Data resource profile: children looked after administrative records in Wales

Abstract

Introduction
Children who are 'looked after' by the State are considered one of the most vulnerable groups in society. Being in State care is associated with poor social, educational and health outcomes. Exploring how to improve the system and better support children in care is key to improving these outcomes. When children and young people come to the attention of children's social services a significant amount of information about their care experience is routinely collected by local authorities. In Wales, routine data are captured in the 'Children Looked After' Census which is submitted annually to the Welsh Government and has recently been shared with the Secure Anonymised Information Linkage (SAIL) Databank.

Method
The aim of this paper is to provide an overview of the main 'Children Looked After' Census dataset and its subsets. These datasets contain rich, situational and individual level data on children looked after, such as information on placement types, education and leaving care. We outline the strengths and limitations of the available information and how to access the data.

Results
The 'Children Looked After' Census has recently been made available for research purposes and access to it will enable researchers to explore and understand at population level the journey through the care system and outcomes of leaving care. There is also the opportunity, through the SAIL Databank, for data linkage to health, education and family justice datasets, allowing research to holistically explore other factors associated with being in care.

Conclusion
These data provide a rich source of information about children and young people who have been in care in Wales. They offer researchers opportunities to better understand the care system and outcomes for this within it. Findings will have important implications for making improvements in children's social care policy and practice.

Introduction/rationale

Children who are looked after by the State are considered one of the most vulnerable groups in society [1–4] and will have suffered significant harm, maltreatment and trauma consequently affecting their physical and emotional wellbeing. Information about children in care is routinely collected by local authorities and submitted to the Welsh Government annually in the form of the Children Looked After Census, which was previously referred to as the SSDA903 Return. In Wales, children in State care (referred to in this paper as ‘children looked after’) are ‘looked after’ by a local authority under the Children Act (1989) or the Social Services and Well-being (Wales) Act (2014)¹. Under the latter Act, specifically Section 76, children and young people can become ‘looked after’ through a voluntary arrangement referred to as ‘voluntary accommodation’. This arrangement is made with parental consent and is not mandated by the courts. Parents retain legal rights and the right to withdraw consent for accommodation at any time. Section 77 of the 2014 Act also requires local authorities to accommodate children who are removed or kept away from home under Part 5 of the Children Act 1989, i.e. those children who are in police protection, detention, or on remand.

The alternative route into the care system is through a court order. In circumstances where there is serious risk of harm to a child or young person local authorities can apply to the court for an order that gives a local authority parental responsibility of the child. In Wales, like England, court orders are applied for and granted under Section 31 of the Children Act (1989). Courts have the powers to issue emergency protection orders, interim care orders or a full care order. A full care order can be mandated by the court where local authorities share parental responsibility with birth parents but have the power to decide if contact with parents is in the best interest of the child and, if it is, how much contact. Children and young people on a care order may be placed with unrelated foster carers and some may be with kinship foster carers or birth parents. Where they are placed with birth parents, this will typically be as part of a process of reunification for children who have previously been removed from home.

Being looked after by the state is associated with short and long term adversity and is a predictor of poor life outcomes [5]. It is well evidenced that children in state care do less well than their non-looked-after peers in several domains related to health, education and other social outcomes [3, 6–8], although much of this disadvantage stems from their pre-care experience rather than from being looked after by local authorities [9]. Therefore understanding determinants and characteristics of outcomes of looked-after children, as well as the journey through the family justice and care system, is important if improvements are to be made and children and young people are provided with the best possible care and opportunities.

This paper follows Mc-Grath-Lone et al’s [10] data resource profile which provides a detailed overview of the English Children Looked After return, including the purpose and scope of the datasets. Their paper outlines the dataset structure, the information collected and details how the dataset has been used to monitor outcomes. Mc-Grath-Lone et al. [10] summarise the content of the dataset under three themes, child characteristics, episode and period details and outcomes of care. They discuss the quality of the data collected and describe some of the weaknesses of the dataset, namely the missing years of full data collection between 1998-2003. Furthermore they acknowledge the dataset does not contain baseline characteristics of children and families or provide information about interventions and ongoing parental contact.

Building on this valuable publication, this paper will describe the Children Looked After Census in Wales, given differences in the structure and content resulting from legal differences and introduction of the Social Services and Wellbeing (Wales) Act 2014 in 2016.

