shows the number of indicators for each domain and Table 7 provides an overview of the thematic focus of the included indicators. As the tables show, the review identified a large number of indicators measuring a broad range of aspects of child public health, with overlaps across domains and thematic focus. Operational agencies mainly measured child health status and health services for children. Nearly two-thirds of routinely collected indicators measure quality or access to care. The two donors who required indicators asked for measures that were spread more evenly across the four domains. Recommending technical agencies, partnerships, and academics placed comparable focus on health status, health services, and social determinants, and less focus on health behaviours. Nutrition indicators were particularly numerous among all categories, comprising 43% of required, 22% of routinely measured, and 23% of all recommended indicators, respectively. Child protection indicators were numerous among recommended indicators (N= 197; 39%) but notably absent among routinely collected or required child public health indicators. One of the required indicators measured the general safety of service delivery but was not specific to child protection. Similarly, 11% of recommended indicators measured an aspect of education, and 2 (14%) were required by donors, however education accounted for only 2 (1%) of routinely measured child public health indicators. Overall, 21 (14%) routinely collected indicators measured social determinants of child health. Agencies with an overt child rights mandate were more likely to measure aspects of child protection as well as social determinants of child health compared with other agencies.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Overview of included indicators by domain.

https://doi.org/10.1371/journal.pgph.0003843.t006

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Overview of included child public health indicators by thematic focus^*.

https://doi.org/10.1371/journal.pgph.0003843.t007

There were overlaps and duplications among the indicators, as well as similar indicators with slight differences in definition that create challenges for pooling data. For example, there were a total of 61 child mortality indicators, which can be broadly grouped into measures of stillbirth (N = 4), newborn mortality (N = 10), infant mortality (N = 1), under-5 mortality (N = 6), adolescent mortality (N = 1), institutional mortality (N = 3), nutrition-related mortality (N = 12), other cause-specific mortality (N = 19), gendered mortality (N = 4), and a measure of the mortality review process (N = 1). One of the 61 mortality indicators was required by a donor, 31 were recommended for routine collection, and 29 were routinely collected by operational agencies – nearly one fifth of all routinely collected indicators measured an aspect of child mortality. Stillbirths were variously defined as “including stillbirths with no foetal heartbeat at admission as well as other stillbirths”, “number of foetuses and infants born with no sign of life and born with birthweight of 1000g or more, or after 28 weeks gestation, or 35cm or more body length in a specified time period”, “number of foetuses and newborns born after 28 weeks gestation or ≥ 1000g with no sign of life”, and “foetal death after 22 weeks of gestation and prior to delivery”. Newborns were again defined variably, some indicators without a definition of the newborn and others providing definitions of <28 days, or ≤28 days, or “the first 28 days of life”. Some of the indicators measured institutional mortality while others measured population-level mortality. Some indicators were specific to a period of time while others were not. Nutrition mortality indicators similarly had overlaps and variations in measures including inpatient and outpatient, severe acute malnutrition and moderate acute malnutrition, and supplementary feeding programme deaths. The definitions were again varied and some indicators were not defined. Age disaggregations for the nutrition mortality indicators appeared to include a typo as well as variations, including no disaggregation, <5, <5 years and ≥5 years, “6-59 months, 6-23 months, 24-59 months”, and “0-5 months, 6-23 months, 24-59 months”. The other mortality indicators had similar variations in definition as well as overlaps. Table 8 provides an overview of mortality indicators.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. Mortality Indicators.

https://doi.org/10.1371/journal.pgph.0003843.t008

The pattern of similar indicators with slightly different definitions was also seen with other measures, most notably the continuum of maternal and newborn care, nutrition, and infectious diseases (S1 Data). These were also more commonly recommended and measured thematic areas. Two of the five operational agencies routinely measured under-five mortality, neonatal mortality, and stillbirth rates. As with the mortality indicators, there was significant heterogeneity in age disaggregation for all other indicators included in the review and for the definitions of children at varying ages and stages of life.

