← Blog

Dialectical Behavior Therapy for Children (DBT-C), Francheska Perepletchikova, and the Yale Child Study Center: parent coaching session narration, child behavioral chain analysis narration, parallel child-parent skills group narration, and developmental assent documentation narration outside psychotherapist-patient privilege

September 18, 2026 · TherapyDraft · 5,900 words

Francheska Perepletchikova, the Yale Child Study Center, and the institutional landscape of DBT-C

The Yale Child Study Center is a division of Yale School of Medicine, founded in 1911, one of the oldest and most prominent academic child mental health research centers in the United States. It is a private academic institution. Yale University is a private Ivy League university; Yale School of Medicine is its private medical school. The Yale Child Study Center has no authority conferred by any federal, state, or local statute to conduct health oversight activities within the meaning of HIPAA § 164.512(d). That exception applies specifically to government agencies conducting health oversight activities authorized by US law: programs administering Medicare, Medicaid, and other government health benefits programs; agencies that license or certify health care facilities or providers; agencies conducting government health oversight investigations; and agencies administering government health programs for specific populations. The Yale Child Study Center is none of these. It is an academic research and training institution affiliated with a private university. A cloud AI scribe vendor that discloses session records to the Yale Child Study Center in response to any research request, training program inquiry, or institutional oversight process is not disclosing under a HIPAA-permitted pathway. Such a disclosure requires patient authorization or a court order.

Weill Cornell Medicine is the medical college of Cornell University, a private research university. Weill Cornell Medicine conducts clinical research and operates training programs in partnership with NewYork-Presbyterian Hospital — a private nonprofit academic medical center. Neither Weill Cornell Medicine nor NewYork-Presbyterian Hospital is a government agency or a health oversight agency within the meaning of § 164.512(d). The same analysis applies: these are private academic and clinical institutions with no government oversight authority over practitioners who use cloud AI scribes in their DBT-C clinical work.

Francheska Perepletchikova, PhD, conducted her foundational DBT-C research at the Yale Child Study Center, where she developed the theoretical basis, clinical protocol, and initial empirical testing for applying DBT to pre-adolescent children. Her published research includes pilot studies, open trials, and randomized controlled trial data testing DBT-C for children ages 7–12 with severe emotional dysregulation, disruptive mood dysregulation disorder (DMDD), and suicidal ideation and self-harm. DMDD was introduced in DSM-5 in 2013 to capture the population of children with chronic, severe, and persistent emotional and behavioral dysregulation who had previously received inappropriate bipolar disorder diagnoses; DBT-C was developed in direct response to the clinical need for an evidence-based treatment for this population. The treatment manual was published as Dialectical Behavior Therapy for Pre-Adolescent Children: Helping Clients Aged 7–12 and Their Parents by Guilford Press in 2020, establishing DBT-C as a manualized, evidence-based intervention with a research foundation.

DBT-C is explicitly distinct from two prior DBT adaptations covered in this series. Standard adult DBT — developed by Marsha Linehan at the University of Washington in the late 1980s for adult clients with borderline personality disorder — is covered in the DBT-LBC credential post in this series, which addresses the vendor archive record types specific to adult individual therapy, adult diary card review, adult phone coaching calls in active crisis, and adult skills training group. DBT for Adolescents (DBT-A), developed by Jill Rathus and Alec Miller at Montefiore Medical Center / Albert Einstein College of Medicine in the Bronx, adapts the adult DBT protocol for adolescents ages 13–18 and their families — a population developmentally and legally distinct from Perepletchikova's pre-adolescent target group. DBT-C for children ages 7–12 is not a downward extension of DBT-A but a separate theoretical and clinical adaptation for a population at a qualitatively different developmental stage: concrete operational cognitive development, pre-pubertal, with limited capacity for abstract self-reflection, limited verbal emotional vocabulary, and dependent on adult caretakers for behavioral regulation support in a way that adolescent patients typically are not. The adaptations required for this developmental stage — shorter chains, visual aids, emotion thermometers, play-based skill practice, parallel parent training as a structural treatment requirement — make DBT-C clinically and documentarily distinct from every prior DBT post in this series.

