Memorandum by Mr David Thrower - continued
Other Current Health Authority Studies
135. Other health authorities that are known
to be actively studying the incidence of autism, and investigating
reported increases in most cases, include:
Bromley (results expected summer
Cornwall and The Isles of Scilly
(results autumn 1999).
Hillingdon (results expected shortly).
Norfolk (results expected shortly).
North Cheshire/South Cheshire (joint
study; results not expected for some time).
136. Other authorities with some data available
include North and East Devon, North Yorkshire, Portsmouth and
South East Hampshire, Somerset and West Kent. Further details
are still being sought. The results of all these studies will
be brought to the attention of the Health Committee. Full acknowledgement
is given to the assistance of each Health Authority concerned.
It is clear that at least some authorities are endeavouring to
fill the "initiative vacuum" left by the Department
Childrens' Act Register and Learning Disability
137. One Health Authority (Lincolnshire)
has candidly admitted that it does not have systems in place at
present for monitoring the incidence of autism, but has stated
that the Children's Act Register and the revised Learning Disability
Register should now begin to address this (letter of 11 May 1999).
138. It is also expected that, in about
five years' time, data may potentially become capable of being
extracted from the proposed National Health Surveillance Network.
However, it is not clear that this will happen in practise, and
that it will not be blocked by the Department on the basis of
alleged "classification difficulties".
An Example of Good EstimatingAutism in
139. Encouraging detailed local assessment
of incidence, as part of a clear structured analysis led centrally
by the Department of Health, is a central part of investigating
the allegations of parents that their child became autistic after
140. An example of a Health Authority where
a determined attempt has been made to obtain data on autism incidence
is Harrow. The example is set out in some detail here, in the
context of the Committee's Inquiry, because the case demonstrates
that there are problems in assembling data on incidence, but also
that these problems can, with an effort, be confronted by the
141. The example exposes the Department
of Health's arguments, as to why the Department cannot confront
the issue of whether there has been an increase in autism post-MMR,
as largely a fallacy.
The Harrow study, "A Report on the Needs
of Autistic Spectrum Children and Adults and Their Families in
Harrow", Simon Jarrett, National Autistic Society 1998,
produced for Brent and Harrow Health Authority and Harrow Social
Services, acknowledges that until a detailed counting exercise
is carried out, it is impossible to confirm nationally-based prevalence
142. The study quotes "national prevalence"
rates at 4.5 per 10,000 (1 in 2,222) classic autism and 26 per
10,000 (1 in 385) other autistic spectrum including Asperger's.
These two rates offer a combined incidence of one in 328, virtually
identical to the California rate discussed later, and very close
to several other UK rates where hard local data has been assembled.
The source of the "national prevalence" rates, which
are very high compared with rates quoted in past published literature,
is assumed to be the NAS itself.
143. But the study does confirm that there
are indicators that enable accurate estimations to be made. This
begs two immediate questions:
why does the Department not initiate
a proper assessment, creating a full autism database, as recommended
by the Health Committee?
in the meantime, why does the Department
not work with health authorities such as Harrow to identify and
refine the indicators of incidence, to enable a refined and better-informed
estimate of prevalence to be established in the meantime, that
could be used by all Health Authorities and the Department itself?
144. The Harrow report also notes that the
Social Services Department has a learning disability register,
but that this register does not include:
children with Asperger's, unless
they have significant other problems;
children with a degree of autism
that still allows them to function in a moderate learning difficulty
children in mainstream schools; and
children whose referral to the learning
disability team has been turned down.
145. The Harrow report notes that there
appears to be a sharply rising incidence in autism at the lower
end of the age range, but then seeks to explain this as being
due to factors such as reduced therapy services with increased
age. It also takes the view that the highest rates of prevalence
it has found, for the youngest children, can then be applied across
the board to all age ranges, but there is no actual local autism
census available to back up this claim.
146. The estimated prevalence rates found
as a result of the Harrow study are quoted as close to or in excess
of 1 in 328 for children aged 0 to four, and 1 in 418 for children
aged five to nine. The report also notes that:
"Both parents and professionals . . . were
virtually unanimous in talking about the high numbers of children
with an autistic diagnosis passing through the system . . . The
unanimity of this view suggests that there is some substance to
it . . . Others maintained that numbers were rising across
the country, and that they had discussed this with colleagues
at conferences." (my emphasis)
147. On the argument that rising numbers
are due to better diagnosis, a view expressed by the Department
of Health and some health authorities, the report stated unequivocally
"Most professionals felt that these factors
might account for a small increase in numbers, but are not an
explanation for the general perceived increase."
