EMDR vs. CBT for Child Trauma Treatment

Blog Home

Blog Archive




May 07, 2017 at 3:10 PM

Is there a best child trauma therapy yet?

Trauma-focused cognitive-behavioral therapy (TF-CBT) appears to be the clear leader in that it has a ton of research support (Cohen, Mannarino, & Deblinger, 2017). Yet eye movement desensitization and reprocessing (EMDR) also has a fair bit of research support, and was more efficient than TF-CBT in two of the three direct comparisons.

One problem with interpreting the findings is that each comparison study used a different version of TF-CBT, and may (or may not) have used a different version of EMDR as well. Another problem is determining the extent to which each of the studies’ findings might apply to clinical practice.

The first study (Jaberghaderi et al, 2004) compared EMDR to a TF-CBT variant for 16 sexually abused 12 to 13-year-old girls in Teheran, who were recruited to volunteer in the study. The therapists – one for each treatment condition – had been appropriately trained, but treatment fidelity was not evaluated. The CBT procedure was based on Deblinger and Heflin (1996) – the precursor to the Cohen et al version – with additional activities from Camino (2000). The treatment was primarily individual, with a single parent psychoeducational session provided in the same way across treatment conditions.

One child dropped out of each group after the treatment had started, so dropout rates were equal. There was a nonsignificant trend for greater treatment effect for EMDR on the various outcomes. Comparing treatment efficiency in this study was not ideal because the CBT condition required a minimum of 10 sessions, whereas the EMDR condition had no such requirement. Even so, only one of the CBT participants ended after 10 sessions; one finished after 11, and the remainder used the maximum of 12 sessions. So the minimum number of sessions requirement did not, for the most part, artificially extend treatment duration. In fact, three of the seven CBT participants were referred for further treatment after the study. Those in the EMDR condition used between four and eight sessions, and none were referred for further treatment. In this study, EMDR was significantly more efficient – calculated as symptom reduction per treatment session – than CBT.

The second study (de Roos et al, 2011) compared EMDR with a Dutch CBT variant for 52 disaster-exposed children and teens. This was a field study conducted with help-seeking clients (and their parents) at the local community mental health center. Eight therapists in the clinic were trained and supervised (by an expert in both treatments) in each treatment condition, and those who screened into the study were assigned a therapist and then randomized to treatment condition. Parental guidance sessions were also provided.

No child dropped out of either group once the treatment had started, so dropout rates were equal. Outcomes were excellent, with no significant difference between treatment conditions. EMDR was significantly more efficient than CBT.

The final comparison study to date (Diehl et al, 2015), also conducted in the Netherlands, compared EMDR to the current (American) version of TF-CBT for 48 children/teens with a variety of single and multiple event traumas, who were recruited to volunteer in the study. Eight therapists were trained and supervised by experts in the respective treatments; participants were assigned to therapists and then randomly assigned to treatment condition. There was some parental involvement in each treatment condition, but this was not equalized; there was more in the TF-CBT condition.

One child dropped out of each treatment condition once the trauma work had started; it’s not known whether this was because they didn’t like the treatment, already felt better, or for some other reason. Both treatments were effective in reducing PTSD and related symptoms, a couple of perhaps-related symptoms reduced more for those in the TF-CBT condition, but there were no significant differences across conditions (I know that seems contradictory). There was a nonsignificant trend for EMDR to be completed in less time.


So what to make of the comparison studies’ findings so far? A few things:

1. EMDR and TF-CBT are both effective treatments for trauma-exposed children and teens. Most participants completed treatment, and most got better. The EMDR participants got a little more better in the Jaberghaderi et al study, but the TF-CBT variant was a bit unique, and treatment fidelity was not evaluated. The TF-CBT participants got a little more better in some tangential symptoms in the Diehl et al study, which also involved more parental involvement in the TF-CBT condition. Regardless, participants overall did well in each of the treatment conditions, across studies.

2. EMDR is probably more efficient than TF-CBT. This was a non-significant trend in Diehl et al, and a significant finding in the other two studies. This finding was especially meaningful in the de Roos et al study, in that the CBT condition achieved results considerably faster than would have been expected in the Cohen et al version of TF-CBT. Furthermore, in the Diehl et al study, the standard TF-CBT was shortened considerably, meaning that standard TF-CBT would be expected to take more time; EMDR was not shortened for this study.

