Uzbekistan Digital Health Platform - Integrations
0.10.0 - draft
Uzbekistan Digital Health Platform - Integrations
0.10.0 - draft
This page is part of the Uzbekistan Digital Health Platform - Integrations (v0.10.0: Releases Draft) based on FHIR (HL7® FHIR® Standard) v5.0.0. This is the current published version. For a full list of available versions, see the Directory of published versions
(No changes yet)
The Cancer Registry now has a FHIR model, shaped after mCODE where the registry's data allows. Cancer Condition carries the diagnosis, registry identifier, laterality and circumstances of detection, Cancer Episode Of Care groups the treatment course and Cancer Encounter a visit within it. Six Observation profiles record tumor morphology, behavior, histologic grade, metastases, TNM category and TNM stage group.
Cancer coding uses ICD-10, ICD-O-3, SNOMED CT and LOINC where they exist, with local code systems for registry concepts such as detection condition, confirmation method and treatment intent. Three ConceptMaps translate the registry's own dictionary identifiers for status, ICD-10 and ICD-O-3 topography into the codes DHP expects.
The hepatitis registry has a model too: Hepatitis Patient, Condition, Episode Of Care, Encounter, laboratory and ultrasound Observation profiles, and a Questionnaire and response for the treatment questionnaire, in place of the unprofiled examples 0.9.0 shipped. Records carry a registry identifier under https://dhp.uz/fhir/core/sid/reg/uz/hepatitis. Hepatitis registry mapping maps every field.
The operation code on a form 066 hospital discharge summary is now an ICHI code. Form 066 Surgical Procedure profiles the Procedure the surgical procedures section references, binding Procedure.code to ICHI Codes Accepted by DMED. Previously the section accepted any Procedure.
ICHI Codes Accepted by DMED enumerates the 6,853 codes DMED carries: 6,176 of the 9,428 stem codes WHO currently publishes, plus 677 WHO has since withdrawn. The withdrawn ones are inactive in the code system and admitted here so records already in DMED stay valid, but should not be used going forward. A procedure the platform codes itself comes from ICHI Codes in UZ Core.
The IHD pre-test probability risk categories are now published here. The IHD pretest questionnaire already answered its risk category item from them, but under a UZ Core canonical that no UZ Core release contains.
The guide is built against UZ Core 0.10.0, which publishes the ICHI code system form 066 needs and relaxes Procedure.code to an extensible binding.
The diabetes, oncohematology, breast cancer, cervical cancer and fertility questionnaires now carry Questionnaire.code from UZ Core screening-code-cs, as the other four screening questionnaires already did.
A fertility questionnaire response with diagnosed infertility shows two ICD-10 diagnoses, each recorded as its own Condition (first, second) whose evidence references the response.
The diabetes and cardiovascular risk questionnaires now score the answers they actually offer. Diabetes compared age, waist and family history against codes no answer option carries, so those items always scored 0; cardiovascular risk compared tobacco use against a stale local code, so smokers were looked up in the non-smoker half of the WHO/ISH chart. Answers are unchanged.
Artifact versions are all MAJOR.MINOR.PATCH. The diabetes screening questionnaire and the screening intake template declared their versions as 1.0 and 1.1; they now declare 1.0.0 and 1.1.0.
Every questionnaire now names the integration it belongs to in a program use context: the nine screening questionnaires and the hepatitis questionnaire carried none and sat ungrouped on the questionnaires page. Integration Area gains hepatitis, and screening now covers the whole screening service rather than cervical and breast cancer screening alone.
