Global development of children’s palliative care: An international survey of in-nation expert perceptions in 2017

Background: The growing interest in tracking the global development of palliative care provision is not matched by research on the development of palliative care services specifically for children. Yet it is estimated that worldwide, 21 million children annually could benefit from the provision of palliative care. We report on a global study of children’s palliative care development and offer suggestions for further improvement in design and method. Methods: Primary data on the level of children’s palliative care development in 2017 was collected from in-country experts through a specific question in an online questionnaire that sought to measure the overall level of palliative care provision globally. Countries were assigned to one of six categories on the basis of the responses obtained. Conflicting responses from the same country were resolved with reference to a hierarchy of preferred respondents. Results: Our data allowed the categorisation of 113 countries, accounting for 65% of the global population aged under 20. Number of countries (% of global child population) in each category were as follows: 1) no known activity, 21 (4%); 2) capacity-building, 16 (24%); 3a) isolated provision, 55 (30%); 3b) generalized provision, 5 (1%); 4a) preliminary integration into mainstream provision, 14 (8%); 4b) advanced integration, 7 (2%). Conclusions: Children’s palliative care at the highest level of provision is available in just 21 countries, accounting for fewer than 10% of the global population aged under 20. It is concentrated in high income settings, whilst the majority of the global need for such care is in low- and middle-income countries. Our study is a useful tool for global advocacy relating to children’s palliative care and a stimulus for the creation of improved indicators to measure it at the country level.


Introduction and context
Children's palliative care -global context The World Health Organization (WHO) has defined palliative care for children as 'the active total care of the child's body, mind and spirit, and also involves giving support to the family' 1 . The WHO states that this should begin when a life limiting condition is diagnosed and should continue whether or not the child receives treatment directed at the disease. It requires health providers to evaluate and alleviate the child's physical, psychological, and social distress; and to be effective, it should be supported by a broad interdisciplinary approach that includes the family and makes use of available community resources. WHO maintains that palliative care for children can be successfully implemented even if resources are limited and can be provided in tertiary care facilities, in community health centres, as well as in children's own homes 2 .
One-third of the global population is aged under 20 years. This is the age range to which WHO applies the idea of children's palliative care in the context of life-limiting and life-threatening conditions. Yet whilst 98% of the global need for children's palliative care occurs in low-and middle-income countries, the majority of such care is available only in high-income countries 3 . It is estimated that around the world, 21 million children annually could benefit from some form of palliative care 4 . The number includes more than 8 million children with complex problems that require specialised attention. Of these the most common difficulties relate to perinatal conditions, followed by congenital anomalies, Human Immunodeficiency Virus (HIV)/ Acquired Immune Deficiency Syndrome (AIDS), and cancer 5 .
Currently we know very little about the scale of development of children's palliative care services in the global context. This stands in contrast to the growing body of information and research that is plotting patterns of growth and development around the world in palliative care for adults. We therefore focus here on this knowledge deficit relating to children's palliative care, report on a recent study that has shed some light on the issues, and reflect on the methodological and conceptual problems involved in research of this type. While the evidence we present is far from definitive, we see it as an important contribution to a crucial and emerging field of enquiry.
Mapping global palliative care development For more than a decade, a series of studies has monitored the level of palliative care development in all the countries of the world -beginning in 2006 6 , followed up in 2011 7 and most recently updated for 2017 8 . These studies have contributed significantly to advocacy, planning and monitoring for the improvement of palliative care worldwide. They have also sought progressively to develop more robust methods for measuring the country-level status of palliative care provision, through continuing identification and refinement of the appropriate indicators 9 . The first study allocated each country to one of four categories of development, using data synthesised from a variety of academic, professional and 'grey' literature, with expert opinion used as a substitute where necessary. The second study used a refined six category classification and was based on the identification of in-country experts or 'champions' who were asked for their opinion on their country's level of development. The most recent iteration of the study develops the method further, again using six categories of development, but now with more detailed information gathered from a survey of national experts across 198 countries, and based on a set of 10 palliative care indicators derived from the literature.
Children's palliative care -current evidence Despite these advances in measuring the overall level of palliative care provision globally, there remains limited evidence on the development of children's palliative care. The first two global mapping studies, for example, attempted to measure only the overall level of palliative care development in each country, with no specific focus on services for children.
Knapp et al. 10 , however, looking specifically at palliative care for children, used a systematic review of peer-reviewed and other published literature, as well as information on numbers of what the authors term 'paediatric specific organizations', to assign countries to the four categories of overall palliative care development used in the first global mapping study. This placed only 11 countries at Level 4, the highest category of palliative care development, while 19 countries were identified as being at Level 3 (isolated provision) and 36 at Level 2 (capacity building with no operational services). The majority of countries -126 (65.6%) -were categorised as being at Level 1, indicating that there was no reported evidence of children's palliative care. Although this process used two reviewers to assign levels of provision, with further review by an expert panel, the work nonetheless shared the limitations of the 2006 global study in relying on published data (limited to English-language literature) supplemented by expert opinion, sometimes from outside the country.