The credentialing landscape for DBT-C reflects the absence of a separate infrastructure. The DBT-Linehan Board of Certification, LLC (DBT-LBC), covered in detail in the DBT credential post, administers the DBT-Certified Clinician credential for practitioners applying DBT, including those working with children using DBT-C. There is no DBT-C for Children Institute issuing separate credentials, no DBT-C for Children board certification program, and no DBT-C for Children practitioner registry. The note here about the DBT-LBC credential abbreviation: DBT-LBC uses the abbreviation "DBT-C" for the DBT-Certified Clinician credential it awards — meaning that a practitioner who holds a "DBT-C" certification from the DBT-Linehan Board of Certification holds a general DBT certification, not a specific certification in dialectical behavior therapy for children. The treatment protocol for children also uses "DBT-C" as its abbreviation. These are different uses of the same abbreviation; the treatment protocol and the certification credential are distinct things that share a letter sequence in the current literature.

Training in DBT-C for children is obtained primarily through Behavioral Tech intensive training programs adapted for the child population, specialized DBT-C workshops developed and taught by Perepletchikova and her collaborators, and supervised clinical training at academic medical centers with active DBT-C research programs. The unlicensed practitioner population applying DBT-C-derived techniques includes play therapists who have completed DBT-C training but have not obtained qualifying state licensure in a privilege-carrying mental health profession; school-based mental health practitioners whose state licensure (school counselor or school psychologist) may carry different — and in some states more limited — privilege protections than the private-practice clinical license; paraprofessional staff in residential treatment programs for children with severe emotional dysregulation applying DBT skills coaching without a qualifying clinical license; and practitioners completing supervised clinical hours toward first licensure who are working in DBT-C programs. For all of these practitioners, the cloud AI scribe vendor archive of their DBT-C sessions carries no psychotherapist-patient privilege protection — privilege requires a qualifying state license in a privilege-carrying profession, not a DBT-C training certificate.

The four DBT-C vendor archive record types absent from all prior 209 posts

Parent coaching session narration

DBT-C's parent coaching component is not merely supportive or adjunctive — it is a structural treatment requirement. Standard adult DBT involves phone coaching, a consultation team, and skills training groups, but these components all directly involve the identified patient. DBT-C adds a qualitatively different component: dedicated parent coaching sessions in which the DBT-C clinician meets with the named parent alone, without the child patient present, and works directly with the parent on developing and practicing DBT-informed parenting skills. The parent is in the room. The child is not. The parent is not the HIPAA patient. The cloud AI scribe that is running in that room — if the clinician uses a cloud AI scribe — is documenting a session whose primary active participant is a person who is not the HIPAA patient.

What the parent says in parent coaching sessions is qualitatively different from what a patient says in their own therapy. The parent in parent coaching discloses: their emotional reactions to their child's behavioral dysregulation in specific named situations; their own mental health history and vulnerabilities that interfere with their capacity to use DBT parenting skills in heated moments; their relationship history with the child and how it affects the current dynamic; their co-parenting conflicts and the ways in which their relationship with the co-parent affects the child's treatment and their own parenting; their responses to their child's suicidal statements and self-harm — the fear, the helplessness, the anger — that are precisely what the coaching aims to help them regulate; their perception of the DBT-C treatment's progress and their doubts about its efficacy; and their own behavioral failures — the moments they escalated rather than de-escalated, the punishments they administered that they knew were not skillful, the times they said things they regretted. These are disclosures made in the context of the child's treatment, about the parent's own behavior and emotional life, by a person who is not the HIPAA patient.

The vendor archive of the parent coaching session contains all of this. It is maintained as a business record by the cloud AI scribe vendor. The vendor did not obtain a separate HIPAA authorization from the parent — the parent is not the HIPAA patient and does not independently control the vendor's data practices through a HIPAA authorization. The parent authorized their child's treatment and in most cases signed a HIPAA authorization that covered the child's treatment records. That authorization did not necessarily disclose to the parent that a cloud AI scribe was being used in parent coaching sessions, that the vendor was maintaining an independent archive of those sessions, or that the vendor's archive could be compelled as a third-party business record in proceedings to which the parent is a party.

