Memorandum 57
Supplementary evidence from Dr James Trussell,
Dr Kate Guthrie and Dr Sam Rowlands
RE: EVIDENCE
ON SCIENTIFIC
DEVELOPMENTS RELATING
TO THE
ABORTION ACT
1967 SUBMITTED TO
THE SELECT
COMMITTEE ON
SCIENCE AND
TECHNOLOGY
The evidence submitted to the Select Committee
on Science and Technology by Professor Patricia Casey cannot possibly
inform anyone about the effect of induced abortion on the subsequent
mental health of the woman. To understand why, it is helpful to
consider the ideal research design to answer this question. In
that ideal research design, women wishing to terminate their pregnancies
would be randomly assigned to receive a termination or have their
request denied without the possibility of their having a termination
elsewhere, as alluded to by Dr Rowlands in his oral evidence at
Question 83. Of course, no study has employed this design; it
is unlikely that such a trial will ever be undertaken. The research
design that is farthest from the ideal is one in which women obtaining
terminations are compared with women who got pregnant because
they wanted to become mothers. These simply are not comparable
groups because it is not only the abortion itself that differentiates
them. They are undoubtedly different in many ways (since one group
wanted to have a baby whereas the other did not), only some of
which are observable. So no matter how many other factors are
controlled in the statistical analysis (such as past mental health),
this design cannot be made to mimic random assignment. The design
that is next farthest from the ideal would compare women obtaining
terminations with all women giving birth, some of whose births
would be unintended; this is the design of the New Zealand study
by Fergusson et al so highly praised by Casey.
The design closest to the ideal would be to
compare women obtaining terminations with women whose request
for a termination was denied. Note that women are not randomly
assigned at all because some have their request denied. Presumably
their reasons for wanting a termination were not so compelling,
so perhaps they are "healthier" on average. The UK study
by Gilchrist et al is the only study with this design,
and it found that the two groups had identical rates of subsequent
psychiatric illness, psychotic illness, and deliberate self harm.
The design of the Gilchrist study is weakened by the fact that
women in each of the groups may have had prior induced abortions.
The data were indirectly standardised for age, marital status,
smoking habit, age at leaving full-time education, gravidity and
previous history of induced abortion at recruitment, since the
comparison groups differed on those characteristics (data available
from the authors). The authors also acknowledge that the study
had little power to detect important effects occurring in the
denied abortion and changed-their-mind groups because there were
too few women in these groups.
Casey cites only studies like the one from New
Zealand, all of which have the fatally flawed research design
identified above. Even if past and subsequent mental health is
measured perfectly, the flawed research design means that the
results cannot possibly be informative about whether induced abortion
is causally related to subsequent mental health. Casey's evidence
therefore does not answer this fundamental question and her review
of recent studies does not give the Committee any new insight.
REFERENCES
Fergusson D M, Horwood L J, Ridder E M. Abortion
in young women and subsequent mental health. J Child Psychol
Psychiatry 2006 Jan;47(1): 16-24.
Gilchrist A C, Hannaford P C, Frank P, Kay C R. Termination
of pregnancy and psychiatric morbidity. Br J Psychiatry 1995
Aug;167(2): 243-8.
October 2007
|