Aim of this work
The aim of the work presented here was to obtain a first global overview of the national development of children's palliative care based on the knowledge of in-country experts. With this in mind, we included two specific questions in the third global mapping study that focussed, respectively, on the level of development of palliative care for children and on its implications.

Amendments from Version 2
In version 3 of this paper we respond in particular to the three points made by reviewer 3. First, the title of the paper has been changed, as recommended. Second we have introduced two new maps which show the level of development of palliative care by 1) maintaining the relative size of areas across the map and 2) re-sizing the area of countries according to population. Third, we have added an extended paragraph detailing how a future, improved, study of children's palliative care development around the world might be designed and delivered.
Any further responses from the reviewers can be found at the end of the article

REVISED
The paper proceeds as follows. First, we set out the research methods and design of our most recent study, along with the protocol for assigning each country, where possible, to one of six categories of children's palliative care development. Second, we present the results of this process and compare them to the relevant countries' overall level of palliative care provision and to the results of the study of Knapp et al. Third, we discuss the strengths and limitations of our approach, with suggestions for future improvement.

Ethical statement
The study was approved by the University of Glasgow College of Social Sciences Research Ethics Committee. Approval was granted on 15 January 2018 (Application No: 400170065). Prior to data collection, the purpose of the study was explained to all participants and written consent for participation in the study was obtained from each participant before access to the questionnaire was granted. Participants were informed that they had the right to withdraw from the study at any point in time, without any repercussions.

Data collection
The design and methods of the third global palliative care mapping study have been presented in detail 11 . Data were collected via an online questionnaire completed by in-country experts actively engaged at the national level in the development, delivery and co-ordination of palliative care activity. These respondents were identified in consultation with international and regional palliative care associations. The survey was administered in 198 territories, comprising the 193 Member States of the United Nations (UN), two Observer States, along with Kosovo, Somaliland, and Taiwan, China. A total of 560 experts from 179 (90%) countries for which contacts could be found were asked to complete the questionnaire. For 19 (10%) countries we were unable to identify an expert contact.

Categorisation
While the questionnaire included multiple questions linked to a variety of indicators that enabled us to assess the overall level of national palliative care development, two questions focused specifically on the level of development of palliative care provision for children. Respondents were asked "Which of the following categories best describes palliative care activity related to children in your country?" A set of detailed category descriptors was then provided (Table 1).
A follow up question -"Please provide here any additional information on children's palliative care services in your country you would like us to know" -allowed respondents to add more information or comments about the level of children's palliative care. Figure 1 explains the categorisation process. Where there was only one completed response per territory, or where there were multiple responses that were in agreement, the territory was allocated a category of children's palliative care development on that basis. Where there were two or more responses with conflicting assessments of the most appropriate category, these were resolved where possible by privileging the assessments of those survey participants closest to the top of the following hierarchy: I) Representatives of the national in-country hospicepalliative care association or nearest professional association (e.g. society for palliative medicine or hospice forum). The person should have an established administrative and/or leadership role in the organisation, making them a reliable source of information.
II) Academic experts with known interests and research experience in hospice-palliative care development in-country and/or beyond, as evidenced by peer-reviewed publications. The person should have an established academic role in hospice-palliative care research or education, making them a reliable source of information.
III) Policy specialists (in or outside government) with experience of and/or responsibility for hospice-palliative care delivery in-country. The person should have an established policy role relating to hospice-palliative care, making them a reliable source of information.
IV) Palliative care representatives, academics or policy specialists from outside the country but with direct knowledge of its hospice-palliative care provision, making them a reliable source of information.
Where it was not possible to resolve conflicts on this basis, that is where survey participants offering differing assessments of their country's level of children's palliative care had similar roles -then the country was excluded from the analysis.