Child behavioral chain analysis narration (pre-adolescent adaptation)

The DBT chain analysis is covered in two prior posts in this series: the DBT chain analysis post covering the standard adult chain analysis, and the adult DBT credential post addressing diary card narration, phone coaching call narration, and skills group narration. Both posts address chain analysis and related documentation in the adult clinical context: adult patients, adult behavioral targets (suicidal behavior, self-harm, serious substance misuse, significant relationship crises), and adult cognitive capacity for the detailed internal-state narration that standard DBT chain analysis requires.

The child behavioral chain analysis in DBT-C is adapted for the pre-adolescent child's concrete operational cognitive stage. The adaptation is not merely cosmetic — simplified language swapped into the same structural framework. The adaptation is structural. Concrete operational children (Piaget's stage, approximately ages 7–11) can reason logically about concrete objects and events but have limited capacity for abstract self-reflection about their internal states. They can typically identify that they felt "mad" or "scared" before a behavioral episode, and they can narrate the concrete events that preceded and followed it, but they cannot produce the nuanced multi-level cognitive and emotional internal-state narration that adult DBT chain analysis elicits. The child behavioral chain analysis accordingly works with concrete behavioral descriptions, shorter chains with fewer links, emotion thermometer ratings for distress intensity rather than verbal descriptions of internal experience, and visual aids — drawings, charts, behavioral thermometers — that help the child identify and sequence the concrete events in their behavioral episode.

What the child behavioral chain analysis generates in its vendor archive narration is a document organized around specific concrete events in the child's daily life. The prompting event in the child's chain is typically an adult behavior: the parent said something specific, the teacher gave a specific instruction, the sibling did a specific thing, the peer excluded the child in a specific way. The vulnerability factors may include a preceding adult interaction: the parent's own emotional dysregulation at breakfast, the argument between the parents that the child overheard. The links in the chain name what the child did in response to each event, what each adult in the chain did in response to the child, and the escalating sequence of reciprocal behaviors between the child and the named adult. The consequences name how the episode resolved: what the named parent did at the end, what the named teacher's disciplinary response was, how the named sibling or peer reacted to the child's behavior.

A year of weekly individual DBT-C sessions, each including a chain analysis of the week's most difficult behavioral episode, generates a vendor archive record spanning 40 to 50 chain analysis narrations, each naming specific adults and documenting their conduct in relation to the child's behavioral dysregulation at each clinical date. This is a contemporaneous, clinically structured, session-by-session account of the child's behavioral episodes in specific named relational contexts — naming the specific adults who were present, what they did, and how the child responded — across the full course of treatment.

Parallel child-parent skills group narration

DBT-C operates two simultaneous groups: a child skills training group in which multiple child patients practice DBT-adapted emotion regulation, distress tolerance, mindfulness, and interpersonal effectiveness skills together; and a concurrent parent skills training group in which the parents of those same children practice the parallel parent-facing DBT skills curriculum. Both groups run at the same time, in adjacent spaces, for the same session duration. A cloud AI scribe used in either or both group sessions generates two simultaneously produced multi-party vendor archive records, linked by the named family relationships between the child group participants and the parent group participants.

The child skills group narration documents multiple named children — each a HIPAA patient — engaged in structured skill practice and group discussion. In a standard DBT-C child skills group, children practice the emotion thermometer, identify the feelings and situations that push their thermometer toward the top, discuss the skills they tried between sessions, and practice new distress tolerance or interpersonal effectiveness skills. Each child's participation — their identified difficult feelings, their reported between-session situations and skill attempts, their interactions with other group members — is documented in the group narration. The group therapy privilege analysis for child patients may differ from the individual therapy analysis: many state privilege statutes limit or eliminate privilege for disclosures made in the presence of persons outside the privileged relationship, and the question of whether group therapy disclosures among multiple child patients are covered by psychotherapist-patient privilege is unsettled in most jurisdictions.