While there are some similarities between the Welsh and English Children Looked After datasets, particularly pre-2016 when both countries operated under the same legal framework, there are also a number of key differences in the scope of the data recorded and the availability of certain data items for research purposes. It is not the intention of this paper to present a detailed comparison between England and Wales Children Looked After data collection and pre- and post-2016 changes; instead we outline how the Welsh datasets are collected and what information they contain currently. This article details the current set of individual-level Children Looked After datasets available for Wales, and held within the Secure Anonymised Information Linkage (SAIL) Databank [11, 12].

Purpose of children looked after datasets

A significant amount of children’s social care individual data is routinely collected in Wales about children and young people in care. This information can provide fruitful insights about the journey through the care system and beyond. The information can be used to influence policy and practice in children’s social care and used to target interventions and support appropriately.

Routinely collected administrative data is used by local authorities and collated by Welsh Government to evaluate the outcome of policy initiatives and to monitor objectives on children looked after, both during their time in care and on reaching adulthood.

SAIL databank

Administrative data collected and maintained by Welsh Government were acquired by the privacy-protecting SAIL Databank [11–13]. The SAIL Databank contains a wealth of anonymised health and administrative data about the population of Wales, accessible via a secure data-sharing platform, all underpinned by an innovative and proportionate Information Governance model.

Each dataset within the SAIL Databank replaces individual identities with a unique field for each person to enable linkage of their records across datasets. SAIL anonymisation and linkage methodology are described elsewhere [12]. All data within the SAIL Gateway are treated in accordance with the Data Protection Act 2018 and are compliant with the General Data Protection Regulation.

The SAIL databank also allows for linkage to other datasets to provide a holistic picture of the experiences of children and young people in care and their onward trajectories. The Welsh Government have shared individual-level Children Looked After data from 2001 onwards with SAIL, including a handful of superseded data collections. We highlight these no longer collected datasets and illustrate their relationship to the equivalent replacements.

All proposals to use these datasets are subject to review and approval by the SAIL Information Governance Review Panel (IGRP). When access has been granted to anonymised data, it is gained through a privacy-protecting safe-haven and remote access system, referred to as the SAIL Gateway. Anyone wishing to access data should follow the application process guidelines available at: www.saildatabank.com/application-process.

A Welsh Social Care Data Users Group has also been established (led by researchers at Swansea, Cardiff and Lancaster Universities) to support use of Welsh administrative data. This group is affiliated with the Children’s Social Care Data User Group (CSCDUG) based in England [14].

Children looked after datasets

In Wales, children’s social care services are delivered at the local government level, via local authorities. Local Authorities are obligated to deliver to the Welsh Government yearly individual-level information about those who come to the attention of children’s social services. The Welsh Government collect several datasets relating to children receiving services from social care, and the majority of these datasets focus on children who are looked after. These datasets contain rich situational and individual-level data items. The information collected in the Children Looked After (formerly Looked After Children) Census allows the trajectories of individual children through the care system to be objectively described in a good level of detail. Extensions to the main Children Looked After dataset, which primarily records the placements that a child has throughout the year, also provide snapshots of information about the educational achievements of individual children and about their experiences upon leaving care if they are adopted or if they age out of the social care system.

Children looked after census (formerly named SSDA 903)

The Children Looked After census, previously the SSDA 903 collection, is the primary dataset relating to children who are looked after in Wales. Its purpose is to provide a complete record of all individual episodes of care that a child has experienced throughout a year and record any changes of legal status and any specific events or milestones in a child’s care history. Data documentation related to this dataset, its content, and variable codes are available from the Welsh Government [15]. The team in Welsh Government who are responsible for the dataset carry out intensive data validation and verification checks each year, meaning that this is a very high-quality dataset with only a tiny number of data errors.

The Children Looked After data collection includes a broader group of children than just those looked after under s.76 and s.77 of the 2014 Act and Part IV of the 1989 Act. It also includes children who receive short breaks of a continuous period of more than 24 hours as part of a Part 6 care and support plan, children placed for adoption, children who are placed with prospective adoptive parents, and children aged 16 or 17 and in transitional settings in preparation for leaving care, among others. Prescriptive guidance sets out the groups of children who are and are not eligible to be included in the census, and an exhaustive list is available in Appendix 1.