Results
The global development of children's palliative care Using this process, an assessment of the level of children's palliative care development was reached for 113 (57%) of the 198 countries included in the study (Table 2, Figure 2). These 113 countries account for 65% of the total global population of children -defined here as being aged 0-19 in line with the WHO convention, so including infants, children and adolescents.
In Figure 2 and Figure 3 we present world maps showing the distribution of the different levels of children's palliative care development. The cartography was made using ArcGIS Pro (ESRI), version 2.6. The layer used for representation is 'Admin 0 -Countries, 1:50m', obtained from Natural Earth. For both maps the equal earth projection was used, which maintains the relative size of areas across the map. For Figure 2 a blue-scale choropleth map was used to visualize the level of development of children's palliative care and a reddish colour was used to denote 'no known activity'. For Figure 3 the same choropleth map was used in combination with a cartogram made by the Gastner-Newman algorithm 12 . The cartogram resizes the countries relative to their population, resulting in countries with high population being bigger in size. A country in this category shows evidence of wide-ranging initiatives designed to create the organisational, workforce, and policy capacity for the development of palliative care services for children, although no service or specific program has yet been established. There are some developmental activities including attendance at, or organisation of, key conferences, personnel undertaking external training in palliative care, lobbying of policy makers and Ministries of Health, and emerging plans for service development.
Category 3a: Isolated children's palliative care provision A country in this category is characterized by the development of children's palliative care activism that is still patchy in scope and not well-supported; sources of funding that are often heavily donor-dependent; limited availability of morphine; at least a service or program can be identified by other professionals in the country as a best practice model for palliative care for children; there are a few children's palliative care services or specific programs, but they are limited in relation to the need of the population. This is used to emphasise the overrepresentation of population in countries with palliative care levels of 'isolated provision' or lower. Dots were used in both maps to visualize various small countries that were included in the study.
Only seven countries, accounting for 42M children (1.7% of the world total), were placed in Category 4b, with the most advanced level of integrated children's palliative care. A further 14 countries were in Category 4a (190M, 7.5%), with preliminary integration of children's palliative care into mainstream services.
In Category 2 were 16 countries, with evidence of some capacity building activity in children's palliative care (610M, 24%).
In the lowest category (1) were 21 countries with no reported children's palliative care activity. In addition, we added to this category 28 countries for which no survey was returned, and 28 countries for which the children's palliative care question was not completed, on the basis that this suggests there is no evidence of any palliative care specifically for children in these countries. Taken together, the countries in this expanded category 1 account for 265M children aged under 20, 11% of all the world's children.
For a further 29 countries, there were multiple responses that provided a conflicting assessment of the level of children's palliative care, and where this conflict could not be resolved with reference to the hierarchy set out above. They were therefore not assigned a category. These account for 646M children, 26% of all the world's children. In the majority of cases these had survey responses from two or three different in-country experts (in two cases there were more than three responses). In 14 of the 29 countries, assessments of the level of children's palliative care varied by one category, and in 10 cases by two categories. The remaining five countries had more significant conflicts, with respondents disagreeing by three or four categories (See Supplementary Table 1, Extended data 13 ).

Comparison with overall palliative care level
There is a broad alignment between countries' overall categorisation of palliative care development, already reported 7 and based on a range of indicators, and their specific categorisation for children's palliative care as presented here. Table 3 shows that all of the countries in Category 4b for children's palliative care are also in 4b overall, as are 10 of the 14 countries in Category 4a for children's palliative care. There is, however, no firm relationship between the national level of overall palliative care development and that of children's services.
There are a number of outliers, in particular a small number of countries in category 4b overall that are in categories 1 or 2 for the level of children's palliative care. In addition, the overwhelming majority of those countries for which there were no data on children's palliative care were placed in the lowest category overall, indicating no known or reported palliative care activity.
Only two of the countries with responses that reported no children's palliative care activity were also placed in the lowest category for their level of overall palliative care development.
On the other hand, of those countries for which no information on children's palliative care was returned, the vast majority (43/56) were in the lowest category overall.

Comparison with other country level indicators
This data can also demonstrate the relationship between the level of children's palliative care development and the United Nations Human Development Index (UNHDI), World Bank Income Level (WBIL), Universal Health Coverage Index (UHCI) and WHO region. Of the seven countries at the highest level of children's palliative care development, all are at high or very high levels of UHCI, WBIL, and in the top two quintiles   Table 4 shows how the 113 countries for which a positive categorisation was possible on the basis of our 2017 survey, were classified (on a four-category scale) by     Table 5 shows some indicative quotes from respondents, arranged by their country's categorisation, that demonstrate the type of commentary provided. Respondents provided broad overviews of institutional frameworks and social contexts, further detail on the number and type of medical facilities and specialists providing palliative care for children, the geographical spread of these services, and narratives about the progress of children's palliative care development in their country.