The parent skills group narration simultaneously documents the named parents of those same named children. The parent group is not a therapy group for the parents as patients; it is a psychoeducation and skills training group for them as caregivers of HIPAA patients. But the boundary between skills training and therapeutic disclosure is not sharp in practice. Parents in DBT-C parent groups disclose the most difficult parenting situations of the week, including situations they handled poorly, emotional responses they were ashamed of, and family dynamics that are causing distress — disclosures that would be clinically relevant in a wide range of legal proceedings involving those parents. The parent group narration documents those disclosures by multiple named parents simultaneously, creating a multi-party record of named parents' own behavioral and emotional disclosures about their parenting of named children who are themselves documented in the concurrent child group narration.

Developmental assent documentation narration

Pre-adolescent children cannot provide legally binding consent to medical treatment. Their parents or legal guardians serve as their HIPAA personal representatives and provide the consent that authorizes treatment and controls the child's HIPAA rights. A child who is seven years old at the start of DBT-C has no independent authority to authorize or refuse access to their treatment records. Their parent holds that authority entirely. A child who is twelve years old at the end of a multi-year DBT-C course has developed substantially, and in some states may begin to have limited HIPAA rights of their own in defined clinical circumstances, but in most states and for most purposes the parent remains the HIPAA personal representative for children under 18.

DBT-C clinical practice requires documented assent from the child patient at intake and periodically throughout treatment. Assent is the child's expressed agreement to participate — not legally binding consent, but clinically and ethically required as part of a treatment approach that is organized around the child's active participation and skill development. The clinician documents the child's assent: what the child said they understood about the treatment, what they said their goals were, whether they agreed to participate, and what they expressed about their own difficulties and what they wanted help with. These assent documentation narrations capture the child's own account of their struggles — in the child's own words, at the child's developmental level, at the specific clinical date when the assent was documented.

Because DBT-C treatment for children with DMDD or severe emotional dysregulation may span a year or more, the developmental assent documentation narrations across the treatment course constitute a longitudinal record of the child's self-account at successive developmental stages. A child who assents at age seven describes their difficulties in the language available to a seven-year-old in concrete operational cognitive development. A child who provides periodic re-assent at ages eight, nine, and ten describes the same difficulties with progressively more developed vocabulary, self-reflection, and insight. This is a developmental record — tracking not just the child's emotional functioning but their developing capacity to describe and understand that functioning.

The cloud AI scribe vendor's archive of the developmental assent documentation narrations is maintained as a business record across the years of treatment. The child, at the time the records were created, had no independent HIPAA authority to restrict access to those records. Their parent held HIPAA authority and authorized the treatment, the HIPAA practices, and — to the extent the HIPAA authorization covered the scribe vendor — the vendor's data retention practices. If the vendor's records are sought in subsequent proceedings — years after treatment concludes, when the child has reached adolescence or adulthood — the developmental assent documentation narrations in the vendor's archive contain the child's own words at each clinical stage, documented while the parent held HIPAA authority over those records, in a business record maintained outside the child's later control.

Five adversarial proceedings reaching the DBT-C vendor archive

Custody and family court proceedings

In contested custody proceedings, each parent's emotional regulation, parenting capacity, co-parenting ability, and behavioral history during the period of contested custody are material to the court's best-interest-of-the-child determination. DBT-C parent coaching sessions generate the most detailed contemporaneous documentation of these matters available outside the therapeutic context: the named parent's own disclosures, in their own words, at each clinical date, about their emotional responses to parenting, their co-parenting conflicts, their behavioral failures, and their capacity to learn and apply parenting skills. This is documentation the parent generated about themselves, in the context of their child's treatment, without necessarily anticipating that it would become adversarially relevant.