Dataset structure

The Children Looked After dataset available inside the SAIL Databank covers 1^st April 2001 – 31^st March 2021 and will continue to be updated each year, offering a minimum of two decades of individual-level data about children who have been looked after at any point during this period. The data available inside SAIL consists of approximately 310550 recorded episodes, reflecting data for 37775 unique children.

This data collection has an ‘episodic’ structure, with the period during which a child is looked after by the local authority being broken down into ‘episodes’ of care. A new episode begins every time a child starts to be looked after; there is a change in the child’s legal status; or the child’s placement changes. Episodes are recorded in consecutive date order.

Episode completion information is only populated if an episode ends within a collection year (1st April – 31st March). If the episode is still open on 31^st March in a census year, then the end date and reason for the end of the episode are left blank, and the following year’s record starts with the same episode details. In cases where a single open episode continues unchanged from year-to-year a record is still submitted every year for that child.

Children’s characteristics

The Children Looked After Census contains information about child gender, date of birth, Unique Pupil Number (UPN), ethnicity, disability status, and asylum-seeking status. The UPN is only recorded for school-aged children, predominantly those over four years old, though some children (e.g. those eligible for the Flying Start programme) may have a UPN when they are as young as two years old [16, 17]. The change in the legal framework from the 1989 Act to the 2014 Act has led to the modification of the information collected in some fields, with ethnicity recorded post-2014 Act being placed in much broader categories than earlier collections. There has also been a change in the recording of child sex and gender;pre-2016, the census recorded child sex (boy/girl), whereas more recent collections record child gender identity at the census date (male/female). Information about the local authority responsible for the child is also included. Postcode data should also be recorded for the household where the child was living when they first became looked after.

Within the SAIL Databank, for anonymisation reasons, the date of birth is transformed to the week of birth, and the child home postcode becomes the Lower Super Output Area (LSOA) of the postcode. Each LSOA comprises households within postal codes aggregated to reach a minimum number of people that satisfy statistical disclosure requirements, generally with an average population of 1500 people or 650 households. Researchers can use LSOAs to answer research questions about deprivation and the likelihood of children coming to the attention of children’s social services [18, 19]. The SAIL Databank can assign an Anonymous Linking Field (ALF) [11, 12] to many children recorded in the Children Looked After Census through the linkage process via a trusted third-party. ALFs are not assigned where the linkage process has been unable to find a matching NHS health record using the individual identifiers included in the Children Looked After Census – date of birth and gender (and name and address for children with a recorded UPN). The ALF can be used to link individuals across datasets from a wide range of health, education, and family justice sources. The Welsh Government collection also includes the UPN, and they share an anonymised version of this – the Individual Record Number (IRN) – with the SAIL. This IRN allows researchers to link children directly to their entries in the Pupil Level Annual School Census (PLASC) and facilitate linkage to education datasets for those children that the ALF process did not succeed for.

Episodes of care

Within the Children Looked After Census, an episode of care consists of the combination of placement and legal status, and a change in either, or both, triggers a new episode. A new episode is also recorded at the start of each census year (i.e. 1^st April in a given census year) even if there was no change in placement or legal status. This is essentially a duplication of the episode record open on 31^st March in the same year.

The commencement of an episode has both the date and reason. Where episodes are consecutive, the new episode will start on the same day as the previous episode finished. For children accommodated under a series of short-term placements, the start date of an episode is the date of the beginning of the first placement under the agreement. Episodes must be longer than one day (i.e. cannot start and end on the same day), and if a child has two changes of legal status in one day only the second legal status is recorded.

Each episode also records the legal status underlying a child being ‘looked after’. The legal status describes, in legal terms, why the child is being looked after. A child may have other legal statuses arising from the youth justice system, for example, but these have no bearing on the ‘looked after’ status of a child and are not recorded in the Children Looked After census. It is important to note that some youth justice legal statuses do not make a child ‘looked after’ even if they require that the child is accommodated in a secure unit.

Similarly to the Welsh Children In Need (CIN) and Children in Receipt of Care and Support (CRCS) censuses [17], the Children Looked After collection also records the main reason a child receives care and support services. Children captured in the Children Looked After census will generally be included in the CIN or CRCS censuses, provided they meet the date criteria for inclusion in these <cite our paper>, and therefore will also have their need for care and support recorded there. It is important to note that the need for care and support might differ between the CIN/CRCS and Children Looked After datasets – in the Children Looked After census, the ‘need code’ is the most relevant at the time the period of being looked after began. In the CIN/CRCS censuses, however, the ‘need code’ indicates the main reason why a child started to receive care and support services. The implication is that the CIN/CRCS ‘need code’ would likely pre-date a child’s entry to care and would be more reflective of the situation when the child was still residing with birth parents.