148. In other words, the view is that there
is a real underlying increase in incidence.
149. Mapping of membership of the National
Autistic Society in Harrow has revealed clusters of cases in West
Harrow/Rayners Lane, in North Harrow, and in Stanmore. The report
noted that the existence of clusters was acknowledged by local
professionals, that some GP practises had far higher numbers on
their caseload than others, and that clustering had been reported
elsewhere in the UK. It again noted the need for a "proper
counting exercise" to be undertaken to shed more light on
150. The report concludes:
"It is imperative that a counting exercise
is carried out to establish as precisely as possible the numbers
and types of autistic spectrum disorders . . . with background
on age, ethnic group, geographical location etc. Once a database
is established it should be maintained and updated to give reliable
information on incidence and trends"
A Parallel Case StudyAutism In California
151. Parents investigating the cases of
their children are also well aware of parallel events elsewhere,
for example in the United States.
152. In the US, a number of autism research
and parents' organisations called in April 1999 for independent
scientific research into whether increased vaccination in early
childhood is a contributingco-factor to the development of autism.
The background to this call, which mirrors the call of Professor
Sir John Pattison in the UK in March 1998, is the apparent rising
tide of autism.
153. According to the California State Department
of Developmental Services, in 1998 there were 1,685 new cases
of children diagnosed with autism, giving an incidence rate of
1 in 312 (offering comparison with 1 in 333 in Bexley, London).
The figures include more than 13,500 children aged 0-3 in the
State's Early Start Program yet to be diagnosed, most of whom
present with language and developmental delays.
154. Autism as a percentage of the total
client population served by California's regional centres nearly
doubled between 1987 and 1998, with the most significant increases
seen in young children less than nine years old.
155. Hitherto, it had been widely accepted
that between two to five cases of autism occur per 10,000 live
births. Based on these estimates, only between 1 in 5,000 and
1 in 2,000 births in California should eventually be diagnosed
with autism, instead of the 1 in 312 which actually occurred in
1998. The California State DDS report stated that:
"The number of persons entering the system
with autism has increased dramatically over the past 11 years
relative to the other three developmental disabilities (cerebral
palsy, epilepsy and mental retardation).
156. It is worth emphasising that the concerns
of parents in the UK, expressed earlier, are mirrored by those
of professionals and parents in the United States:
"There has been no concerted national effort
to determine the actual incidence of autism. . . . The most frequently
quoted incidence figure, based on large scale surveys in the United
States and in England, is 4.5 per 10,000 . . . (or 1 in 2,222)
. . . not including other spectrum disorders." (quoted source:
"A Report to the Legislature: Changes In The Population
of Persons With Autism And Pervasive Developmental Disorders In
California's Developmental Services System, 1987-98", Department
of Developmental Services, California Health & Human Services
Agency, Sacramento, 1 March 1999, section IV, Rates of Occurrence,
"Presently the CDC admits to an estimated
autism incidence of 1 in 500 (contrasted with an earlier estimate
of 1 in 10,000). The actual incidence is suspected by the autism
community to be at least this high, perhaps 1 in 200, and to be
increasing exponentially". (source: Laura J. Ruede, parent,
Fort Worth, Texas)
157. The Assessment Evaluation and Support
Unit of the California Department of Education's Special Education
Division gave the following statement about increases in the numbers
of special education students with autism.
"We first started collecting enrolments
counts on autism the 1992-93 school year. On 1 April 1993 there
were 2,157 students (out of a total of 540,472) reported . . .
. (to have autism). Five years later, on 1 April 1998, there were
8,084 students (out of 632,238) reported".
158. In other words, incidence of registered
autism in enrolment registers has increased in California, using
consistent diagnostic criteria and using the same local health
professionals, from 1 in 251 to 1 in 78 in only five years, a
rise of 222 per cent.
159. Other parts of the United States report
steep rises. The Illinois State Board of Education lists figures
for the incidence of autism, rising from 317 in 1991 to 2,305
in 1997, a rise of 627 per cent in six years.
