The DTD field trial replication (DTDFT-R)

Julian D. Ford (UConn School of Medicine), Joseph Spinazzola (Foundation Trust) and Bessel van der Kolk (Boston University), European Journal of Psychotraumatology 12(1):1929028, 28 June 2021. T1 — peer-reviewed, structured interviews, full numbers published, and a replication of a study the same team ran before.

This is the evidence behind the one sentence psychological-trauma could not source: whether developmental trauma disorder is a real category distinct from PTSD, and where it stands officially.

The status question, answered

DTD is not in the DSM-5 and not in ICD-11. It was proposed for DSM-5 by van der Kolk and colleagues out of the National Child Traumatic Stress Network and rejected; the DSM subcommittee's stated reason was a lack of evidence for the diagnosis in its submitted form. What the WHO did admit, into ICD-11, is complex PTSD — PTSD's three clusters plus disturbances of self-organization (affect dysregulation, negative self-concept, relational difficulty). C-PTSD is not in DSM-5 either.

So the field has two competing framings for the same children and one of them has an official code. This paper's authors concede the collision in their own discussion: ICD-11 C-PTSD's self-organization symptoms are "a potential alternative or complement to DTD," they did not measure it, and they call for a study putting PTSD, DSO and DTD side by side. That study is not here.

What was measured

271 children aged 8–18 (M 12.1, 47% female) in intensive mental-health treatment, recruited October 2014 – November 2016 across six sites — three northeastern, three west coast, urban to rural — where the original trial ran fewer. 84% outpatient, 16% residential; a third living with both birth parents, a fifth in foster or adoptive care. Every child screened with the DTD Semi-Structured Interview (v10.6) and KSADS-PL modules.

The finding is a count of other diagnoses. Mean comorbid diagnoses: 5.7 for DTD with PTSD, 5.2 for DTD without PTSD, 3.8 for PTSD without DTD, 2.1 for neither (F[3,267] = 55.49, p < .001). A child who meets DTD criteria and not PTSD still carries more comorbidity than a child with PTSD alone — which is the whole argument, since it is what a merely redundant label could not do.

Unadjusted odds ratios for DTD are large: major depression 9.05 [4.03–19.13], oppositional defiant disorder 9.00 [4.78–16.96], separation anxiety 5.25, ADHD 5.09, conduct disorder 4.12, suicidal ideation 4.04. After controlling for demographics and every other psychiatric disorder, three survive as independently DTD-associated — oppositional defiant disorder (OR 3.47, p = .005), depression (2.88, p = .041), separation anxiety (2.31, p = .034) — against one for PTSD (separation anxiety, 3.74). Both predict suicidality.

A team replicating itself

The corpus keeps a running note that nobody re-runs anybody's numbers. This is a re-run, and it is a third kind of one, distinct from the two the note has been counting. It is not an independent team checking someone else's result, and it is not an absent re-measurement — it is the same three authors replicating themselves at new sites with the same instrument. That buys generalization across geography and setting, which was plainly the point. It does not buy independence: the people who designed the diagnosis, wrote its interview and would benefit from its adoption are also the ones scoring it. The paper is honest about the design; it simply cannot answer the question its own existence raises.

Rated T1 on method and reporting, and read with that conflict on the page rather than in a footnote.

Related

psychological-trauma · bessel-van-der-kolk · julian-d-ford · joseph-spinazzola · the-body-keeps-the-score · replication-crisis · synthesis