3. Looking only at these studies, dropout rate seems to be about the same for each treatment. One problem in interpreting these results is that participants tend to drop out of studies at lower rates than they drop out of trauma treatment in practice settings (Najavits, 2015). Therefore the de Roos et al study is most relevant here because it was a field study with help-seeking clients; still it is only one study.

Looking beyond these studies, I confess to being skeptical of TF-CBT’s applicability to clinical practice. Yes, my first claim to fame was as an expert on EMDR for children, but I don’t think I’m biased; I’ve also published on several other trauma treatments (and taught a couple of them). I don’t care what it’s called or who invented it, as long as it works. And TF-CBT does work, but not nearly as frequently as one might hope. When I speak with people whose clinics advertise TF-CBT as their primary treatment model, I generally learn that only a tiny percentage of their clients actually make it through the trauma work. This may also be an issue with EMDR, though it can probably be overcome with proper therapist training (Greenwald, 2006). A “gold standard” treatment cannot be proclaimed until we are confident that it can be routinely implemented in clinical practice, with a low dropout rate (Najavits, 2015).


For now, I prefer EMDR for its efficiency. If I were going to learn a CBT method for treating traumatized children, I’d probably choose to learn the Dutch version, which is apparently well tolerated as well as relatively fast. And we still need large-scale implementation studies to determine whether either EMDR or TF-CBT (or variants thereof) can reliably and effectively treat children and teens in clinical practice.

Camino, L. (2000). Treating sexually abused boys: A practical guide for therapists & counselors. San Francisco: Jossey-Bass.

Cohen, J. A., Mannarino, A. P., & Deblinger, E. (2017). Treating trauma and traumatic grief in children and adolescents, 2nd Edition. NY: Guilford.

de Roos, C., Greenwald, R., den Hollander-Gijsman, M., Noorthoorn, E., van Buuren, S., & de Jongh, A. (2011). A randomized comparison of CBT and EMDR for disaster-exposed children. European Journal of Psychotraumatology, 2, 5694. DOI: 10.3402/ejpt.v2i0.5694.

Diehle, J., Opmeer, B.C., Boer, F., Mannarino, A.P., Lindauer, R. J. L. (2015). Trauma-focused cognitive behavioral therapy or eye movement desensitization and reprocessing: what works in children with posttraumatic stress symptoms? A randomized controlled trial. European Child & Adolescent Psychiatry, 2, 227-36. doi: 10.1007/s00787-014-0572-5.

Greenwald, R. (2006). The peanut butter and jelly problem: In search of a better EMDR training model. EMDR Practitioner.

Jaberghaderi, N., Greenwald, R., Rubin, A., Zand, S. O., & Dolatabadi S. (2004). A comparison of CBT and EMDR for sexually abused Iranian girls. Clinical Psychology and Psychotherapy, 11, 358-368.

Najavits, L. M. (2015). The problem of dropout from “gold standard” PTSD therapies. F1000Prime Reports, 7, 43. DOI: 10.12703/P7-43


1708 hits

Please add a comment

Posted by Elise on
Briefly, common factors (S. Rosensweig) may be relevant. The search for a gold standard misses the fact that some treatments work for some individuals.
Posted by Ricky Greenwald on
Common factors are important, but trauma-specific treatments outperform other treatments, e.g., Ehlers, A., Bisson, J., Clark, D. M., Creamer, M., Pilling, S., Richards, D., Schnurr, P. P., Turner, S., & Yule, W. (2010). Do all psychological treatments really work the same in posttraumatic stress disorder? Clinical Psychology Review, 30, 269-276.

While it's important to take common factors into account, it's also worthwhile to determine whether or not some treatments outperform others. Once that has been determined, then the next step is, as you suggested, to determine whether there are definable individual characteristics that would lead a generally less preferred treatment to be more suitable for certain people.
Posted by Robert Olcott on
I didn't see a specific reference to the O'Hare/Paulsen [EMDR] child Protocols..., but I only first quickly scanned the article.... My library's public terminal is closing...
Posted by Ricky Greenwald on
Right. Because those have not been researched. This post was focused on the leading research-supported child trauma therapies.
Leave a Reply

(Your email will not be publicly displayed.)

Captcha Code

Click the image to see another captcha.