For children who receive a series of short-term breaks, the number of breaks received by the child is recorded in the dataset.

The placement code indicates where, in an episode, a child is living. Any change in placement code requires the beginning of a new episode, but it does not necessarily mean that a child has moved locations (though this is the most common type of placement change). For example, a child might be placed for adoption with previous foster parents. Any researcher using placement code to derive information about placement stability should account for these sorts of changes. This field also records unauthorised absences of more than 24 hours.

The data received by Welsh Government also include details about the postcode of each child placement, except where the placement is an adoption placement. As with the child home postcode, placement postcode is aggregated into the LSOA of the placement postcode when the data is shared with the SAIL Databank. The quality of this data item is variable, particularly for older episode records, meaning in some cases there is no placement LSOA derived or that the placement LSOA is incorrect. Incorrectly recorded placement LSOAs, and by extension incorrectly recorded placement postcodes, are particularly evident where children are recorded as placed outside the Local Authority (LA) boundary, but their reported LSOA is inside the LA boundary. Where this field is populated it can be helpful to link to the Welsh Index of Multiple Deprivation (WIMD) for studies that may wish to take this into account. However, caution is advised to ensure that the recorded LSOA is legitimate in the context of the placement type.

There are two further details recorded at the end of an episode – the date the episode ended and the reason for the episode’s end. In many cases, one episode will end and another will start on the same day, and if an episode does not end during a year this field is left blank. When a child ceases to be looked after because of adoption this date is the date the court grants the adoption order. When a child ceases to be looked after because they have died, the episode end date is the same as the date recorded on the death certificate.

Adoptions of children looked after

This separate dataset reports information about all children who cease to be looked after in a collection year upon the granting of an adoption order. Children are placed with adoptive parents before adoption but are still considered ‘looked after’ for the main Children Looked After data collection – it is only on the granting of the adoption order that they cease to be looked after. All dates in this collection refer to the final decision date of the local authority and not the date of the adoption panel recommendations.

The dataset in SAIL contains information about approximately 5240 individual children and 5255 individual adoptions. It includes details of all adoptions of children looked after in Wales from 1^st April 2002 – 31^st March 2021 and is continuously updated each year. Due to the nature of the information contained in this dataset it is considered sensitive. All applications for it must be proportional and will be carefully considered by the data owner, Welsh Government.

Dataset structure

Within SAIL, this dataset must link to the primary Children Looked After Census to get child characteristics (week of birth, gender, ALF) and so, logically, this captures a subset of the children recorded in the full Children Looked After dataset.

All dates recorded in the Adoption collection refer directly to the various steps in the adoption process. They include the date that the local authority decided adoption was in the child’s best interests, the date that the child was matched with prospective adoptive parents, the date of placement for adoption, and the date that the child was adopted and ceased to be looked after.

An amount of information is also recorded about the characteristics of the adopter(s). These characteristics include adopter gender, the number of adopters, their legal status (i.e. marriage status), and ethnicity. There is also a flag to indicate if the child has been adopted by their former foster carers.

Care leavers aged 16 and over

The introduction of the Social Services and Well-being (Wales) Act 2014 did not simply cause the renaming of the main Children Looked After collection. Due to the change in underlying legal frameworks, it also necessitated the end of a dataset that recorded information about care leavers on their 19^th birthday (detailed below) and the beginning of a dataset recording information about care leavers on their 16^th birthday. This change came about because of the implementation of the Care Leavers (Wales) Regulations 2015. These regulations aimed to strengthen the support given to young people leaving care.

Within SAIL, this dataset currently covers 1^st April 2016 – 31^st March 2021 and includes all children who ceased to be looked after during a collection year if they were aged 16 years or older at the time of care ending. Only the last occasion that care ceased is included. There are 3315 records reflecting 3240 individual children.

The discrepancy in the number of children versus the number of records in this dataset is explained by the fact that some children in the ‘care leavers 16 and over’ dataset leave care in one year (for example, the 2019/20 financial year) and then return to care in the next financial year (e.g., 2020/21) after a brief period of not being looked after. However, this is very uncommon effecting approximately 80 children over the entire dataset.