Psychotherapist-patient privilege protects the child's psychotherapy records. The child is the patient. The child's individual DBT-C sessions — including the child behavioral chain analysis narrations — carry psychotherapist-patient privilege to the extent that privilege protects the child's PHI. But the parent coaching session narrations present a different privilege analysis. The parent is not the patient. The parent's disclosures in parent coaching sessions are not the child's confidential communications to a therapist; they are the parent's communications to a clinician in the context of the parent's participation in their child's treatment. Many state privilege statutes protect confidential communications made by a patient to a therapist in confidence; they were not written to address communications made by a non-patient third party attending the patient's treatment as a collateral participant. The extent to which the child's privilege extends to protect the parent's collateral disclosures in parent coaching sessions is a genuinely contested legal question — and one that the therapist's own records and the cloud AI scribe vendor's archive both raise, but which the vendor's archive raises in the additional context of a separately subpoenable third-party business record.

The cloud AI scribe vendor is not the therapist. The vendor is not the covered entity. The vendor's independently maintained archive of parent coaching sessions is a business record subject to civil subpoena as a third party without the need to directly overcome the psychotherapist's privilege — because the vendor's records are not the therapist's privileged communications, they are the vendor's business records of processing activity that the vendor performed. The custody attorney who obtains a civil subpoena to the cloud AI scribe vendor can obtain the parent coaching session narrations documenting both parents' disclosures without necessarily establishing a legal basis to overcome the therapist's privilege, because the subpoena targets the vendor's business records rather than the therapist's clinical file.

Child protective services and mandated reporting proceedings

DBT-C practitioners are mandated reporters. When information disclosed in treatment — including in parent coaching sessions — meets the threshold for a mandated report, the practitioner reports to the appropriate CPS agency. But the mandated reporting process and the CPS investigation that follows are separate from the question of what records the CPS agency can access through the investigation.

CPS investigations operate under HIPAA § 164.512(b), which permits disclosure to public health authorities authorized by law to receive such information, and under § 164.512(c), which permits disclosure about victims of abuse, neglect, or domestic violence to government authorities authorized to receive reports. Under these exceptions, CPS can access protected health information in the course of an authorized investigation without a patient authorization or court order. The cloud AI scribe vendor's independently maintained archive is subject to the same HIPAA framework as the covered entity's own records when the vendor has entered into a BAA — meaning the vendor's archive is accessible to CPS under §164.512(b) and §164.512(c) in the same circumstances as the therapist's own records.

The adversarial significance of the parent coaching session narrations in CPS proceedings is direct. Parent coaching sessions are the clinical setting in which parents are most likely to disclose the parenting behaviors that are closest to the CPS threshold: the disciplinary measures they used that they feel uncertain about, the moments they lost control of their emotional responses to their child's behavior, the incidents in which they did something they immediately regretted. These disclosures are made in the therapeutic context of learning to parent more skillfully — the parent discloses the failure in order to get help doing better. But the disclosure is documented in the vendor's archive as a business record, and that record is accessible to CPS under the HIPAA framework for abuse and neglect investigations.

Dependency court and termination of parental rights proceedings

Dependency court proceedings — including adjudications of child abuse or neglect and termination of parental rights (TPR) proceedings — require evidence about the named parent's parenting fitness, their response to services designed to remediate identified parenting deficiencies, and their capacity for change. DBT-C parent coaching sessions, conducted specifically to help the named parent develop parenting skills, are directly relevant to each of these questions.

A longitudinal set of parent coaching session narrations from a year of treatment documents: whether the named parent engaged with the coaching, attended sessions consistently, and attempted to apply the skills; how the parent's emotional regulation in parenting situations changed or did not change across the treatment course; whether the parent's disclosures about the child's behavioral episodes became more skill-informed or continued to reflect the same patterns of emotional dysregulation; and whether the parent's account of their co-parenting conflicts, family dynamics, and parenting responses evolved in the direction of greater skill and self-awareness or remained static.

This is exactly the evidence that dependency courts evaluate in assessing whether a parent has made meaningful efforts toward reunification, whether services have been adequate to remediate identified deficiencies, and whether the threshold for termination of parental rights has been met. The parent who participated in DBT-C parent coaching as a component of a child welfare service plan may have believed that their participation in the child's treatment demonstrated their commitment to parenting improvement. The cloud AI scribe vendor's archive of those parent coaching sessions provides a contemporaneous record of the content and quality of that participation — at each clinical date, in the parent's own disclosed words — that is independently subpoenable as a third-party business record in the dependency proceedings.

