I recently mentioned discussions about “parental
alienation” (PA) on the World Health Organization website, where there is
ongoing discussion of the 11th edition of the International
Classification of Diseases (ICD-11). PA advocates would like to include PA as
an index term—a term that can be found in the index, but in this case that
leads the reader to a much more general classification. People concerned about
the potential harm of PA “diagnoses” and treatments have opposed this.
If you are interested in this discussion you can go to
the WHO website and set up an account that lets you read and comment on
proposals.
I have just commented on one statement there, and I
want to elaborate here on what I said. On
the WHO website, a PA advocate cited a paper I published in Journal of Child
Custody, in which I said that the outcome research on PA treatments was at no
better than a “promising” level. The advocate used this statement to assert
that I had said PA treatments were promising and therefore ICD-11 should
include PA.
Unfortunately, this PA advocate apparently does not
understand that a “promising” treatment is not in fact particularly promising,
in the everyday sense of that term. The “promising” category is one that has
real meaning for people who do and use outcome research in psychology, but its
name is deceptive for others. This is a nuisance but I am not sure what we can
do about it now—except to educate people.
Research on the effectiveness of psychotherapies can
be evaluated as being at one of five levels of evidence. These levels are
essentially measures of the quality of the research and the extent to which
confidence can be placed in its results. The two highest levels of evidence,
randomized controlled trials and clinical controlled trials, are not identical
but share some important features. The crucial feature is that each method
builds in standards of comparison, so that the results of a treatment are compared to the results shown
in other circumstances, which may be no treatment at all, established
treatments, sham treatments (placebo conditions), etc. Because treatments
may—indeed are supposed to—have long-term effects, and for other reasons as
well, these two high-quality methods compare people receiving the treatment to
other, similar people who receive the control/comparison condition. Using such
comparisons is especially important in work with children and adolescents, who
are developing rapidly and whose natural changes may easily be mistaken for the
effects of a treatment.
When research at those two high levels of evidence
shows positive effects of a treatment, and when those effects are confirmed by
independent researchers, the treatment can properly be called evidence-based. “Promising”
research is work that has been designed and carried out at a lower level of
evidence, for example a simple before and after comparison of a group of
people, prior to and following their treatment. The “promising” nature of a treatment that
shows positive results under those circumstances is that it is worth doing more
complex research on the treatment, although there are no guarantees that the
treatment will be shown to be effective with more challenging designs.
PA treatments are “promising” because positive effects
have been reported when parents’ views of children’s attitudes after a
treatment have been compared to their views before a treatment. As there is no
standardized way to evaluate a child’s PA as present or absent, much less to
quantify it, there have been no comparisons of the children’s attitudes
themselves before and after treatment. And there certainly are no published
studies reporting randomized controlled trials or clinical controlled trials of
PA treatments, using controls which are essential for work on treatments for
quickly-changing children and adolescents. These facts are among the many
reasons why ICD-11 will not contain PA as a disorder.
When you see that a treatment is “promising”, be sure
to consider exactly what is being promised!