Turkish Version of DIAMOND-KID Interview Proves Reliable for Diagnosing Youth Mental Disorders

Diagnosing a psychiatric disorder in a child is one of the most consequential acts in clinical medicine. A label such as generalized anxiety disorder, major depression, or obsessive-compulsive disorder can open the door to evidence-based treatment, but it can also follow a young person through school records, family expectations, and self-image. That is why the instruments used to reach these diagnoses are held to such exacting standards, and why a new study from researchers at Atatürk University in Erzurum, Türkiye, published in BMC Psychiatry, matters well beyond its national borders. The team reports that the Turkish adaptation of the DIAMOND-KID, a semi-structured diagnostic interview covering anxiety, mood, obsessive-compulsive and related neuropsychiatric disorders in children and adolescents, shows good to excellent agreement with independent clinical diagnoses, along with strong reliability across raters and repeated administrations.

The DIAMOND-KID is the child and adolescent version of the Diagnostic Interview for Anxiety, Mood, and OCD and Related Neuropsychiatric Disorders, a family of interviews built directly on the diagnostic criteria of the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, DSM-5. Unlike brief screening questionnaires, which flag the possible presence of a problem, semi-structured diagnostic interviews are designed to establish whether a young person actually meets full diagnostic criteria. Trained interviewers follow a prescribed sequence of probes and follow-up questions, allowing them to distinguish, for example, between the intrusive thoughts and rituals of OCD and the worries of generalized anxiety, or between the low mood of depression and the irritability that can accompany a range of conditions. The DIAMOND-KID covers an unusually broad diagnostic territory for ages 6 to 18, spanning panic disorder, social and separation anxiety, specific phobias, major depression, premenstrual dysphoric disorder, post-traumatic stress disorder, tic disorders, trichotillomania, skin-picking and hoarding disorders, body dysmorphic disorder, avoidant and restrictive food intake disorder, and related conditions.

Despite that breadth, the instrument’s developers and users have faced a persistent gap: relatively little evidence about how well the interview performs outside the cultural and linguistic settings in which it was first validated. Diagnostic interviews are not immune to translation problems. Idioms of emotional distress differ across languages, and questions that sound natural in one culture may feel alien or misleading in another. A Turkish child describing somatic complaints, for instance, may frame anxiety in bodily terms that a literal translation would fail to capture. Adapting an interview therefore involves far more than word-for-word translation; it requires demonstrating that the adapted version still identifies the same disorders with the same precision. The new study set out to supply exactly that evidence for Turkish-speaking clinical populations.

The research team enrolled 195 children and adolescents aged 6 to 18 who were receiving psychiatric care, together with their parents or caregivers. Each young participant was assessed with the Turkish DIAMOND-KID and, separately, with an independent clinical interview conducted by a clinician, providing a benchmark against which the interview’s diagnostic judgments could be compared. Within this overall sample, the investigators carved out specialized subsamples for reliability testing: 22 participants took part in an interrater reliability analysis, in which two interviewers assessed the same child and their diagnoses were compared, and 51 participants returned for a test-retest analysis, in which the interview was administered a second time to see whether the same diagnoses emerged. Families also completed a battery of validated self-report questionnaires targeting the symptom domains the interview covers, including measures of depression, multidimensional anxiety, obsessive-compulsive symptoms, eating attitudes, hair-pulling, skin-picking, hoarding, and attention-deficit/hyperactivity symptoms.

The headline result concerns diagnostic validity, quantified with Cohen’s kappa, the standard statistic for agreement between two classification systems after correcting for chance agreement. Kappa values are conventionally interpreted as moderate around 0.41 to 0.60, good around 0.61 to 0.80, and excellent above 0.80. Across the diagnostic categories with sufficient numbers of cases to permit analysis, agreement between the DIAMOND-KID and the independent clinical interview ranged from 0.61 to 0.93, meaning that every comparison reached at least the good range and the strongest reached near-perfect concordance. In practical terms, when the interview said a child met criteria for a disorder, the treating clinician’s independent assessment very often said the same thing, and vice versa. For a tool intended to standardize diagnosis across clinics and studies, that level of concordance is the core evidence that it measures what it claims to measure.