The child welfare system's interest in the cloud AI scribe vendor's archive of children's therapy records is covered in an earlier post in this series. The DBT-C parent coaching session narration is structurally distinct from that coverage because it documents the parent's participation in the child's treatment as a collateral — generating a record of parental fitness evidence that is not protected by the child's psychotherapist-patient privilege and that sits independently in the vendor's archive as a third-party business record.

School IEP and IDEA proceedings

Children with DMDD or severe emotional dysregulation sufficient to require DBT-C typically experience significant functional impairment in school settings. Aggressive outbursts in class, inability to regulate emotional responses to academic frustration, interpersonal conflicts with teachers and peers that result in school discipline, and school refusal are common in the DBT-C clinical population. Many of these children are referred for special education evaluations, pursue IEP or 504 plan accommodations, or are the subject of IDEA eligibility determinations.

The child behavioral chain analysis narrations in the DBT-C vendor archive are directly relevant to school proceedings. A chain analysis conducted in the week after a significant school incident — a physical altercation in class, an aggressive episode triggered by a teacher's correction, a school refusal meltdown — documents the specific named teacher or school staff member whose conduct appeared as a prompting event or link in the child's behavioral chain. The chain analysis documents what the named teacher said or did, how the child responded, what the named teacher did next, and how the episode escalated or resolved. This is contemporaneous clinical documentation, generated within one week of the incident, of the child's account of the school event and the conduct of named school personnel.

IEP team meetings review the child's functional performance across school settings, assess the impact of disability on the child's educational performance, and design services and accommodations based on the child's documented needs. The child behavioral chain analysis narrations in the vendor archive may be sought by the school district's attorney as contemporaneous documentation of the child's behavioral functioning in specific school contexts and of the named school personnel's conduct in relation to those behaviors. Records shared with the school district in the IEP process become FERPA-covered copies in the school's possession. The original records in the therapist's file and in the cloud AI scribe vendor's archive remain HIPAA-governed. The vendor's archive is accessible to the school district only through a HIPAA authorization or civil compulsory process — not through the FERPA educational records request process that governs the school's own copies. The cloud AI scribe vendor is a separately subpoenable third party in any civil proceedings arising from the school's IEP process or from any adverse action taken by the school in relation to the child's behavioral incidents.

Insurance medical necessity review and managed care clinical review

Children's behavioral health services are subject to intensive managed care review. DBT-C is a multi-component treatment that requires significant resources: individual child therapy sessions, parent coaching sessions, child skills group, and parent skills group, typically delivered weekly across six months to a year or more. Insurers and MCOs conduct ongoing medical necessity review for such services, requiring documentation of the child's diagnosis, functional impairment, treatment response, and continued medical necessity at regular intervals.

Medical necessity review for a minor patient requires a HIPAA authorization from the parent as the child's HIPAA personal representative. When the parent signs the authorization for medical necessity review, that authorization typically covers the child's treatment records. A HIPAA authorization drafted broadly enough to cover all treatment records includes the cloud AI scribe vendor's independently maintained archive, because the vendor is a business associate of the covered entity and the vendor's records are part of the child's designated record set for HIPAA purposes when the vendor has a proper BAA in place.

The insurer's or MCO's clinical reviewer who receives the vendor's archive for medical necessity review receives a record set that may include parent coaching session narrations documenting the named parent's own mental health status, emotional regulation, and parenting behaviors — disclosures the parent made in the context of their child's treatment that are not strictly necessary for evaluating the child's medical necessity. The parent who signed the medical necessity authorization for their child's records may not have understood that they were authorizing the disclosure of their own disclosures from parent coaching sessions to the insurer's clinical review team. The insurer's clinical reviewer, upon reviewing the vendor archive, is now in possession of the named parent's own health disclosures — their emotional dysregulation history, their parenting failures, their co-parenting conflicts — as part of the child's treatment record authorized for medical necessity purposes.