None of the operational agencies who shared indicators included measures of early child development (ECD) or disability outcomes in children, nor did the review identify recommended or required indicators for these topics. Among the recommended indicators, only agencies focusing on child protection or education recommended measurements of education (N = 55) or access to services for children with physical impairments or disabilities (N = 12). There were no recommended measures of ECD by any international organisation or partnership, regardless of the organisation/partnership thematic focus (i.e., health, public health, protection, education, or child rights). Health and public health agencies and partnerships did not recommend measurement of disability outcomes in children. There were, however, 14 indicators for which a disaggregation by disability status was recommended (S1 Data).

Agencies who provided routinely measured indicators in FCV settings placed a heavy focus on outcomes for mortality, nutrition, the continuum of maternal and newborn care, and infectious diseases (Table 7). A broad range of indicators for clinical child health services that did not fall into one of the other categories was coded as “paediatrics”; these paediatrics indicators account for one-fifth of the routinely measured child public health outcome indicators. There was also a heavy focus on adolescent sexual and reproductive health (n = 15, 9.9%), and broadly in the social determinants of child health (n = 21, 13.8%), again with a notable lack of measurement of ECD or disability outcomes. The two noncommunicable disease (NCD) indicators that were routinely measured are most relevant for adults (unspecified type of diabetes, and hypertension), however, because they included age disaggregation for children, they met inclusion in this review. There were no routinely measured childhood NCDs such as asthma, epilepsy, type 1 diabetes mellitus, or childhood cancers.

Discussion

This scoping review provides a comprehensive overview of the data landscape for child public health situation, outcome, and impact indicators in FCV settings. A key finding of this review is that a large number of child public health indicators are routinely measured in FCV settings in spite of the contextual challenges. Routinely collected indicators by operational agencies included all four domains of health status, health service, social determinants of health and health behaviours. Most of the indicators that are included in this review were specifically recommended by technical agencies and partnerships, required by donors, and/or used in humanitarian contexts. As such, they are indicators that are recommended or actively measured in places characterised by interruptions in access, subject to short-term funding and/or interrupted funding for interventions, and subject to logistical and technical challenges that relate to the specific crisis and setting.

The sheer number of indicators that are recommended for routine collection by technical agencies and partnerships is striking. Given the paucity of routinely available data on child public health worldwide and the well-described challenges for data collection in FCV settings [9,20] it is concerning that over 500 indicators are recommended by major supporting agencies and partnerships for routine collection. It is important to note that the indicators included in this review are limited to situation, outcome and impact indicators. Output and process indicators, which often form the bulk of reporting by operational agencies [9], were not included in this review. The very large number of recommended indicators would require a significant amount of human capital, financial resources, and infrastructure to measure, analyse and interpret. It is not realistically achievable by governments and nongovernmental organisations working in FCV settings and it suggests competing priorities between sectors and limited coordination of activities across sectors.

The current collection of indicators is also not appropriate, as many indicators overlap, and some are similar but not exactly the same. It not clear whether all of the recommended indicators would be useful for decision-making by operational actors. It is noteworthy that specific sectors, such as nutrition and child protection, recommended particularly large numbers of indicators (Table 7), suggesting a lack of prioritisation of topics by actors working within the respective sectors as well as lack of coordination and collaboration for prioritisation, standardisation, and harmonisation of indicators across sectors and agencies. The use of intersectoral indicators would help to both reduce the number of indicators and can provide a more human-centred overview of child public health. Finally, the large number of recommended indicators may be confusing for monitoring and evaluation teams, and likely reinforces variations in measurement. Heterogeneity in measurement by teams and agencies reduces the reliability and validity of the data and makes pooling of data challenging. As such, measurement variability hampers meaningful accountability to populations, by limiting our ability to understand whether an intervention was safe, appropriate, and effective for the target population. Data practices that hamper accountability serve to reinforce a colonial approach to aid, where the outcomes and impacts of interventions on the affected population are not prioritised.

Establishment of consensus about priority topics across sectors that address child public health, alongside standardisation of prioritised indicators could both drastically reduce the number of recommended indicators and simultaneously improve indicator quality, and ultimately, accountability to affected populations. Indicator prioritisation should consider usefulness for operational decision-making in the dynamic environment of FCV and humanitarian settings. Prioritisation of indicators may be most effective if based upon decision-making needs.