Dataset structure

Within SAIL, as with the adoptions dataset, this must link to the main Children Looked After dataset to get child characteristics (week of birth, gender, ALF) and the children captured in this are a limited subset of those in the main dataset.

The link to the primary Children Looked After dataset makes several fields recorded in this collection - the latest date started to be looked after, date ceased to be looked after, and reason care ceased - somewhat extraneous as they can be retrieved from the episode information in the main Children Looked After dataset. However, they are helpful if the user of the dataset is primarily interested in the tail-end of a child’s care experience.

Several data items offer a fuller picture of the support offered to a child as they transition out of care and into independent living. These include information about the assignation of a personal adviser and if a care leaver had been assigned a Pathway Plan. These fields only apply to children defined as category 2, 3, or 4 care leavers². The type of accommodation that the child is living in on the census date is also recorded, as is the suitability of the accommodation. There are no data items detailing the education, employment, or training status of the child; likely due to the change in local authority legal responsibilities brought about by the introduction of the 2014 Act.

OC3: care leavers at 19th birthday

Prior to the introduction of the 2014 Act, the OC3, Care leavers on their 19^th birthday, collection was used to record information about the education, employment, and accommodation of former care leavers on their 19^th birthday. This collection took place while Wales operated under the legislation introduced by the 1989 Act, which is the legal framework still in place in England, and so the Welsh OC3 collection is the same as the English OC3 collection.

As with the other subset Children Looked After datasets available in SAIL, this must be linked to the primary dataset to retrieve child characteristics. This dataset includes care leavers who turned 19 years old between 1^st April 2002 – 31^st March 2016. It is no longer updated. There are 4940 records for 4890 children.

Dataset structure

The primary and most well-populated field in the OC3 collection represents whether a young person was in touch with the local authority responsible for them when they left care. Contact with the young person must have taken place within the three months prior to or one month after their 19^th birthday for them to be considered ‘in touch’. This applies regardless of whether the young person had a pathway plan on leaving care. Of the 4890 young adults, 4460 are recorded as ‘in touch’.

There are a further two data items for those young people who are in touch. The main activity of the young person describes their involvement in employment, education, or training, or reasons for why they are taking part in none of these activities. The accommodation of the young person is a two-part code where the first digit is alphabetical and represents the type of accommodation, and the second digit is numerical and indicates whether the social worker judged the accommodation suitable or unsuitable. Accommodation is suitable if it provides safe, secure and affordable provision for young people. Accommodation that exposes the person to the risk of harm or social exclusion because of its location or other factors is coded as ‘unsuitable’. Of the 4460 ‘in touch’ young adults, approximately 4110 are in suitable accommodation.

While there is no specific accommodation-level information collected (e.g. postcode, LSOA), using the SAIL databank, it is possible to use the ALF of the young person to compare to the LSOA of the registered address on GP records at the point at which the OC3 collection was completed. However, it is unlikely that young people in unstable or unsuitable accommodation will be registered with a GP or that their registered address will be accurate, so consideration will need to be applied in this case.

OC1: educational qualifications of care leavers

This dataset existed under the 1989 Act, and, as with the OC3 collection, it was ended with the introduction of the 2014 Act. The Wales OC1 collection is the same as the English OC1 collection due to the legal frameworks previously being identical. However, there were likely differences in practice and collection, making it inappropriate to consider them equal.

Within SAIL, this dataset currently covers 1^st April 2002 – 31^st March 2016 and includes all children who ceased to be looked after during a collection year if they were aged 16 years or older at the time of care ending. There are approximately 7150 records reflecting 7080 individual children. It must be linked to the primary Children Looked After collection to retrieve child characteristic information.

Due to its inclusion within SAIL Databank, the ending of this dataset represents a minor issue. By linking to the Pupil Level Annual School Census (PLASC) dataset, using the ALF or the IRN from the primary Children Looked After collection, we can anyway retrieve information about the number of GCSEs attained by an individual and the grades they achieved in each subject.

Dataset structure

The primary purpose of the OC1 collection is to record the qualifications that young people had gained at the point they ceased to be looked after. The main fields of interest set out the number of GCSEs attained by the young person, broken down into A*-G and A*-C ranges, the number GNVQs and the number of other educational or vocational qualifications. For those young people who attain no qualifications a reason is given, accommodating situations where a young person sat at least one exam but did not attain a qualification, where a young person sat no exams due to illness or disability, and where a young person sat no exams not due to an illness or disability.