Insurer and MCO records of medical necessity reviews are not governed by HIPAA in the same way as the treating clinician's records — they are held by the insurer under its own HIPAA obligations, which permit use for payment and health care operations. If the insurer's records of the medical necessity review, including the vendor archive materials reviewed, are subsequently subpoenable in proceedings involving the family — a disability claim, a custody proceeding, a malpractice claim against the child's treating clinician — the chain of disclosure from the original vendor archive through the medical necessity review creates an additional custodian of the sensitive information from the parent coaching sessions.

On-device processing and what it changes for DBT-C

The adversarial significance of the DBT-C vendor archive across all five proceedings above rests on one structural fact: the cloud AI scribe vendor maintains an independently subpoenable business record of sessions that the therapist, the parent, and the child did not intend to place in the permanent possession of a third party. The vendor archive exists because the AI scribe needs to process the audio to generate the note, and because cloud AI scribe vendors retain transcripts, summaries, or both in their systems — under their own data retention policies, accessible under their own HIPAA and legal obligations, and separately subpoenable without overcoming the therapist's psychotherapist-patient privilege.

On-device processing eliminates this third-party archive entirely. When the DBT-C clinician uses a tool that processes audio locally on their own Mac — with audio, transcript, and note text never transmitted to a server — there is no vendor archive. The cloud AI scribe vendor does not come into possession of the parent coaching session audio or transcript. The vendor cannot produce records it does not hold. The custody attorney who subpoenas the vendor receives a response that the vendor holds no records for the relevant practice. The CPS investigator, the dependency court, the school district's attorney, and the insurance carrier's clinical reviewer all face the same result: the vendor has nothing.

What remains accessible through legal process is the therapist's own clinical record — notes, assessments, and documentation maintained in the therapist's own records system. That record is subject to psychotherapist-patient privilege, and the privilege analysis for the parent coaching session narrations applies to the therapist's own record as it always has — through the established privilege framework, not through the collateral route of third-party business record subpoena. The privilege questions about the parent's non-patient disclosures in parent coaching sessions remain legally interesting regardless of whether a cloud scribe is used; what on-device processing removes is the additional independently subpoenable third-party vendor archive that makes those questions practically pressing in a way they would not otherwise be.

For a treatment modality specifically designed for families navigating the most difficult possible circumstances — children with severe emotional dysregulation, parents who are struggling to manage crises that would overwhelm most adults, family systems that are frequently in contact with child welfare systems, schools, and legal proceedings — the elimination of the vendor archive is a structural protection that aligns with the treatment's clinical goals. The parent who discloses in parent coaching that they handled a situation badly and needs help doing better is not disclosing to create a record that can be subpoenaed in subsequent custody proceedings. The child who provides assent and describes their difficulties at each developmental stage is not consenting to the creation of a longitudinal vendor archive that will outlast their childhood. The design of the documentation system should serve the treatment's goals, not undermine them.

Practical implications for DBT-C clinicians

Clinicians delivering DBT-C who use or are considering cloud AI scribes should evaluate several threshold questions before adopting such tools for this population. First, is the cloud AI scribe vendor's business associate agreement clearly negotiated to cover all DBT-C treatment components, including parent coaching sessions and parallel skills groups — and does the BAA address data retention practices for multi-party session records that include both patient PHI and non-patient third-party disclosures? Second, does the HIPAA authorization obtained from the parent as the child's personal representative clearly disclose that a cloud AI scribe vendor will maintain an independently archived record of all treatment components, including parent coaching sessions, and that the vendor's archive may be subject to compulsory process in legal proceedings? Third, has the clinician assessed the likelihood that the family is navigating or may navigate custody proceedings, child welfare involvement, school disputes, or insurance disputes — because the standard DBT-C population has substantially elevated exposure in each of these proceedings?