Strengths and limitations
The research presented here is the most comprehensive attempt yet to map the country-level development of children's palliative care around the world. In basing this assessment on responses from in-country palliative care experts from 113 territories, the method moves beyond a reliance on published data and the judgement of external parties. As part of a broader study of global palliative care, this therefore represents a significant contribution to evidence and understanding about the spread of palliative care provision for children.
Nevertheless, there remain a number of limitations with this approach and these should be addressed in future research.
• Although this method -based on the assessment of 'in-country' experts with detailed knowledge of palliative care in their territory -can be considered more robust than previous approaches, categorisation on this basis is nevertheless informed to some extent by the professional judgements of survey participants. Such self-reported judgements cannot be independently verified, and some participants may have over-or under-estimated their country's levels of development. There are, for example, instances of countries at the highest level of overall palliative care development, of human development and income that have been identified as having a very low level of children's palliative care development. The deficiencies of this approach could therefore potentially be strengthened through triangulation with other sources of evidence, though we lacked the resources to do this in the present study. However, while confidentiality principles do not allow us to give more details, many survey responses came from impeccable organisational sources and were signed off by senior persons, sometimes of global standing in the field of palliative care, and very well placed to provide an assessment of the situation in their country.
• Furthermore, some readers may raise the question of relevant expertise in the respondents to our survey, and take the view that national experts in palliative care will not necessarily have in-depth knowledge of palliative care services available for children. This is a reasonable assertion but could be subject to challenge, particularly in poorer countries where there is a less developed divide between 'adult' and 'children's' palliative care. We acknowledge the distinct and perhaps contested status of children's palliative care in our discussion of definitional issues below; given this, we propose that a survey specifically about children's palliative care would be the desirable next step in developing understanding of this area. It would, however, be unreasonable to discard the assessments of these palliative care There are not any activities for palliative care for children in my country.
Palliative care for children is usually provided by the programmes providing palliative care for adults. No specific paediatric palliative care programmes are currently available in the country.
There is only one children's general hospital. This hospital provides non-specialized palliative care services to children in need e.g. HIV/AIDS and malnutrition.

2
Activities are begun by NGOs in some places. Professionals (doctors, nurses, management) in the oncological department in [capital city] Children's Hospital make efforts to establish children's palliative care. There is a symposium about children's palliative care held every year. Also, the recognition of children's palliative care by the national association has achieved a significant impact. There is also an active palliative care committee at the Ministry of Health.
We created this year a hospital committee for children's palliative care; decrees have been signed for the national hospital and we are now working on different protocols to formalize them.

3a
There is only one children's onco-hematological hospital. This offers chemotherapy and symptom control for children in need, but there is not any palliative care or hospice for children in the country.
There are a few complete palliative care services but they are in big cities. Palliative care for children is growing fast but in a very heterogeneous way at the national level. There are provinces/states that do not have any paediatricians trained in palliative care and other provinces who have several teams in big cities.
There are several specialized paediatric onco-hematological hospitals that provide specialized palliative care services for children.
There is only one NGO providing home palliative care to children and young adults as well as psychological support to parents, siblings and relatives.
There is only one paediatric palliative care service; it is a well-established service; provides 24/7 service with inpatient and outpatient plus one home team The service is only for the patients that are seen at the centre, mostly cancer patients. Trying to help neighbouring countries by providing training to physicians and nurses.
Progressing slowly. Health care professionals need to be sensitised and trained to refer and support children with palliative care needs. Still lagging behind adult service provision particularly in rural areas.
The children's palliative care services are currently limited to two University Teaching Hospitals but there is a Government willingness and plan to spread the services to District Hospitals in the nearest future.
Hospitalized children suffering for cancer have a palliative care support program but this is generally missing for other specialties.

3b
The Ministry of Social Affairs and Health recommendations for organizing palliative care includes a recommendation for building up children palliative care network and services. The official group for implementation of these recommendations and another group to revise the legislation/need for a new legislation of end-of-life and palliative care include plans for children's palliative care. University Hospitals have home care teams for children, but in general there is no official plan for children palliative care so far.
There are palliative care standards at national level in children's care. Palliative care specialists are trained, before they are certified, both in children's and adult palliative care on a one year course.