The Screening identifier systems now follow the https://dhp.uz/fhir/core/sid/{namespace}/{country}/{type} pattern. Two had no namespace segment, and the public health case identifier sat in the organization namespace although it identifies cases, not organizations. Records and cases move to the doc namespace, the program type to the general prg program system, which any health program can reuse with its own value. Identifier values are unchanged; senders switch the system:
| Identifier | Old system | New system |
|---|---|---|
| Screening record | https://dhp.uz/fhir/core/sid/uz/screening |
https://dhp.uz/fhir/core/sid/doc/uz/screening |
| Public health case | https://dhp.uz/fhir/core/sid/org/uz/screening |
https://dhp.uz/fhir/core/sid/doc/uz/screening-case |
| Screening program type | https://dhp.uz/fhir/core/sid/uz/screening-program-type |
https://dhp.uz/fhir/core/sid/prg/uz/program |
The Uzbekistan PINFL naming system is removed. PINFL is the national identifier UZ Core already defines as https://dhp.uz/fhir/core/sid/pid/uz/ni; send it under that system instead of https://dhp.uz/fhir/core/sid/uz/pinfl.
The fertility questionnaire's infertility-icd-diagnosis item is now coding bound to ICD-10 and repeats, instead of free-text string. Responses must send one valueCoding per diagnosis instead of valueString.
The IHD risk category codes low, medium and high keep their codes but their system changes from https://terminology.dhp.uz/fhir/core/CodeSystem/ihd-risk-category-cs to https://terminology.dhp.uz/fhir/integrations/CodeSystem/ihd-risk-category-cs.
Nine screening questionnaires used by DMED are now published as definitions: cardiovascular risk, cerebrovascular, diabetes, ischaemic heart disease pretest probability, oncohematology, breast cancer, cervical cancer, helminthiasis and fertility. Each scores its answers with an SDC calculatedExpression into a risk category - breast cancer, cerebrovascular, cardiovascular and diabetes - that drives the recommended follow-up. The cardiovascular questionnaire and its risk categories moved here from UZ Core 0.9.0.
Their answer terminology covers the cerebrovascular form's blood pressure, cholesterol, diabetes, heart condition, family history, smoking, alcohol and stress and physical activity answers, the IHD chest pain type, and the diabetes, tobacco use, cervical cancer and fertility answer value sets, the last three SNOMED CT coded with Uzbek designations.
Resources belonging to one screening programme now carry an identifier from the new screening program type naming system, valued with the programme's SNOMED CT code. DMED and the HPV screening system write it on clinical and workflow resources, not on Patient, Practitioner or PractitionerRole.
Pathology reporting gained the local remainders SNOMED CT does not cover: cervical histology morphology subtypes, cervical and breast cytology material types, and breast surgical procedure types. Breast histologic type grew by eight subtypes.
The screening questionnaires now name the integration they come from, in a program use context coded from integration area: GET [base]/Questionnaire?context-type-value=program$https://terminology.dhp.uz/fhir/integrations/CodeSystem/integration-area-cs|screening lists one service's forms, and the questionnaires page groups its picker the same way.
Local codes have been replaced by SNOMED CT wherever a concept exists, leaving the local code systems with the remainder: four compound colour-and-consistency codes in vaginal discharge type, one in urination characteristic, grade IV in obesity degree and "relatively satisfactory" in general condition; axillary and intramammary node status are now entirely SNOMED CT.
Value sets widened for answers the source forms needed: a SNOMED CT concept each in breast palpation assessment and pregnancy outcome, "not changed" and "pronounced" in retromammary space and subcutaneous fat, and six designations in the body site and diagnostic procedure supplements.
The screening laboratory LOINC supplement now declares version 2.82.0 rather than 2.82.
Citizenship has been dropped from the hepatitis patient example: iso-3166-2-vs in UZ Core declares a required supplement on urn:iso:std:iso:3166 that no FHIR package ships, so the validator fails any instance bound to it. It returns once the core value set is fixed.
The guide is now built against UZ Core 0.9.0. None of its breaking changes reach this guide.
meta and meta.source are now 1..1 and Must Support on Screening Composition, DiagnosticReport, DocumentReference, Observation and ServiceRequest, with an invariant requiring https://dhp.uz/fhir/source/screening or https://dhp.uz/fhir/source/dmed, naming the creating system. Existing instances must add it.