Fourth, and most structurally: does the clinician's informed consent process for DBT-C treatment accurately represent the documentation and data practices that apply to each treatment component? Many standard therapy informed consent forms were drafted for individual adult therapy; they do not contemplate the specific multi-party, multi-component documentation landscape of DBT-C. An informed consent form that accurately discloses the cloud AI scribe vendor's practices for individual child sessions may not accurately disclose what happens to the parent coaching session audio, the parallel skills group narrations, or the developmental assent documentation — because those components are each structurally different from the individual adult therapy session for which most cloud AI scribe vendor disclosures were drafted.

Clinicians who are delivering DBT-C in contexts where child welfare involvement, custody litigation, school disputes, or intensive insurance review are likely — which describes a substantial share of the clinical population for whom DBT-C is indicated — should evaluate whether the documentation architecture they use serves the clinical and ethical obligations of the treatment or creates collateral records whose adversarial implications they have not disclosed to the families in their care. The technical reality of what cloud AI scribes transmit and retain applies to DBT-C's complex multi-party session structure in ways that the clinician who adopted a cloud scribe for straightforward individual adult therapy may not have anticipated.

Frequently asked questions

Does training in DBT-C create a separate credentialing pathway from DBT-LBC?

No. DBT-C does not have a credentialing body separate from DBT-LBC. There is no DBT-C for Children Institute, no DBT-C for Children board certification, and no DBT-C for Children practitioner registry. Training through Behavioral Tech workshops and specialized DBT-C programs does not create a § 164.512(d) health oversight authority. Yale Child Study Center and Weill Cornell Medicine are private academic institutions with no government health oversight authority over cloud AI scribe vendors. Psychotherapist-patient privilege requires a qualifying state mental health license, not a training completion certificate.

How is the parent coaching session narration different from other multi-party session records in the series?

In every prior post in the series, the active session participant is the HIPAA patient. In parent coaching sessions, the active participant is the named parent — not the child patient — disclosing their own behavioral and emotional experiences as a non-HIPAA-patient collateral in the child's treatment. The privilege analysis for the parent's own disclosures in parent coaching sessions is distinct from the privilege protecting the child's own clinical records. The cloud AI scribe vendor's archive of parent coaching sessions is independently subpoenable without overcoming psychotherapist-patient privilege for the child's records, because the parent's disclosures are not the child's privileged communications.

What makes the child behavioral chain analysis narration distinct from adult DBT chain analysis?

The child behavioral chain analysis in DBT-C is adapted for concrete operational cognitive development: shorter chains, simplified language, visual aids, emotion thermometers, and behavioral targets specific to pre-adolescent children (aggressive outbursts, school refusal, emotional meltdowns). Most critically, the prompting events and links in the child's behavioral chains name specific adults — parents, teachers, siblings — whose conduct appears in the chain at each clinical date. A year of weekly chain analyses generates 40 to 50 narrations, each naming specific named adults in behavioral chain links, across the treatment course.

What is the parallel child-parent skills group narration and why does it create a dual-archive structure?

DBT-C operates a child skills group and a simultaneous parent skills group whose members are directly linked by named family relationship. A cloud AI scribe used in either group creates two simultaneously produced vendor archive records: the child group narration (containing PHI from multiple named child patients) and the parent group narration (containing disclosures from the named parents of those same children). The two narrations together constitute a dual-archive account of the family's behavioral functioning from two perspectives at the same clinical date — independently subpoenable as third-party business records, accessible without overcoming the privilege protecting the child's individual therapy records.

Why does the developmental assent documentation narration have no analogue in the prior 209 posts?

Every prior post covers treatment of patients who have or are approaching independent HIPAA rights. Pre-adolescent children have no independent HIPAA rights — their parent holds HIPAA authority entirely. Developmental assent documentation narrations capture the child's own account of their difficulties in their own words at successive developmental stages across the treatment course — created while the parent held HIPAA authority over those records, maintained in the vendor's archive across years, and potentially relevant in proceedings initiated after the child has reached adulthood. No prior record type in the series is organized around the minor patient's evolving capacity to understand and assent to treatment at developmental stages before they had HIPAA rights of their own.