4a
Services for children are provided by the hospital, and upon discharge, quality care is given. It should be noted that there is not a high percentage of children in need of palliative care. Should a case arise, families are very keen to give primary care with our monitoring support.
All 6 departments of paediatric oncology in the leading university hospitals have developed an advanced service for palliative care.
We have one children's hospice. Some children are taken care of in advanced home care teams around the country.
In most cases families take the total responsibility and provide the end of life care to their dying child. Despite availability of opiates in country, there isn't proper system of prescribing and consuming opiates for reducing pain of patients. Parents get pain killers, even sometimes opiates, on an ' over the counter' basis from pharmacies to reduce pain of their loved one. experts on the grounds that they might not be specialists in children's palliative care, especially when the latter may not even be present in some countries.
• Related to this, the survey question did not define 'palliative care activity related to children in your country', meaning there was the potential for the question to be interpreted in different ways, influenced by different understandings of what constitutes palliative care and who are considered to be 'children' in different national contexts. We explore this in more detail in the section below on matters of definition.
• A significant proportion of countries included in the global mapping survey have been excluded from these results on the basis that participants provided conflicting categorisations that could not be reconciled. The exclusion of these countries limits the extent to which the results presented here can be considered a comprehensive global assessment of palliative care development. While there are no straightforward ways of addressing this, a number of possible approaches may be considered. At its simplest, this could take the average value of multiple categorisations. Alternatively, more detailed information could be sought about the characteristics of survey participants to allow a judgement to be made on which were best placed to assess their country's level of development, or the views of external experts could be sought to moderate the proposed categorisations. We do not consider this approach self-defeating, but regard it as an honest and open statement of limitation which we mitigated by not including the relevant countries in the final analysis.
• In some instances, it was not possible to identify palliative care experts in-country, and in others, where eligible respondents were identified, they did not complete the questionnaire or the specific question relating to children's palliative care.
• Data limitations were compounded by language constraints. Questionnaires were only available in three European languages (English, Spanish and French). This both impeded communication with reluctant respondents and might also have added to the barriers for non-English speakers willing to complete the questionnaire. Terminological issues may also have been further exacerbated by translation factors.
Some readers may also question the grading assigned to their own country in our analysis. In previous iterations of the world map of palliative care, where expert opinion has been sought, we have been gratified by how few of such concerns have been raised. In the present study, and within the bounds of research ethics approval, we would be happy to have dialogue with those questioning their country grading, or indeed who may be able to shed light on countries where we did not give a grade due to conflicting returns from in-country respondents.
We have given careful thought to how a future study might be made more robust than our current attempt. One starting point for this would be to establish a specific collaboration with experts in children's palliative care development, around the world, building on our existing links with the Worldwide Hospice Palliative Care Alliance 14 . Such a collaboration might be with an organisation like the International Children's Palliative Care Network (ICPN) 15 . This approach has already yielded good results in the WHO-Europe context 16 . Individuals in the collaboration would then constitute a panel for a bespoke survey of children's palliative care in each country. In some countries it will be difficult to identify such experts, but this would be a finding in itself. The indicators to be used in such a survey would be determined by use of a consensus process with the collaborative members. Thereafter and in parallel with the in-nation expert survey, it would also be possible to survey, in partnership with WHO, representatives from national governments who hold responsibility for children's palliative care. The 'experts' survey might concentrate more on indicators of service delivery and outcomes, education and the 'culture' of children's palliative care. The 'government' survey might give greater weight to policies, service availability, outcomes and workforce capacity. Such approaches might still require moderation of the data, and would require more investigative techniques, with feedback to and further questioning of respondents with regard to 'conflicting' data. We are optimistic about this line of development, which has been prompted by reviews of earlier versions of our paper.

Matters of definition
Children's palliative care is a complex domain that is similar to, but also distinct from, adult palliative care. Understanding the specificities of children's palliative care, and situating it globally, requires attention to the gamut of life-threatening and life-limiting illnesses that can affect children and youth, myriad culturally diverse conceptions of childhood, and the vast range of national, ethnic, and legal contexts of care provision.
The diagnostic categories, illness trajectories and prognosis that fall under the umbrella of children's palliative care are vast, with significant variation in severity, symptoms, prognosis, and disease trajectories. Some illnesses requiring children's palliative care are 'life limiting' (i.e., those with no reasonable hope of cure and from which children will die); others are 'life threatening' (i.e., those for which curative treatment may be feasible but may fail). These conditions have been divided into the following four categories for the purposes of research and to better focus care: 1. Life-threatening conditions (e.g., infectious diseases, cancer, organ failure and transplant, long-term ventilation); 2. Conditions in which premature death is inevitable but intensive treatment can prolong life (e.g., cystic fibrosis, Duchenne muscular dystrophy); 3. Progressive conditions without curative treatment options (e.g., metabolic conditions, incurable cancer); 4. Irreversible non-progressive conditions causing severe disability (e.g., cerebral palsy, brain or spinal cord injury) 17 .
Children's palliative care therefore includes a greater diversity of illnesses and a longer duration of survival than adult palliative care 18 . Many of these children live with physical, neurological and cognitive impairments, therefore requiring intensive family and/or institutionalized caregiving 17 . Advances in rehabilitation and disability studies have shifted focus beyond functional impairments to better appreciate the interplay between a child's capacities and the accommodations available in their local environment.
Conventional biomedical 'paediatric' models parse the category of childhood into neonates, infants, children, and adolescents based upon psychological developmental models. Children's palliative care extends beyond these categories starting with the unborn child (e.g., perinatal palliative care) and into young adulthood (e.g., many children will live with life-limiting illnesses beyond 19 years old). Children's palliative care, as defined by leading authorities -e.g., the WHO 21 , and the UK's Together for Short Lives 17 -is attentive to the family as the unit of care, offering support for the child as embedded within a family unit. This support starts with diagnosis, and proceeds throughout the illness, to end-of-life care and death, and beyond into the bereavement period. Understanding who is important to the child's life and care (e.g., extended family, parents, sibling, community), and at what age and for which kinds of treatments the child should be afforded some agency, will depend on local logics, legal systems, and cultural practices.
While conventions in the global north are to provide children's palliative care through inter-professional and interdisciplinary teams, services in rural contexts as well as in low-and middle-income countries are often without access to specialist care. Instead, care is often provided through child-focused general health professionals, with the assistance of adult-specific palliative care specialists if available 22 .
All of these dimensions challenge our ability to map and categorise levels of children's palliative care development at the national level and globally. Significant further work is needed to generate indicators of children's palliative care development that result in measurable outcomes but at the same time incorporate the nuanced and complex world of palliative care for children. The next step in developing a more comprehensive and robust map of global children's palliative care provision would therefore be a bespoke global survey online, adapting the approach taken in the global studies of palliative development and making use of new and more imaginative indicators, which capture the complexities of defining palliative care for children, the varied age groups and development categories to which it applies, and the specific forms of co-production that may be involved in its design and delivery.

