Eleven codes and one standard
- Since 1 January 2025 an adult in Russia carrying any of eleven ICD-10 personality disorder codes, F60.0 through F60.9 and F61, is treated under one national document: the clinical guideline "Specific personality disorders", entry 397 of the Ministry of Health rubricator, third version, posted on 23 August 2024. The Russian Society of Psychiatrists wrote it, the ministry's Scientific-Practical Council approved it, and the working group of five is drawn entirely from the Serbsky Centre in Moscow. Target readers are named as four groups: psychiatrists, psychotherapists, forensic psychiatric experts and clinical psychologists.
- Sixteen statements in the document carry evidence grades. Thirteen of them are addressed to every patient with a personality disorder, without naming which disorder. Three name borderline personality disorder: amitriptyline at 100–175 mg a day, a course of 15 transcranial magnetic stimulation procedures, and long-term psychotherapy. No statement is addressed separately to the paranoid, schizoid, dissocial, anankastic, anxious or dependent disorder, although each holds its own code in the same list.
- Seven of the sixteen sit at the lowest certainty level the rubricator uses, the one it reserves for expert opinion. On transcranial magnetic stimulation the document writes that randomised trials with a sufficient sample of borderline patients are absent and that this calls for caution, and describing its evidence work as a MEDLINE search 12 years deep plus expert consensus.
- The 2024 version is shorter than the 2021 one it replaced. Graded statements went from 23 to 16 and the treatment section from 18 to 9. The treatment section named 53 medicines in 2021 and names 25 now, with 28 names dropped and none added. Quality criteria went from nine rows, each carrying its own grades, to seven rows answered yes or no. Two rating scales, one from the DSM-5 alternative model and one built on ICD-11, entered as required tests at first presentation, and the 2021 version carried no scale appendix at all.
Russia has run personality disorder care from a single document since 1 January 2025. An adult carrying any of eleven ICD-10 codes, F60.0 through F60.9 and F61, falls under the clinical guideline "Specific personality disorders", entry 397 of the health ministry rubricator, third version, posted on 23 August 2024. The Russian Society of Psychiatrists drafted it, the ministry's Scientific-Practical Council approved it, and the working group of five comes entirely from the Serbsky Centre. It fixes what has to be assessed, which medicines may be given and at what doses, how long psychotherapy runs, from 1 to 36 months in courses of 10 to 40 sessions, and where the patient is treated, outpatient by default and inpatient for no longer than 30 days. Seven quality criteria close it, answered yes or no.
Sixteen statements carry the rubricator's two grades, one for how certain the evidence is, on a scale of 1 to 5, and one for how strongly the statement is put, from A to C. Thirteen of the sixteen speak to every patient with a personality disorder and name no diagnosis at all. Three name the borderline one: amitriptyline at 100–175 mg a day (certainty 2, strength A), a course of 15 transcranial magnetic stimulation procedures (certainty 4, strength C), and long-term psychotherapy (certainty 1, strength A). Nothing is addressed separately to the paranoid, schizoid, dissocial, anankastic, anxious or dependent disorder, though each holds its own code two sections earlier.
The grades show what those sixteen stand on. Seven sit at certainty 5, the level the rubricator gives to expert opinion; strength runs A five times, B twice and C nine times. The text says so itself. Its comment on transcranial stimulation reports that the shortage of studies makes the effect hard to assess or explain. Its section 3.3 on other treatment opens by pointing to the insufficient effectiveness of drug therapy in personality disorders. One line states that the absence of randomised trials with a sufficient sample of borderline patients calls for cautious appraisal. Appendix A2 sets out the evidence work as a MEDLINE search 12 years deep plus expert consensus.
Against the 2021 version, the current one is a subtraction. Graded statements fell from 23 to 16, and in the treatment section from 18 to 9. That section named 53 medicines in 2021 and names 25 now: olanzapine, aripiprazole, lamotrigine, topiramate, quetiapine and risperidone are among the 28 names that went, and nothing came in to replace them. Quality criteria went from nine graded rows to seven answered yes or no. Numbered references run to 91 in the old text and to 81 in the new one.
What did arrive points at ICD-11. Every patient at first presentation is now to be tested on the Self and Interpersonal Functioning Scale, which grades how far functioning is impaired (certainty 3, strength B), and on the Five-Factor Personality Inventory for ICD-11, which grades five maladaptive traits (certainty 5, strength C). Both sit in appendices that the 2021 version did not have. ICD-11 is named five times in the new text and not once in the old. Coding stayed put: section 1.4 lists the ICD-10 rubrics, and section 1.5 calls that classification consistent with the traditional Russian approach.
The service therefore stands where the trials do not. The document puts prevalence at 5–10% for all specific personality disorders together and at 0.5–7.8% for a single type, and for that population Russia has fixed an assessment, a medicine list, a session count, a bed-day ceiling and seven checkable criteria. None of that settles efficacy. Seven of the sixteen statements rest on expert opinion, and the guideline says as much on its own pages.
Thirteen of the sixteen graded statements are addressed to every patient with a personality disorder without naming which one, and seven of the sixteen sit at the certainty level the rubricator reserves for expert opinion.
The source is a national regulatory document, not a study. It reports no patients, no cohort and no outcome: every number here counts what the guideline contains, not what happened to anyone treated under it. There is no DOI, and the document is not indexed in PubMed, which makes it unverified through that route rather than unverifiable; the machine-readable copy sits at the rubricator's own API and the same text downloads as a PDF. The counts of graded statements, medicines and references were made by reading both versions in full, and a different counting rule could shift them by one or two. Both versions mark some medicine names with a symbol that neither text defines anywhere in its pages, so what that symbol means is not stated here and the medicine counts are given as names present and names absent. The document itself carries no data on how many patients in Russia are treated under it, how many centres apply it, how long anyone waits, or what the seven quality criteria have measured since they took effect. Prevalence of 5–10% is carried over unchanged from the 2021 version, with two references behind it, and is not a Russian measurement. The rise from 23 statements to 16 says what the text now contains and nothing about why any single statement was dropped.