Taking a count of vulnerable children
The effective collection of data on the number and location of
vulnerable children is a highly pertinent issue as government
guidance has called for the setting up of local strategic
partnerships to address a wide definition of children’s needs.
Researchers Philippa Hare, Martin Baxter and Elizabeth Newbronner
report on one children’s planning group’s attempt to get its sums
A principle of the Children Act 1989 is local authorities’ duty
to promote the welfare of “children in need” and their families.
Vulnerable children are defined in draft guidance issued in 2000,
opening the way for a multi-agency response to a much wider
category of children.1
The guidance defined vulnerable children as those “who have
acquired or encountered some difficulty which requires additional
help if their life chances are to be optimised or the risk of
social exclusion is to be averted”.
The guidance proposes co-ordinated planning for vulnerable
children, using local strategic partnerships to “jointly address
levels of need, determine priorities, articulate shared objectives
and agree targets and milestones”.
Accurate and reliable information is fundamental to identifying
the needs of local children and the nature of services required to
meet those needs. A mapping exercise recently carried out for a
children’s services planning group in northern England has
highlighted some of the challenges involved in sharing information
about children between agencies, which is a pre-requisite to joint
The project aimed to estimate the number of vulnerable children
in the area and determine their geographical distribution. It also
sought to describe the services available to support them and their
families, and to identify how many were receiving services from
more than one agency.
The project team began by identifying the key services, both
statutory and voluntary, which provide support services for
“vulnerable children” in the area. It was soon appreciated that not
all of these services kept records on individual children. Those
which did were asked to provide identifier data for the children
known to them whom they considered to be vulnerable. Three
identifiers (gender, date of birth and postcode) were selected
which would enable the project team to attribute individual
children to wards and age bands, while maintaining their anonymity
and avoiding double counting.
Organisations were also asked to provide data on the ethnic
origin of their clients, as nationally Quality Protects management
action plans have been found weak in terms of references to
services for children from ethnic minority groups.
The data collected were collated and analysed for the area as a
whole, and then separately for each of the 39 local authority
wards. The results of the analysis were then compared with the
total number of children in the ward, national estimates of the
number of vulnerable children, and indicators of deprivation from
the Department for Transport, Local Government and the Regions.
Although the data were incomplete for reasons summarised later,
the study yielded some important information. In particular it
highlighted those wards in which either a large number, or large
proportion, of the child population was accessing services, and
those wards which appeared to have disproportionately high levels
of need, and which might therefore require more intensive input
Although the results presented few surprises to those working
with vulnerable children, it is hoped that the detailed information
presented on the different age groups and the use of services will
help to inform future planning decisions.
The analysis also indicated that the number of children actually
known to services was significantly lower than the number of
children estimated as vulnerable. This was further emphasised by
the individual ward level figures taken from the DTLR (formerly
DETR) Child Poverty Index.2 It is acknowledged that some
of these children may be recognised as vulnerable by other agencies
or groups, for example their school or a voluntary organisation
working with their family, but, as much of this data is not held
centrally, it was not incorporated in the overall figures for the
area. Agencies may need to think about how they share this kind of
information at a local level in order to identify vulnerable
children at an early stage.
The process of collecting the data for this project highlighted
practical issues which have important implications for the work of
the new strategic partnerships.
Definitions and the availability of data
Agencies that were asked to supply data for vulnerable children
were given the definitions contained in the assessment framework
3 and the draft guidance for children in need. These
definitions had not, however, been used by any of the agencies in
deciding which children were recorded. Agencies categorised
children in many different ways. Some took the view that a child
was automatically vulnerable if she or he was being seen by them.
Others used categories based on their statutory responsibilities
(for example, children who were in need, looked-after, on the child
protection register, or statemented). Some voluntary agencies
recorded all families who came to them for advice, so they were
self-defined as opposed to being professionally assessed as in need
In those organisations which held information centrally,
retrieval was relatively straightforward. In those which did not,
the individual caseloads of staff had to be collated. This was
naturally more time-consuming, and possibly less accurate than
centrally collected information as staff who were asked to do the
job had other priorities.
Some agencies, such as counselling services, saw many of their
clients through self-referral. As many clients did not know, or did
not wish to give, their postcode, the data they provided was
inevitably incomplete. Information about the ethnic origin of
children appeared to be routinely collected only by the social
services department and the NSPCC in the survey area.
Some of the data provided was incorrect. For example, up to 6
per cent of the postcodes provided by one of the larger agencies
were invalid. Dates of birth were incomplete in the data provided
by at least one of the other organisations. Some agencies provided
data which contained several sets of the same identifier
indicators, suggesting that individuals had been recorded more than
Information regarding the children who lived in families with
less than half the average household income was not available from
the local Benefits Agency, as they could not discern from their
database which of the income support claimants had dependants. Data
from education departments about children in receipt of free school
meals was used as a proxy for low income, though it is collected in
school catchment areas rather than wards.
Most organisations could provide their data electronically,
making input to a common database relatively simple. Some, however,
were only able to supply their data on paper, making manual input
into the common database necessary which was a time-consuming and
error prone task.
Attitudes to confidentiality varied substantially between
organisations in both the statutory and voluntary sectors. Most
organisations were willing to pass on the data requested once they
had been given appropriate assurances about how the data would be
treated. Some, however, felt able to provide only limited data;
these could not then be used to check for double-counting or to
attribute children to wards.
Others took the view that they could only release the requested
data if the parents of the children had given written permission.
Though this view was respected, such an approach obviously reduced
response rates, while at the same time risking raising fears
unnecessarily among parents as to how the data would be used.
Within the NHS, the Caldicott guidelines govern the release of
confidential information.4 However, even here there is
clearly room for interpretation, as one of the local health
providers released the information in full, while the other would
not do so without parental permission.
The study has highlighted a number of issues which may be
particularly pertinent to those involved in the new strategic
partnerships. In the light of our experience, it would seem
sensible for partnerships to agree a set of ground rules before
starting the process of mapping the vulnerable children in their
area. We suggest that these should include:
The definitions and categories to be used for vulnerable
A common data set, covering, at the minimum, date of birth,
gender, full postcode and ethnicity.
Systems to ensure that data are reliable, complete, and easily
retrievable by administrative staff.
Joint policies and protocols on confidentiality.
Philippa Hare is senior research officer, Martin Baxter is
project officer, and Elizabeth Newbronner is director at primary
care consultant Acton.Shapiro
1 Department of Health. Draft for Consultation, New
Guidance for Children’s Services Planning Consultation document
issued with LASSL (2000), April 2000
2 Department of Transport, Local Government and the
Regions, Indices of Deprivation 2000: Ward Level Presentations,
Index team at Oxford University for the DTLR (formerly DETR)
3 Department of Health, Framework for the Assessment
of Children in Need and Their Families, Stationery Office, 2000
4 Department of Health, The Protection and Use of
Patient Information, HSG (96) 18/LASSL (96) 5, DoH, March 1996
1 Draft guidance (reference 1) at www.doh.gov.uk/pdfs/cspconsultation.pdf
2 Indices of Deprivation (reference 2) at www.statistics.gov.uk/neighbourhood/metadata18.htm
3 Assessment framework (reference 3) at www.doh.gov.uk/scg/cin.htm