Data quality-strengths and limitations (in all datasets)

The main strength of the Welsh Children Looked After datasets is that they have whole population-level coverage. Records held within the SAIL Databank are also fully anonymised. The use of administrative data for research is highly desirable. The data are not subject to reporting bias, such as stigma, that can apply to self-reporting certain kinds of contact with social services [20]. The data collected are longitudinal, allowing for change over time to be assessed. Finally, administrative data remove the burden on research participants to disclose sensitive issues [21].

One significant limitation of the Children Looked After dataset is the lack of routinely collected information on the socio-economic circumstances of children and families requiring state intervention. Such information would provide a more holistic picture of the context and circumstances in which children require care and their backgrounds [10]. However, through the SAIL Databank and linkage to other data sources there is opportunities to explore a diverse range of issues that are pertinent to understanding the challenges facing children and families. Data in other datasets within the SAIL Databank include, for example, free school meal receipt as an indicator of low income for school-age children. A further limitation is that the data do not contain information about children post-adoption. When an adoption order is granted child’s identifiers are changed – this includes their UPN, unique ID within a local authority, and their NHS number. The impact of this change on the SAIL ALF and the ability to link pre- and post-adoption records is currently unknown; but it is likely that it will not be possible to follow up on incidence of a return to care, or the educational outcomes of this group.

It is very encouraging that the Welsh Children Looked After dataset is mostly complete with very little missing data. However, as with most administrative data sets there is some missing information, or the information recorded may not be particularly useful. In the Welsh dataset there is some missing data on home and placement postcodes effecting a number of children, primarily those in older collections. There is also some variation in recording between local authorities; and LA child identifiers do not continue with the child if the responsibility for their care is transferred to another local authority. This can make it challenging to link records for individual children who have this experience. Researchers can mitigate this by using the IRN and the SAIL ALF to identify children who have multiple local authority identifiers and reunify their care records. It is important to note that data collection ceased in 2016 for two of the datasets; Care Leavers at 19^th Birthday and Educational Qualifications of Care Leavers.

The inclusion of the Welsh datasets in the SAIL Databank does afford it a number of strengths over the English dataset. Through the allocation of an ALF we are able to link to a large number of other health and education datasets that could not be linked with the English dataset. This allows researchers to identify siblings of children looked after, to examine education records even when a child is only looked after prior to school-age, and to explore some characteristics of birth families and residential/foster carers. Ongoing work in England with the The Education and Child Health Insights from Linked Data (ECHILD) [22] [citation here] database does allow the linkage of education, social care, and some English health datasets, but does not yet include the range of datasets that would allow research into these key questions. That said, the ECHILD database represents an invaluable resource for English research, and we hope that research into these topics will become possible as the database evolves and grows.

Implications, usage, data linkage opportunities

The Welsh Government release a statistical return annually based on the Children Looked After census data in aggregate form. Statistical tables, annual reports and additional documentation relating to the Children Looked After data are publicly available from the ‘Stats Wales’ website [23]. Welsh Government provide data documentation that describes the content, and lookup’s for variable codes which is helpful for researchers new to the data [24]. There are also opportunities to link the Welsh Children Looked After data anonymously to other sources of information, such as health and education data through the SAIL databank. Existing research using administrative records for children looked after in Wales are limited although a growing field of enquiry [19, 25–27]. However, there is huge potential to explore a range of issues relating to children’s social care through data linkage. Many agencies that work alongside children social care services generate administrative data that can be used to understand who comes into the care system, what happens to them before, during and after leaving care. To get a complete picture of such trajectories over time, it is crucial to link datasets [28]. By doing so researchers can quantify specific risk factors that are associated with entry into and the long term consequences of being in care [28]. Using evidence derived from population data can be used to target where in the system improvements can be made both on a policy and practice level.

Ethical statement

An application for access to the Welsh social care datasets held in the SAIL Databank, to explore the nature of the data and to generate the simple descriptives for this resource, was reviewed by an independent Information Governance Review Panel (IGRP). The IGRP considers each project to ensure the proper and appropriate use of SAIL data. This work was approved, and access was granted through a privacy-protecting safe haven and remote access system.

Conflict of interest

One of the authors on this paper works in collaboration with Welsh Government to provision the Children Looked After data in the SAIL databank and to make it accessible for researchers.