Conclusions
We have been able to identify a total of just 21 countries in the highest categories of development for children's palliative care (7 in 4b and 14 in 4a); these countries contain 232 million children and young people age 19  Accordingly, less than 10% of those under 20 years old live in countries where palliative care for children is of the highest current standards. Almost a third of children live in countries where such care is highly limited in provision, and incommensurate with need. One quarter of all children live in countries that are only beginning to mobilise efforts for children's palliative care. The rest live in countries where it is proving difficult even to assess available levels of children's palliative care.
We have previously shown that the overall development of palliative care, despite some gains over time, is incapable of meeting existing need. For the world's children who could benefit from palliative care, the situation is even more serious and requires a global intervention of massive scale for its rectification.
We contend that the in assessing the limitations of our study, the perfect cannot be the enemy of the good. We know, based on our experience of previous mapping studies we have conducted, that despite the limitations and our gradual efforts to overcome them, these studies have proved enormously valuable to the field. Indeed, they constituted the underlying evidence base for the World Health Assembly Resolution on palliative care, which was passed in 2014, and which of course supports the provision of palliative care across the entire life course. No other group has attempted to map the global development of children's palliative care in the manner we present in our paper. Until such an attempt is made with an improved methodology, ideally by an international team of specialists in the field of children's palliaitve care, we argue that the current work should be seen, not as definitive, but as the best evidence we currently have, and which moreover is highly likely to be of benefit to the field on which it is focussed. To take a phrase from the paediatrician Winnicott, it is 'good enough'. This project contains the following underlying data: -SurveyResponsesQ26_995.xlsx -NB in response to one review of Version 1 who commented on one element in the underlying data as an example of a problem with this data, we noted a survey response wrongly assigned to the USA on the basis that it came from the representative of an organisation that works in one of the poorest countries in the world, but who for some reason gave the USA as the host country for the organisation -an issue we noted in some other responses. We dealt with these case by case in the analysis, but have amended the relevant record in the underlying data. The authors are unable to make all data publicly available because of the terms of data sharing included in the consent process for this study. Furthermore, it is not possible to effectively de-identify all the participants and their organisations in the free-text responses to the survey. However, access will be granted on a case-by-case basis upon request to the corresponding author (david.clelland@glasgow.ac.uk). Access to the full underlying data will be granted upon request from a researcher for the purposes of further research providing the requesting researcher is in possession of a protocol that has been approved by an ethics committee and which satisfies the guarantees of anonymity originally given to the research participants.

Disclaimer
Neither the inclusion or names of any territories for which results are presented or the boundaries and designations used on the map imply the expression of any opinion whatsoever on the part of the authors concerning the legal status of any country, territory, city or area or of its authorities or concerning the delimitation of its frontiers or boundaries.

Open Peer Review
This was an incredibly ambitious effort and I applaud the authors for this work. With regards to the in-country experts, what qualified an individual as an "in-country expert"? They are described as individuals engaged in the development, delivery and co-ordination of palliative care at the national level and the hierarchy was well described. However, in reading through this I did not see clinicians who are actively practicing in the country of interest who may have particular insights in the actual care delivery on the ground, especially when it comes to care delivery for children.

○
The second comment is with regards to the response rate. We have the response rate from a country level (113 of 179 countries) but we do not have a response rate of the 560 experts. This is an important piece of information when thinking about survey-based research methodologies. ○

Results:
No comments as they are well presented and described.

Strengths and Limitations
The authors have done well in addressing many of the limitations associated with this particular study and highlight the fact the deficit of not having pediatric specific experts as an area of further focus. Additionally, within the matters of definition it is critical to highlight the differences in cultures and views of who would be considered "children." ○ Additionally, I would add that as mentioned in the WHO definition, quality pediatric palliative care includes support of the family and what is defined as the family unit is also unique between cultures and individual units.

Conclusions:
There is great discussion about the need for further research in understanding the gap of pediatric palliative care globally and ways to ameliorate those research gaps. However, there is less discussion about the actual gap in clinical care for children with life-threatening conditions. While further research is needed, this paper describes the deficit, and it would be helpful ways to scale up palliative care for children in resource-limited settings.