It is important to consider that standardisation of indicators may not be feasible for each type of data source. Child public health data are collected in FCV settings by a range of methods, such as population-based surveys, review of clinical information, facility assessments [21], rapid needs assessments, qualitative interviews with facility staff, patients and/or affected populations, and quality of service assessments. Each type of data source contains different information that may be used for baseline data, monitoring changes in population needs, and/or evaluating the impact of interventions. However, definitions of both the numerator and denominator may differ by data source. For example, the denominator in a population-based survey, a census, or the quantification of a population in need quantified by UN OCHA, are likely to differ substantially. The accuracy of denominators may be uncertain and difficult to assess in FCV settings, which are often characterised by weak vital registration systems, fluid population movement, displacement of populations across administrative or even national borders, barriers in access to the affected population, and lack of expertise and resources needed to estimate population size [22]. Some of the indicators included in this review either lacked a definition for the denominator entirely or provided a definition that required significant interpretation. Some indicators suggested multiple, differing data sources for denominators. These practices limit the reliability of indicator measurement and preclude pooling of data. Efforts to establish consensus on a shortlist of priority child public health indicators for routine measurement in FCV settings must therefore consider the types of data sources that will be used and develop clear definitions of the numerator and denominator.

It is encouraging that 152 indicators were feasible for five agencies to routinely collect despite the contextual challenges in FCV settings (Table 7). It is important to reiterate that these indicators are actively measured by five large operational nongovernmental agencies working in FCV settings (S4 Text) and is likely to differ from the kinds of measurements made by smaller organisations. However, the finding indicates that the measurement of child public health outcome, impact and situation indicators is feasible despite the significant and changeable contextual challenges. This lends further support to the argument for the need to develop adaptations of “gold standard” outcome indicators for measurement in settings where the highest standard of measurement is not possible to consistently collect. In the absence of existing routine data on child public health in FCV settings, researchers and the public health community have developed methods to create estimates based on exiting data, such as data from the Demographic and Health Surveys [23, 24]. Our findings suggest that in many settings, reliance on estimates is not necessary – that it is possible to collect data in these settings. There are notable exceptions, in contexts where health systems have been severely destroyed and/or access is restricted, such as the current situations in Gaza and Sudan, respectively [25, 26]. In such contexts, the use of adapted indicators for proxy measures is a realistic solution that could enable ongoing routine measurement of outcomes. In addition to the creation of adapted indicators, there is a need for global consensus on a shortlist of prioritised indicators for routine measurement across agencies and contexts, and a system to support the pooling and sharing of these data.

During the course of this review, a number of indicator banks were also identified, some containing hundreds of indicators for a range of sectors and topics [27, 28]. Indicators from these banks were not included in the review because they were not recommended for routine collection across FCV settings but serve as a resource of potential measures for actors, often with specific focus areas, such as infectious disease outbreak. A shortlist of priority indicators feasible for routine measurement in FCV settings with specific adapted proxy indicators, supplemented by a bank of high-quality indicators for routine measurements may be a preferable alternative to the current longlists.

It is also encouraging that only two of the six approached donors, USAID BHA and ECHO, had requirements for the measurement of child public health; the other donors advised that they do not have specific requirements, but that operational teams are asked to propose measures for reporting. Interestingly, one donor agency responded to our enquiry with a request for recommendations about child public health outcome indicators that they should consider requiring. Donor reporting requirements are often cited as the main driver for indicators measured by operational teams [29, 30]. The findings from this review suggest that operational agencies and child public health professionals have an opportunity to influence donor requirements and expectations for data collection. To this end, a core list of child public health indicators for measurement in FCV settings would be of value.

If a core list of child public health indicators were routinely measured by all actors, including ministry, nongovernmental organisations, and other operational groups, and if these data were reliably measured, we could understand the corresponding child public health status and needs of children in a given crisis context and globally. For the first time, we would be able to base our response to crises on actual data on child public health, rather than estimates.