Reliability results were equally encouraging. Interrater reliability, which asks whether two trained interviewers reach the same diagnostic conclusions from the same child, ranged from fair to excellent across the disorders examined, indicating that the interview’s structured probes constrain subjective judgment enough to keep different clinicians largely on the same page. Test-retest reliability, which asks whether the interview yields stable results over repeated administrations, ranged from 0.70 to 1.00, a good-to-excellent band in which the upper bound represents perfect stability. Stability matters because diagnostic interviews are often used at baseline and then again after treatment; if the instrument itself wobbles between administrations, clinicians cannot tell whether a change in symptoms reflects real improvement or measurement noise.

The convergent evidence came from the questionnaire battery. If the interview correctly assigns a diagnosis, children who receive that diagnosis should score higher on questionnaires measuring the corresponding symptoms than children who do not. That is precisely what the researchers found: effect sizes from comparisons of self-report measures between children with and without current diagnoses were generally large. Large effect sizes indicate that the differences were not statistical subtleties visible only in aggregate data but substantial separations between diagnostic groups. This triangulation, in which a clinician-administered interview, an independent clinical judgment, and child and parent questionnaires all point in the same direction, is the strongest practical demonstration of validity available in a clinical sample of this kind.

The study’s significance extends into research infrastructure. Cross-cultural validation of diagnostic instruments is unglamorous work, but it is the foundation on which multinational studies of child psychopathology are built. When a Turkish research team can now use the DIAMOND-KID and know that its diagnoses correspond to those made with the same instrument in other languages, pooled analyses become possible: prevalence estimates can be compared across countries, treatment trials can recruit comparable samples, and genetic or neuroimaging studies can combine cohorts without worrying that diagnostic criteria were applied differently. The interview’s DSM-5 grounding is central here, because DSM-5 considerably refined the obsessive-compulsive and related disorders category, grouping OCD with hoarding, body dysmorphic disorder, trichotillomania, and excoriation disorder, and the DIAMOND-KID was designed to capture that contemporary structure rather than the older DSM-IV framework still used by some legacy interviews such as the Kiddie Schedule for Affective Disorders and Schizophrenia or the Mini International Neuropsychiatric Interview for children.

The authors are careful about the boundaries of their findings. Agreement statistics were computed only for diagnostic categories with sufficient case numbers, so rare diagnoses in the sample could not be independently evaluated, a limitation inherent to any single-clinic study of 195 participants. The sample was clinical, meaning the results speak to the interview’s performance in settings where disorders are already suspected, not necessarily to community screening. The published version is also an early-release, citable record subject to final editorial processing. Still, the pattern of results, good-to-excellent diagnostic agreement, strong test-retest stability, and large convergent effect sizes, gives clinicians and researchers in Türkiye a validated, DSM-5-aligned instrument for one of the widest diagnostic spans available in child and adolescent assessment.

For the wider field, the study is a reminder that the quiet machinery of psychiatric science, the interviews, translations, and kappa coefficients that rarely make headlines, is what allows the dramatic findings to mean anything. Every large trial of a new treatment for pediatric anxiety or depression begins with a diagnostic decision, and the quality of that decision depends on tools like the DIAMOND-KID and the painstaking work of teams who test them in new languages. With this Turkish adaptation now supported by solid psychometric evidence, clinicians in Erzurum and beyond can make those decisions with greater confidence, and the growing international family of validated DIAMOND-KID studies gains another link, extending reliable psychiatric diagnosis to millions of Turkish-speaking children and adolescents aged 6 to 18.

Subject of Research: Validation of the Turkish adaptation of the DIAMOND-KID semi-structured diagnostic interview for child and adolescent psychiatric disorders

Article Title: Validity and reliability of the Turkish adaptation of the diagnostic interview for anxiety, mood, and OCD and related neuropsychiatric disorders: child and adolescent version (DIAMOND-KID)

Article References: Ürük, T., Akıncı, M. A., Kaya, B., Kayıkçı, S., Yıldırım Demirdöğen, E., & Bozkurt, A. (2026). Validity and reliability of the Turkish adaptation of the diagnostic interview for anxiety, mood, and OCD and related neuropsychiatric disorders: child and adolescent version (DIAMOND-KID). BMC Psychiatry. https://doi.org/10.1186/s12888-026-08566-0

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08566-0

Keywords: DIAMOND-KID, diagnostic interview, child and adolescent psychiatry, DSM-5, anxiety disorders, mood disorders, OCD and related disorders, psychometric validation, interrater reliability, test-retest reliability, Cohen's kappa, Turkish adaptation