160. The above is quoted at some length
because it underpins the view of UK parents that their children
may be part of a wider international health problem, connected
with vaccines that are manufactured and distributed by multinational
companies, the recognition of which is being resisted by the Department
of Health, and its overseas counterparts.
161. However, the UK Department of Health
portrays a wholly different view of MMR and autism in the United
"Millions of children have received the
MMR vaccine in other countries, such as Finland and the United
States, prior to its introduction in the UK. No serious long-term
complications have been identified". (My emphasis. Source:
letter of 19 March 1998 from UK Department of Health Immunisation
and Communicable Disease Branch).
162. The UK Department's letter may reflect
the previously-expressed views of the US CDC, but clearly does
not acknowledge the fact that these views are in complete contrast
to the current views of many US parents nor does it acknowledge
the major problem being experienced in at least parts of the US
in terms of autism incidence.
DfEE StatisticsAnother Source of Data?
163. Although the Health Committee's inquiry
is focused upon issues of health rather than education, the Department
of Education (for England) and its equivalents could in theory
provide valuable information to enable the monitoring of health
164. One way of parents trying to investigate
their child's degeneration is to look at the statistics maintained
by the Department of Education on statements for children with
special educational needs. However, again the picture is one of
variation between active action (Scotland) and apparent policy
165. The position in England is revealed
through the following quotes:
"The Department still does not break down
the numbers of children being statemented with special educational
needs by type of special need . . . Statementing practice varies
widely across the country, making collection of comparable statistics
extremely difficult . . . (We have) previously discussed theories
to explain the increase in the number of statements in recent
years. There has been no scientific research into the reasons
behind the increases" (letter of DfEE of 18 February
1999, my emphasis).
"The increase in the number of pupils with
statements is a matter of concern for the DfEE . . . There has
been a steep increase in recent years . . . It is not possible
to say why there has been such a huge increase in statements .
. . The Department has not made any investigations in the past
as to exactly what is driving these numbers upwards."
(letter of DfEE of 30 April 1998, my emphasis.
"The number of pupils with statements rose
from 153,228 in 1991 to 232,995 in 1995 (England only). It is
not possible to say whether the increase in statements has been
due to a rise in the numbers of one particular type of special
need, such as severe learning difficulties . . . With regard to
autism, a 1997 Health Select Committee Report said that no data
is collected centrally by the DoH giving information on the number
of pupils diagnosed as having spectrum disorders. The report
recommended that the DoH should take the necessary steps to redress
this" (letter of DfEE of 14 April 1999, again my emphasis).
(Footnote: the Department of Health has, as already
noted, refused to do this. A postscript to this correspondence
is that the updated figure for statements, England only, climbed
further in 1998, to 242,294).
166. In connection with the 58 per cent
figure quoted above, it is also interesting to note that the number
of children with statements in Northern Ireland increased by 60
per cent between 1990 and 1997 (source: Special Education Branch,
Department of Education Northern Ireland, letter of 18 March 1999).
The figures for Scotland and for Wales are being sought but have
not yet been obtained.
167. The two increases quoted must beg obvious
questions in relation to whether the incidence of learning disability
is being better recognised, or whether the statementing system
itself is being more proactively used by parents, or whether there
is an underlying health problem, or a combination of these three
168. The Department of Education quotes
the Warnock Report as reason for not breaking-down total national
(England) figures for learning disability or statementing into
categories that might give an insight into the incidence of autism
amongst school age children. The position was updated in a letter
of 19 May 1999 to Lord Clement-Jones from the Minister of State,
"The Warnock Committee recommendation which
led the Department to discontinue collecting statistical data
on pupils by "type" was that recommending abolition
of statutory categorisation of handicapped pupils . . . The Warnock
Committee did recommend a feasibility study into the use of a
more complex grid approach for collecting data. Its purpose would
be to measure numbers of children with different combinations
of special educational needs and numbers with particular degrees
of any given special educational need . . ." (source: Lord
169. As the numbers of children, including
children with autism, that have been statemented has grown in
the past seven years by 58 per cent, this refusal to investigate
the underlying causes of any increase in statements, or even to
categorise the figures into broad areas of disability, seems incomprehensible
to parents trying to investigate whether the condition that they
suspect their child has degenerated into is part of a wider syndrome,
and whether the level of incidence of that syndrome has recently
increased or not.