○
Finally, there is a grammatical error in the first sentence of the last paragraph "We contend that the in assessing the limitations of our study, the perfect cannot be the enemy of the good." ○ In summary this is an extremely important topic that I believe is a valuable contribution to the literature. While there have been significant advancements in palliative care globally and pediatric palliative care in high-income settings there is a significant deficit in LMICs. The authors do a solid job of presenting their data without overstating the findings and identify ways of focusing of pediatric palliative care expertise with collaborations with WHPCA and ICPCN among others.

Is the study design appropriate and is the work technically sound? Yes
Are sufficient details of methods and analysis provided to allow replication by others? Yes

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes pursue this agenda with scientific rigor.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound? Yes

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
In these regards, the project conducted by the authors, and their report, succeed. Their main conclusions, as summed up in the abstract, is that children's palliative care is not widely available (with high levels availability in less than 10% of the global child-age population), and that the greatest opportunities to improve access are likely to be found in low-and middle-income countries. They also found, in a comparison that they appropriately do not overstate, that some degree of progress appears to have been made over the past decade. Lastly, the authors spot a clear opportunity to continue to improve the methodology by which such global assessments can occur. Even with the clearly stated limitations of the project's methodology kept in mind, these are reasonable conclusions substantiated by the data and findings from their project.
I offer 3 suggestions that I think would further strengthen the report.
First, make clearer from the outset of the report that the data is from key informants in each country, and not objective comparative data. Specifically, change the title to: "Global development of children's palliative care: An international survey of in-nation expert perceptions in 2017." This accurately states the data that the study is based on; and as currently written, the manuscript clearly addresses the strengths and limitations of this data. Given this more qualitative than qualitative, more subjective than strictly objective, nature of the data, I am not surprised or overly perturbed that the classification of specific countries may be less than absolutely accurate; the data does not have sufficient precision to offer precise level classifications. The overall conclusions of the project, though, do seem justified by the data.
Second, reconsider the projection of the Figure 2 map. At the risk of sounding pedantic, the choice of map projection is worthy of though, as each projection emphasizes one attribute while inevitably distorting other attributes. The current map projection is "high latitude biased", enlarging area nearer to the poles. An equal-area projection (Equal Earth) would be a more appropriate projection, given that one of the most important points of the report is that palliative care for children is not readily available on most of the global landmass. (One might even consider 2 maps, one with equal area, the other with equal population being depicted (see for example here), if the team is able technically to create this alternative projection, as that most important point is about population, not really landmass. This projection would draw attention to the fact that the two most populous countries either do not have palliative care readily available or are "unresolved", underscoring two opportunities.) Third, the authors might be more explicit about what "other sources of evidence" they would want to use to "triangulate" the findings (as they suggest in the second bullet under the "Strengths and Limitations" heading The study addresses an important topic. It focuses on the state of palliative care for children at the global level. The aim of the study was to: "obtain a first global overview of the national development of children's palliative care based on the knowledge of in-country experts". It reports on the methodological and conceptual challenges in mapping palliative care for children.
The paper is well written, and the process of categorisation is well-explained. The study is useful in advocating for children's palliative care, and to improve indicators to measure it at the country and international levels.
The methodological and conceptual limitations outweigh the authors' abilities to map and categorise levels of children's palliative care development at the national level and globally that they undermine the utility of the study results: The study describes the limitations of self-reported data and suggests the triangulation of data sources to address this limitation. However, this could have been considered in this study given that similar work has been done previously.
On pages 8-9, the authors state: "A significant proportion of countries included in the global mapping survey have been excluded from these results on the basis that participants provided conflicting categorisations that could not be reconciled. The exclusion of these countries limits the extent to which the results presented here can be considered a comprehensive assessment of palliative care development". The statement is self-defeating and raises questions about the study results.
The authors state that lack of definition of palliative care for children may have resulted in different interpretations, and this could have resulted in under or over-reporting, making the data unreliable.
The categorisation of children's palliative care development is useful, it also illustrates some of the limitations resulting from the reliance on the country experts sampled in the survey. For example, South Africa is categorised as 3a, yet there is an existence of a national palliative care association; and palliative care is provided by multiple providers ranging from community-based, to hospice and hospital-based services. This raises questions about categorisation of palliative care development for some of the countries.
reference to some generic indicators that are not specific to children's palliative care. It is worth noting that our grading of adult palliative care for South Africa, published elsewhere ( https://www.jpsmjournal.com/action/showPdf?pii=S0885-3924%2819%2930664-5 ) gives a grade of 4a. This is a form of 'triangulation' which may indeed demonstrate the veracity of our findings -children's palliative care in the view of in-country experts in South Africa does not reach a level of development that those same expert source provided across 10 sets of indicators when assessing adult provision.