With over 500 indicators recommended for routine collection, it is unsurprising that there are overlaps and similarities in recommended indicators from technical organisations and partnerships. While a detailed assessment of quality was beyond the scope of this review, the included indicators were assessed for the degree of interpretation required in order to measure them. Less than half of all indicators were of sufficient clarity that they can be measured without some degree of interpretation (Table 5. Included indicators by category). Given the vast number of indicators recommended by technical agencies, partnerships and academics, it is concerning that more than half of them required either some or significant interpretation to measure. Several of the recommended indicators contained typos in the definition or title that led to discrepancies between the title and the definition and led to uncertainty in what was meant to be measured. While typos are unfortunate, they are easy to correct. A more challenging correction would be the development of consensus on a shortlist of priority child public health outcome, situation, and impact indicators for measurement in FCV settings. The latter is essential if we are to hold ourselves accountable to children and populations affected by crisis.

A similar trend in was observed in the routinely collected indicators, with just under half requiring at least some interpretation. When we enquired about definitions from one agency, we were informed that indicators were deliberately not defined and that country teams were encouraged to use “common sense”. The failure to define what is being measured renders the entire measurement process unreliable. This relaxed approach to tracking routine indicators suggests a systemically colonial approach to crisis response without accountability for the resulting outcomes from the services that are provided.

Age disaggregation practices and recommendations vary widely and were even found to vary among indicators recommended or used by the same organisation. There were no clear trends from any sector, organisation, or donor, except that nutrition indicators are most often recommended for measurement in children aged 6-59 months. Even the indicators for infant and young child feeding used variable age disaggregation, all of them measuring some group(s) under 2 years of age, but with significant variations in the definition of these groups. Diaz et al have published a strongly argued recommendation for standardised age disaggregation [31]; widespread adoption of standardised age disaggregation across all public health programming in FCV settings would improve the ability of governmental, nongovernmental, and academic institutions to collect, compare, share, and pool data.

The only NCD measures routinely collected by operational teams focused on diseases that are more commonly seen in adults. The lack of child specific NCD indicators routinely measured suggests that the decisions for what to measure are made by individuals or teams that have limited child public health knowledge. With children accounting for a significant proportion of the population in crisis-affected areas - and an even higher proportion of health care contacts in humanitarian settings [32] - a lack of child public health expertise among international NGOs is unacceptable. The findings of this review are consistent with other papers that have identified a lack of child health expertise in humanitarian settings more generally, and a need to train our colleagues who are tasked with providing care and designing and delivering interventions to support children in crisis contexts [33].

Sexual and gender-based violence (SGBV) indicators tended to be recommended by a range of different sectors, indicating a broader recognition across sectors of the relevance of SGBV to child public health. Of the SGBV indicators, mental health and psychosocial support (MHPSS) and access to care were the main areas of focus. Agencies routinely collecting MHPSS indicators tended to focus on SGBV and on change/improvement in mental health outcomes after service provision. Recommended indicators explored MPHSS service needs, outcomes, and referral pathway functioning, as well as caregiver and teacher MHPSS needs and support. The inclusion of indicators on caregiver and teacher MHPSS among recommended indicators is encouraging, as it indicates a recognition of caregiver mental health as a major determinant of child health, development, and safety.

The lack of indicators measuring early child development (ECD) and disability indicates a major technical gap in child public health expertise by operational agencies, donors, and technical organisations supporting programmes in FCV settings. There were a small number of indicators on injury (N = 5) and disability (N = 9) recommended by child protection actors. No measures of ECD were identified by the review. Early child development and disability are both markers of child health and also determinants of child health and protection; they have significant impacts on child health and children’s risk of violence and exploitation throughout childhood, adolescence, and adulthood [34, 35]. The cumulative harmful effect of early childhood adversity on physiological, neurological, and socioemotional development are well described [4]. As such, developmental status and disability have important implications on child health and child public health needs and outcomes in all settings, and especially in FCV settings where children are at higher risk of experiencing multiple forms of severe adversity. For example, it can be argued that children with severe acute malnutrition (SAM) have experienced multiple forms of severe adversity that resulted in them developing SAM. It is well documented that children with direct exposure to armed conflict experience a range of severe forms of adversity including witnessing and/or being subjected to physical, psychological and emotional violence, lack of access to basic needs, loss of one or both parents, forced displacement, and early marriage, to name just a few. These children are also known to be at significant risk of disabling injuries from combat activities, including blast injuries, penetrating wounds including penetrating head injuries from bullets, shrapnel and other objects, crush injuries, and burns [36]. Children can also be injured by combat activities long after hostilities have ended; up to 50% of injuries from unexploded ordnance are in children, sometimes decades after the end of the conflict [37]. These are just a few examples of why ECD and disability are important markers of child health and child public health. It is concerning that ECD and disability are not prioritized in measurement of child public health outcomes in FCV settings. Given the pervasive and potentially profound impacts of crises on child development and disability, the consequences of failing to measure ECD and childhood disability are significant. At best, programmes and funding decisions will be inadequately informed. Prioritisation of the use of resources may be inappropriate or may fail to address the needs of children. At worst, ill-informed programme design and implementation may reinforce existing inequalities, exacerbate health harms to children, or cause further harm.