Funding and acknowledgements

This work was supported by Health and Care Research Wales, as art of the CASCADE partnership infrastructure funding (517199), and Administrative Data Research (ADR) Wales (ES/W012227/1). LJG and KB are also funded by the Nuffield Family Justice Observatory (FJO/43766).

This study uses anonymised data held in the Secure Anonymised Information Linkage (SAIL) Databank, which is part of the national population data research infrastructure for Wales. We would like to acknowledge all the data providers who make anonymised data available for research. ADR Wales, part of the ADR UK investment, unites research expertise from Swansea University Medical School and WISERD (Wales Institute of Social and Economic Research and Data) at Cardiff University with analysts from Welsh Government. ADR UK is funded by the Economic and Social Research Council (ESRC), part of UK Research and Innovation.

We would like to thank colleagues from Welsh Government, who have had sight of this paper before its publication and endorse the information presented as accurate.

References

1. Brand SL, Morgan F, Stabler L, Weightman AL, Willis S, Searchfield L, et al. Mapping the evidence about what works to safely reduce the entry of children and young people into statutory care: A systematic scoping review protocol. BMJ Open. 2019;9(8):1–7. 10.1136/bmjopen-2018-026967

   https://doi.org/10.1136/bmjopen-2018-026967

2. Harrison N. Moving on Up: Pathways of care leavers and care-experienced students into and through higher education. 2017;(November):1–84. Available from: http://www.nnecl.org/resources/moving-on-up-report?topic=guides-and-toolkits.

3. Clark D, King A, Sharpe K, Connelly G, Elliott L, Macpherson LMD, et al. Linking routinely collected social work, education and health data to enable monitoring of the health and health care of school-aged children in state care (‘looked after children’) in Scotland: a national demonstration project. Public Health [Internet]. 2017 Sep;150:101. Available from: https://search.proquest.com/scholarly-journals/linking-routinely-collected-social-work-education/docview/2059719261/se-2?accountid=14680.

4. Sebba J, Berridge D, Luke N, Fletcher J, Bell K, Strand S, et al. The educational progress of looked after children in England: linking care and educational data [Internet]. Oxford: University of Oxford. Rees Centre for Research in Fostering and Education; 2015. 35 p. Available from: http://reescentre.education.ox.ac.uk/wordpress/wp-content/uploads/2015/11/EducationalProgressLookedAfterChildrenOverviewReport_Nov2015.pdf.

5. Viner R, Taylor B. Adult health and social outcomes of children who have been in public care: population based study. Paediatrics. 2005;115:894–9.

6. Martin A, Ford T, Goodman R, Meltzer H, Logan S. Physical illness in looked-after children: A cross-sectional study. Arch Dis Child. 2014;99(2):103–7. 10.1136/archdischild-2013-303993

   https://doi.org/10.1136/archdischild-2013-303993

7. Griffiths LJ, Johnson RD, Broadhurst K, Bedston S, Cusworth L, Alrouh B, et al. Maternal health, pregnancy and birth outcomes for women involved in care proceedings in Wales: a linked data study. BMC Pregnancy Childbirth. 2020;20(1):1–13. 10.1186/s12884-020-03370-4

   https://doi.org/10.1186/s12884-020-03370-4

8. Berridge D, Luke N, Sebba J, Strand S, Cartwright M, Staples E, et al. Children in Need and Children in Care: Educational Attainment and Progress. 2020;(April). Available from: www.nuffieldfoundation.org

9. Forrester D, Goodman K, Cocker C, Binnie C, Jensch G. What is the impact of public care on children’s welfare? A review of research findings from England and wales and their policy implications. Vol. 38, Journal of Social Policy. 2009. p. 439–56. 10.1017/S0047279409003110

   https://doi.org/10.1017/S0047279409003110

10. Mc Grath-Lone L, Harron K, Dearden L, Nasim B, Gilbert R. Data Resource Profile: Children Looked After Return (CLA). Int J Epidemiol. 2016;45(3):716F–717F. 10.1093/ije/dyw117

    https://doi.org/10.1093/ije/dyw117

11. Ford DV, Jones KH, Verplancke JP, Lyons RA, John G, Brown G, et al. The SAIL Databank: Building a national architecture for e-health research and evaluation. BMC Health Serv Res. 2009;9:1–13. 10.1186/1472-6963-9-157

    https://doi.org/10.1186/1472-6963-9-157

12. Lyons RA, Jones KH, John G, Brooks CJ, Verplancke JP, Ford DV. et al. The SAIL databank: Linking multiple health and social care datasets. BMC Med Inform Decis Mak. 2009;9(1):1–8. 10.1186/1472-6947-9-3

    https://doi.org/10.1186/1472-6947-9-3

13. Jones KH, Ford D, Ellwood-Thompson S, Lyons R. The UK Secure eResearch Platform for public health research: a case study. Lancet. 2016;388.