Overall, we would suggest that this reviewer's comments and the assessment that we 'partly' addressed two of the key dimensions under review and fully addressed the three other relevant criteria, seems incommensurate with an overall judgement of 'not accepted'.
pediatric palliative care. Being an expert in adult palliative care does not necessarily mean you have in-depth knowledge of palliative care services available for children. It is problematic that categorization of countries came directly from the responses of these experts. We therefore have serious concerns about the validity of the results.
Let us consider pediatric palliative care in the United States (US) as an example. As pediatric palliative care experts based in the US we wonder why it was categorized as 4a rather than 4b given the category descriptors and our knowledge of other countries categorized as such. Specifically, the US has activism for children's palliative care nationally. For example, as a result of advocacy, in 2010 with the initiation of the Affordable Care Act, policy change led to the opportunity for children to receive disease-directed therapy concurrent with hospice, a provision not available for adults 1,2 . The American Board of Pediatrics has recognized Hospice and Palliative Medicine as a formal specialty since 2008. Interdisciplinary pediatric palliative care is now a specialty service available in a majority of freestanding children's hospitals across the US. Most health professionals, local communities, and members of our society are aware of pediatric palliative care as exemplified by media coverage of these services 3,4 . The US has unrestricted access to strong pain-relieving drugs, including morphine. The American Academy of Pediatrics has published guidelines around children's palliative care 5 and multiple academic centers have integrated palliative care into pediatrics residency curricula. Finally, pediatrics is a strong and integral part of the American Academy of Hospice and Palliative Medicine. Despite this level of PPC integration in the US, one of the "in-country experts" ranked the US as a 1 "No known palliative care activity for children" (Extended Data: SurveyResponsesQ26_995). We use the United States to exemplify the potential flaws of relying on adult experts to rank pediatric palliative care development across the globe. While there remains room for further pediatric palliative care integration in the US, this is also the case for the other countries categorized as 4b. And there is substantial misclassification of other countries. For example, Russia was ranked a 4a based on a single survey respondent. Our knowledge of the country is that it should be categorized as a 3a, isolated children's palliative care provision.
We do commend the authors for a paper that brings attention to the important issue of mapping children's palliative care across the globe. While we raise serious issues around the methodology and in turn the results, we appreciate the discussion around the complexity and nuances of palliative care in children and the inherent challenges around measuring it globally. As mentioned by the authors, we agree that in order to more accurately capture the current state of pediatric palliative care, there needs to be a global survey using measures and experts specific to children.
The reviewers also question the grading of their own country, the USA, and make reference to the underlying data as an example of the problem they perceive, specifically SurveyResponsesQ26_995. In fact, this response came from the representative of an organisation that works in one of the poorest countries in the world, but who for some reason gave the USA as the host country for the organisation -an issue we noted in some other responses, which we dealt with case by case. This particular questionnaire was therefore quickly excluded from our analysis of the USA data. Moreover, the reviewers may be reassured to know (and we regret that confidentiality principles don't allow us to give more details) that the other two responses from the USA, not only both gave the same grading (4a), but also came from impeccable organisational sources and were signed off in one case by a senior clinician of global standing in the field of palliative care.
In sum, we reiterate our thanks to the reviewers, who clearly concur with us on so many issues. We contend however that the reviewers make 'the perfect the enemy of the good'. We know, based on our experience of previous mapping studies we have conducted, that despite the limitations and our gradual efforts to overcome them, these studies have proved enormously valuable to the field. Indeed, they constituted the underlying evidence base for the World Health Assembly Resolution on palliative care, which was passed in 2014, and which of course supports the provision of palliative care across the entire life course.
No other group has attempted to map the global development of children's palliative care in the manner we present in our paper. Until such an attempt is made with an improved methodology, we argue that the current work should be seen, not as definitive, but as the best evidence we currently have, and which moreover is highly likely to be of benefit to the field on which it is focussed. To take a phrase from the paediatrician Winnicott, it is 'good enough'.

Competing Interests:
No competing interests were disclosed.

Comments on this article Version 1
Reader Comment 25 Jun 2020 Joan Marston, Palliative Treatment for Children South Africa, Cape Town, South Africa Thank you for this paper and for highlighting the need for unique research to map the development of palliative care for children; and for identifying the low level of provision worldwide that requires urgent attention.. This sort of mapping is always difficult for many of the reasons that the authors have identified when listing the limitations. From previous experience in research to identify the need for children's palliative care, the experience we had was there will always be conflicting views. While I do not agree with all the country classifications , including in South Africa and certain other countries in which I have worked quite extensively ,this research can be used to advocate for both further research and development. It s a pity that conflicting views led to the exclusion of certain countries like Belarus where palliative care for children is well-established, extensively available and of a high level. Unfortunately experience has shown that not all country leaders in palliative care have knowledge of development for children. This paper also , for me, highlights the need for those working in the field of palliative care for children to develop a stronger voice. I look forward to the further research this paper should generate.
Competing Interests: I have no competing interests