170. A wholly different approach has been
taken in Scotland, where specific questions on number of autistic
pupils, broken down by junior/senior school and by sex, were included
in the September 1998 Scottish schools census. The results were
anticipated to be available by mid-1999 (source: Scottish Office
Education and Industry Department), and will be passed to the
Committee in time for its Inquiry.
171. This difference of approach appears
inexplicable to parents trying to understand whether their child's
condition is part of a wider national/international health problem.
The approach taken in Scotland is referred to by Baroness Blackstone:
"The Scottish education system does of course
differ in significant respects . . . there is no reason why each
country should collect identical statistics. We are not convinced
that a list akin to that introduced in Scotland would give sufficiently
useful information to merit our introducing it . . . It is unclear
how data on local incidence of autism would assist in policy decisions
. . ."
172. The above explanation appears unconvincing
in the extreme. Baroness Blackstone has, however, put forward
a positive suggestion:
"The regional groupings of local education
authorities which we are promoting . . . will be piloting the
collection of data on incidence of particular needs. We shall
be looking at their experience, and whether we can build on it
to put together meaningful statistics at national level."
173. Key ConclusionsIt is extraordinary
and wholly inexplicable that the Department of Health has failed
to properly monitor the incidence of autism and associated disorders,
despite having already been urged by the Health Committee to do
174. It is even more extraordinary that
they have no future plans to do so, nor are they actively encouraging
by example the monitoring of autism at the local level. This failure
cannot be helpful to parents investigating their child's condition,
nor to central and local government planning future demand for
services. Parents are sceptical as to the Department's reasons
for inaction, and believe it may have ulterior motives for not
investigating prevalence and causes.
175. The efforts of a few Health Authorities
will eventually bear fruit, but the failure of the Department
of Health to properly co-ordinate their efforts and use consistent
criteria and mechanisms for data-gathering will limit the value
of this data, by preventing it being used as building-blocks to
form part of a national picture.
176. Many Health Authorities still seem
to have no data at all, and some are using estimates based upon
outdated textbooks. It seems inconceivable that autism in Bexley
really is 75 times more prevalent than in Salford, and this suggests
that there is a great deal of catching-up in data collection to
be done by some authorities.
177. If the levels being found in Bexley,
West Kent, West Sussex and other leading-edge health authorities
eventually turn out to be nation-wide, then it implies that there
may have been a substantial but unrecognised increase in autism,
either because of increased incidence or better diagnosis, but
probably both. In the context of the implications of the data
from California, proper data-gathering is essential.
178. Issues for Discussion:
how can local monitoring be structured
better, and brought up to the level of the best being undertaken
by a minority of health authorities?
if several health authorities have
very high incidence rates (between 1 in 300 and 1 in 500), does
this imply much higher incidence than is currently realised in
how will the development of the Children's
Act Register, the Learning Disability Register and the National
Health Surveillance Network interrelate, and how will these enable
the Department and local health authorities (and other local agencies)
to properly assess incidence of autism, monitor it for the future,
and provide data for meeting needs?
does the recent study in Harrow,
and the efforts of several leading-edge health authorities, offer
a way forward, and how can the experience of these be quickly
put to a wider use in advance of the National Health Surveillance
can the benefits of the Scottish
education census approach be quickly and simply applied to the
remainder of the UK, as a further starting-point?
will LEA regional groupings work
closely together so that their information on autism and other
related disorders (both for educationally-statemented children
and non-statemented children) is gathered on a consistent UK basiseven
if not 100 per cent compatible with data elsewhere in the worldso
that it can form a useful database for health monitoring purposes?
can the Department of Health be very
strongly encouraged to co-operate with the DfEE and regional LEA
groupings on this issue?
179. Suggested Recommendations for Action:
rapid agreement of a nationwide "template"
for monitoring autism, so that locally-gathered data is compatible
between different parts of the UK;
early implementation of a nationwide
early networking of local data and
best health authority practice as part of the above, including
all authorities where very high rates of incidence are now apparent;
extension of the Scottish education
census in relation to autism, to the remainder of the UK through
the LEA regional groupings, but to national standards of consistency
(issue for education rather than health, but with health data
joint working group between DfEE,
DSS and DoH to investigate autism, including education and care