14. CSCDUG. Children’s Social Care Data User Group [Internet]. 2020. Available from: https://cscdug.co.uk/

15. Welsh Government. Children looked after census 2020-2021 [Internet]. 2021 [cited 2022 Mar 17]. Available from: https://gov.wales/sites/default/files/statistics-and-research/2021-07/looked-after-children-census-2020-21-guidance.pdf

16. Welsh Government. Flying Start [Internet]. 2021 [cited 2022 Mar 17]. Available from: https://gov.wales/sites/default/files/publications/2019-07/flying-start-health-programme-guidance_0.pdf

17. Lee A, Elliott M, Scourfield J, Bedston S, Broadhurst K, Ford D, et al. Data Resource: Children Recieving Care and Support and Children in Need, adminstrative records in Wales. Int J Popul Data Sci. 2022;

18. Cusworth L, Bedston S, Alrouh B, Broadhurst K, Johnson R, Akbari A, et al. Uncovering private family law: Who’s coming to court in England? 2021.

19. Elliott M. Child Welfare Inequalities in a Time of Rising Numbers of Children Entering Out-of-Home Care. Br J Soc Work. 2020;50(2):581–97. 10.1093/bjsw/bcz154

    https://doi.org/10.1093/bjsw/bcz154

20. Zhang ML, Boyd A, Cheung SY, Sharland E, Scourfield J. Social work contact in a UK cohort study: Under-reporting, predictors of contact and the emotional and behavioural problems of children. Child Youth Serv Rev [Internet]. 2020;115(October 2019):105071. Available from: 10.1016/j.childyouth.2020.105071

    https://doi.org/10.1016/j.childyouth.2020.105071

21. Bronnell MD, Jutte DP. Administrative data linkage as a tool for child maltreatment research. Child Abus Negl [Internet]. 2013;37(2–3):120–4. Available from: 10.1016/j.chiabu.2012.09.013

    https://doi.org/10.1016/j.chiabu.2012.09.013

22. Mc Grath-Lone L, Libuy N, Harron K, Jay MA, Wijlaars L, Etoori D, et al. Data Resource Profile: The Education and Child Health Insights from Linked Data (ECHILD) Database. Int J Epidemiol. 2021;1–7. 10.1093/ije/dyab149

    https://doi.org/10.1093/ije/dyab149

23. Welsh Government. Stats Wales [Internet]. Statistical Return. 2022. Available from: https://statswales.gov.wales/Catalogue/Health-and-Social-Care/Social-Services/Childrens-Services/Children-Looked-After

24. Welsh Government. Children receiving care and support census 2018-19 [Internet]. Statistics for Wales. 2019. Available from: https://gov.wales/sites/default/files/statistics-and-research/2019-05/children-receiving-care-and-support-census-2018-19-notes.pdf

25. Allnatt G, Elliot M, Scourfield J, Lee A, Griffiths L. Use of linked administrative children’s social care data for research: a scoping review of existing UK studies. Br J Soc Work. 2022;

26. Bywaters P, Scourfield J, Jones C, Sparks T, Elliott M, Hooper J, et al. Child welfare inequalities in the four nations of the UK. J Soc Work. 2020;20(2):193–215. 10.1177/1468017318793479

    https://doi.org/10.1177/1468017318793479

27. Hooper J, Cusworth L, Whincup H. Permanently Progressing? Building secure futures for children in Scotland. 2019.

28. Jay MA, Woodman J, Broadhurst K, Gilbert R. Who cares for children? Population data for family justice research. 2017;(October). Available from: http://wp.lancs.ac.uk/observatory-scoping-study/

29. Social Services and Well-being (Wales) Act 2014 [Internet]. 2014 [cited 2022 Mar 17]. Available from: https://www.legislation.gov.uk/anaw/2014/4/section/104