Background

Comprehensive DBT for adolescents (DBT-A) is among the best-supported interventions for reducing life-threatening behavior in young people, delivering its skill-acquisition and generalization functions through individual therapy, group skills coaching, intersession phone coaching, and a clinician consultation team. Yet group skills training has known limitations: scheduling and insurance barriers, a generalized format that may not fit an individual’s real-world triggers, and group norms that discourage discussing target behaviors in depth. These gaps raise a practical question — how do we move a client from knowing a skill to reliably using it under acute distress?

Our team piloted an augmentation we call Enhanced DBT-A: standard comprehensive DBT-A plus Individualized Skills Coaching Sessions (ISCS). This post summarizes what we found and, just as importantly, what the study cannot tell us.

The intervention: ISCS

ISCS are 45–60 minute one-on-one sessions delivered by a skills coach — typically someone other than the client’s primary therapist — and deliberately distinct from individual therapy. Coaches do not conduct emotional processing or problem assessment. Instead, sessions target a specific skill deficit, often one surfaced by a behavioral chain analysis in individual therapy, and center on behavioral rehearsal: having the client teach the skill back, over-practice it, troubleshoot barriers, and adapt it for acceptability. Commitment, reciprocal, and stylistic DBT strategies (including game-based practice) are used to sustain engagement. Frequency is not fixed; coaches recommend more sessions for higher acuity or greater skill deficits.

Methods in brief

• Sample: 47 adolescents and young adults (M age = 17.8; 78.7% female; 85.1% white) in OP (87.2%) or IOP (12.8%) care; predominantly privately insured; major depressive disorder was the most common primary diagnosis.
• Design: Single-site, within-subjects; assessments at intake, 6 months, and 1 year; linear mixed-effects models (which retain participants with missing data) plus regression to compare predicted change under standard vs. Enhanced DBT-A.
• Measures: C-SSRS screener (suicidal ideation severity/risk, scored 0–10), Reasons for Living Inventory for Adolescents (RFL-A), and PHQ-9/PHQ-A (depressive symptoms).
• Dose: Over year one, participants completed on average ~54 individual sessions, ~31 group sessions, ~6 family sessions, ~1 phone-coaching contact, and ~25 ISCS (range 1–60).
• Registration/oversight: gov NCT07256496; approved by the Duke Health IRB.

Key findings

Across the first year, participants improved significantly on all three measures, with the largest gains between intake and 6 months:

The dose signal was specific to suicide risk. In a mixed model controlling for the standard DBT-A modalities (individual, group, and phone sessions), the number of ISCS predicted lower C-SSRS scores (F[1,93] = 7.41, p = .008), with a significant time-by-ISCS interaction (p = .031). Roughly 15 ISCS over six months (about one per week) corresponded to an estimated 2-point C-SSRS decrease. Regression estimates illustrate the contrast: predicted intake-to-6-month C-SSRS change was −2.18 under Enhanced DBT-A versus +0.67 under standard DBT-A, and cumulative one-year change was −2.90 versus −1.08. ISCS dose did not significantly predict RFL-A or PHQ change, so the dose-response pattern is best understood as specific to suicidal ideation in this sample.

Limitations — read before you generalize

1. No control group. We cannot attribute improvement causally to ISCS, nor claim superiority over standard DBT-A.
2. Increased treatment contact is a plausible confound. More ISCS means a higher overall dose of clinical contact, which could account for some of the effect independent of session content.
3. Limited generalizability. A single, resource-rich, predominantly white, female, privately insured sample treated by highly trained DBT clinicians may not represent community, school-based, residential, or lower-resourced settings, or boys, gender-diverse youth, and minoritized populations.
4. NSSI was not formally assessed with a validated measure; findings pertain specifically to suicidality.
5. Missing data and the paper-to-EHR transition affected some service records; missingness was more common among older participants.

Clinical implications

We read these results as promising but preliminary. The most defensible takeaway is that ISCS may function as a dose-increasing adjunct within programs already delivering adherent DBT-A and equipped to add skill-focused sessions — not as a stand-alone, independently evidence-based intervention, and not as a replacement for or improvement on standard DBT-A. Several mechanisms here travel well across settings: structured targeting of chain-analysis-identified skill deficits, repeated individualized behavioral rehearsal, and engagement-oriented use of commitment, validation, and stylistic strategies — none of which depend on unusual resources.

For lower-resourced or higher-volume programs, the model could be adapted with briefer (20–30 minute) coaching, event-triggered scheduling (e.g., post-crisis or post-discharge) rather than weekly cadence, prioritization for the highest-risk youth, culturally responsive chain-analysis prompts, multilingual and telehealth delivery, and supervised trainees working from structured session templates to maintain fidelity.

Future directions

Controlled designs are the clear next step, along with assessment of additional outcomes (emotion-regulation skills, broader quality-of-life and BPD-related symptoms), validated NSSI measurement, qualitative input from youth and caregivers, and parallel caregiver-report measures. We share these findings as a contribution to the DBT-A literature and an invitation for replication.