15 May, 2013

Evidence for treatments for abused/neglected children - part 2

This is part 2 of a series of posts looking at the treatments that parents of abused and/or neglected children might be offered. Part 1 which explains a bit of the background, and a bit more about why evidence matters, is here.

So, moving on from the names mentioned in the previous post, we come to Bessel Van der Kolk.  Here, I'm starting to feel a little more confident.  The research (and there is actually research, with actual data) is grounded in well-established theories - I'm not a specialist in PTSD, but I do know that it's a well-established disorder, that we know it happens in children as well as in adults, and that a lot of research groups work on the disorder. This all gives us confidence both that the disorder exists, and that no one group of psychologists is going to be the only ones saying something, which could just be their personal theory.

A couple of interesting background articles include this one which talks about multiple instances of trauma in childhood (e.g. through repeated abuse, or different types of abuse) and whether this might increase the complexity of PTSD. As I say I'm not a specialist but they include neglect and abandonment as two categories of childhood abuse (but importantly they are ONLY two categories and they weren't the most common either).

They found that more instances of abuse did increase the complexity of PTSD, and interestingly the more trauma there was in childhood the worse were symptoms - but more trauma in adulthood didn't mean more symptoms.

I know many parents whose children were neglected but there's no evidence of abuse are interested in how this can affect them - it's hard to say specifically from this paper but because there were several possible traumas in childhood, it doesn't look like "only" being neglected is likely to lead to complex PTSD.  I'd have to do more reading to say what the link was between neglect and any PTSD at all, though.

The diagnosis of complex PTSD does seem to be applied to a proportion of children who have been abused and/or neglected and Van der Kolk also has some ideas on treatment, such as this chapter. The suggestion is for what he calls Phase Oriented Treatment, which briefly seems to start with symptom management and move on via other steps to therapies including Exposure Therapy and EMDR.

However, I started to get a little sceptical when I looked up this treatment programme (it's not so much a treatment as a programme of different treatments, that must be applied in a particular order). I can't find a group that has evaluated the programme in a valid way - no controlled trials. Van der Kolk in this same chapter says that his programme of treatment for adults (note not children, and we have to be very wary of recommending treatments for children on the grounds that they work for adults) works better with PTSD patients, citing a controlled trial. I can't find the precise trial he talks about, but another paper  (pdf) by the same group says that the two treatments were just as effective as each other, and the part of that study that Van der Kolk calls "prolonged exposure and cognitive restructuring" - his recommended treatment - seems to be CBT, about which a lot is known, and seems to be effective.

I find it a bit odd that Van der Kolk uses this study to suggest that CBT (or PE/CR as he calls it) is no use for PTSD while the original researchers suggest it is good for PTSD.  And it's all very well to say (even if it's true) that "lots of patients drop out in conventional PTSD treatment" (the figure was 30%) but to go on to say, which he does, they are "harmed" by it seems a bit strong.  The study he's citing here does not actually include the programme of treatment he recommends, anyway. So I'm not sure what to conclude about his treatments.

Although Van der Kolk doesn't seem to be saying that his treatment recommendations in general are all about EMDR (which is an eye-movement based treatment, but which does have something in common with CBT. It has a bit of a bad press among a few quack-watch types but there does seem to be some evidence for it), he himself has published a controlled trial (pdf) of EMDR in adults (versus a drug treatment or a drug placebo - there wasn't a therapy control group), and there's another study of EMDR versus a waiting list group in children. As I said in my previous post, waiting list (or no therapy control groups) are not great to evaluate psychological therapies - just talking generally to a helpful therapist is likely to make people feel better.

Both of these studies though seem to show that EMDR helps PTSD symptoms - and continued to help after 2 months in the child study and 6 months in the adult study.

Phew! After posting this the very nice Keith Laws sent me a metanalysis (a research study that analyses other research studies) here (pdf) specifically looking at EMDR and PTSD in children, which does look at some studies that compare EMDR to other therapies. The conclusion is it's a lot better than waiting list or other types of treatment that aren't particularly tailored to PTSD, and it's actually a little better than CBT (about which more anon). But there were, really, very few studies that did a full, direct comparison.

Recommend to a friend test: I think a cautious yes on the EMDR, but most of the evidence is only versus no therapy.  Though Van der Kolk's theories on the development of PTSD in children are backed up, there doesn't seem to be anything at all backing up his ideas on his programme of therapy - that's a definite No, especially as it seems to be so very time consuming and lengthy (and therefore expensive. Am I very wrong to hypothesise that he charges a lot? And charges a lot to train people to do it?). He himself criticises dropout rates in other therapies but doesn't tell us how many people drop out from his multi-phase treatment.

It's time to talk more here about treatment for PTSD in general. The NICE recommendations (pdf - the NHS best clinical evidence recommendations) say that there isn't enough evidence of complex PTSD being different from "regular" PTSD so they look at treatments for all kinds of PTSD (other researchers say this too. Let's be conservative and say more trauma in childhood probably = worse PTSD). Because it's difficult to diagnose in children (Van der Kolk's first study is of outcomes in adults whose trauma was in childhood), and there aren't that many studies, most of the studies are on those who have experienced sexual abuse - though behaviour and mental health difficulties are, as Van der Kolk says himself, similar.

There are some controlled trials here - phew!

For older children, there seems to be some help from CBT (compared to "supportive therapy", in other words talking to a therapist without a specific type or plan for treatment) on some aspects of children's behaviour and mental health, but not others. For children under 7, it's hard to measure mental health directly but CBT also seems to help children's behaviour. None of the studies went on for long enough to tell if the CBT affected children's behaviour or mental health in the long term. As families often drop out of therapy when they don't find it's helpful, they also looked at whether the dropout rates differed.  Again, it wasn't possible to tell.  These all seem to have been studies of trauma-focussed CBT, in particular.

Interestingly some practitioners  and some of the trials, compared what happened when the mother had CBT with giving it to the child.  I wasn't too surprised to read that it didn't help to give this instead or as well!

RtaF: As far as we can tell this particular type of CBT does help PTSD in children, at least in the short term, and it helps quite young children too. I would recommend this, but would love to see more data.

Before I finish what was supposed to be a concluding post but is going to be one of at least a couple more, I can tell, I can see (!) I wanted to comment on another theory that Van der Kolk has, which is that as well as the category of PTSD there should be a category of Developmental Trauma, for children who have suffered prolonged abuse and neglect and have consequent behavioural and mental health problems.  This may or may not be a separate category of disorders according to the "official" manual used to classify psychiatric problems (the DSM). A new version is about to come out and I can't find out if it is going to include this.

Van der Kolk, among others, has suggested this would be a good idea and that it should form a separate diagnosis.  He thinks it would be a better diagnosis than PTSD, and he's probably right in saying that there are more symptoms in a lot of children than would be covered under this.  But as he also says (pdf), many children who have suffered abuse meet  another diagnostic criterion.

And there hasn't been any particular treatment proposed for this group as a whole - who's to say that lumping all these children together isn't a mistake too, and that you need to look more closely at their symptoms and/or other diagnoses they might have, to treat them effectively? The consensus among some researchers is that if the diagnosis exists with an emphasis on cause, there might be too much temptation to diagnose all children with this background with the same thing, regardless of what problems they have. Children not only may have differing problems but vary in their resilience in the face of trauma - so will have differing degrees of the same problems.