Limitations

This review compiled recommended indicators from large organisations and partnerships that provide technical guidance and recommendations for operational agencies. The literature review was restricted to English language publications. Similarly, the routinely collected indicators were from a handful of large international nongovernmental organizations, and reporting requirements were sought from six large government and private donors that fund interventions in humanitarian contexts. As such, the indicators included in this review are by no means an exhaustive list; they are more representative of large and well-funded institutions and partnerships that are primarily based in Europe and North America and support and/or operate in FCV settings.

Inclusion of qualitative indicators was outside of the scope of this review. Both quantitative and qualitative data are necessary to support safe, relevant, and effective program decision-making. Further research on qualitative child public health measures recommended, required, and actively measured in FCV settings would be helpful and provide further depth to our understanding of child public health data practices, gaps, and needs in FCV settings.

The paucity of ECD and disability indicators measured by operational teams may be in part due to bias from agency contacts who shared indicators with the research team. The lists of routinely collected indicators were provided primarily by health and nutrition staff; our sample may be missing some non-health or nutrition indicators that are routinely collected. During data collection, the study team provided written definitions of child public health and the domains included in this review, as well as verbal explanations to participating agencies. If the shared indicators were focused on child health and nutrition, follow up was made to enquire about other indicators to measure child public health. However, the siloed nature of aid work may have impacted the kinds of indicators we received. It should be noted that three of the five organisations that shared their routinely collected indicators included education, protection and/or MHPSS indicators. If there is a gap in our dataset for child public health indicators not strictly falling into health or nutrition, it would suggest that public health workers may not recognise the child public health work done by teams in other sectors outside of health and nutrition and/or the value of the data they are collecting, and a more general lack of child public health expertise.

Conclusions and recommendations

The routine measurement of child public health indicators in fragile, conflict-affected, and vulnerable settings is not only possible, but it is already being achieved. Measurement of child public health in FCV settings is characterised by an excessive number of recommended indicators, duplication, poor definitions, and siloed sector-specific focus. Improving the measurement of child public health in FCV settings requires recognition of the child public health work done by actors in sectors other than health and nutrition. While there are recommended and routinely collected indicators on social determinants of child health, important gaps remain. The most conspicuous gaps include early child development and disability, which are both child public health outcomes and also determinants of child health, wellbeing, safety, and life trajectory. There is a need to establish consensus on priority child public health topics for measurement and a shortlist of high quality, appropriate and relevant measurements that governmental and nongovernmental actors can be reasonably expected to measure routinely in FCV settings. The prioritisation of indicators should be based on decision-making needs in a dynamic context. Proxy indicators should be developed for periods where routine measurement is hampered. Clearly defined indicators and consensus on age disaggregation would enable pooling and comparison of data that could inform an evidence base for data-driven decision-making in FCV settings. Child public health expertise and data are not a “nice to have” in FCV settings, where a significant proportion of the population are children; they are essential for appropriate and relevant measurement of child public health, guidance on interventions, evaluation of interventions, and accountability to affected populations.

Supporting information

Acknowledgments

The authors would like to thank Bethany Tapster at the Leeds Teaching Hospitals NHS Trust for support in developing the database searches. We would also like to thank the organisations that spoke to us and shared their indicators.