The obstetrician-gynecologist code in screening roles has been shortened from his_centre_obstetrics_gynecology_obstetrician_gynecologist to his_centre_obgyn_obstetrician_gynecologist.
Local codes that SNOMED CT covers have been removed and must be restated: scrn-0019-00003, scrn-0019-00005, scrn-0019-00006 and scrn-0019-00008 (vaginal discharge), scrn-0016-00001, scrn-0016-00002 and scrn-0016-00003 (urination), and scrn-0048-00001 and scrn-0048-00002 (axillary node status). Obesity degree also swapped 162864005 and 83911000119104 for 443371000124107, 443381000124105 and 819948005.
The screening, tuberculosis, narcology, psychiatry and sick leave pages now open with a resource model diagram, in all three languages.
The Forms menu entry and page title have been renamed to Questionnaires.
The narcology registry is now covered. Narcology EpisodeOfCare is the patient's registration and carries the dynamic observation group they are placed in, Narcology Observation the decision issued by the medical-consultation commission, and Narcology and Psychiatry Flag the registry information a provider has to know before treating the patient, such as social danger or compulsory treatment. Observation groups are coded with episode-of-care type groups and the flags with registry flag and flag status codes, each of them mapped to SNOMED CT: type groups, registry flags and resource types.
The psychiatry registry follows the same model and shares its terminology and Flag profile. Psychiatry EpisodeOfCare carries the registration, its service type and the observation group, Psychiatry Condition the diagnosis, and Psychiatry Commission Observation the commission's decision notes. Registrations are identified through the narcology and psychiatry registry naming systems.
The screening integration has been improved considerably. It now covers the pathology pathway end to end: Screening Pathology Service Request orders the examination and states in its order details what material is being sent; Screening Breast Biopsy Service Request and Screening Breast Biopsy Procedure cover the biopsy itself, and Screening Specimen the material collected. Results come back as breast histology, cervical histology, breast cytology and immunohistochemistry observations, with Screening Special Treatment Observation recording the treatment the patient received before the specimen was taken.
Breast imaging is structured rather than free text: Screening Mammography Observation and Screening Breast Ultrasound Observation record whole-breast parameters and the characteristics of a focal lesion as components.
A screening cycle can now be published as a document. Screening Composition aggregates the questionnaire answers, the results of every step and the final diagnosis into a downloadable summary, shipped as an example document Bundle. The questionnaires themselves are published as definitions with example responses: breast cancer risk, cervical cancer risk and the women's examination form.
Sixty code systems and a hundred and six value sets supply the terminology all of this needs. Breast ultrasound reporting has a value set per reported parameter, from ACR breast composition and fibroglandular echogenicity through lesion shape, contour, echogenicity, orientation and vascularization to axillary and regional node status. Pathology reporting is covered by breast histologic type, cervical histologic grade, pathological T, N and M stage, resection margin status, lymphovascular invasion, HER2 immunohistochemistry score, p16 result and the Yokohama category. The gynecological examination has its own sets, among them cervix condition, vaginal discharge and vaginal wall prolapse, and every question in the risk questionnaires has a value set for its answers.
Four naming systems record the identifiers Screening exchanges - the questionnaire identifier and questionnaire code assigned to a questionnaire definition, the public health case identifier and the form 025 source identifier - alongside one for the Uzbekistan PINFL.
Screening DiagnosticReport now marks supportingInfo as Must Support, so a pathology report can carry the treatment the patient had before the specimen was taken alongside its results, rather than on the request that ordered them.
The screening body site and diagnostic procedure SNOMED CT supplements gained designations for nipple structure and histologic test, and screening observation codes now also admits the breast cytology result code. The screening body site supplement declared 5.0.0, the FHIR version, as the version of SNOMED CT it supplements; it now declares the SNOMED CT release 2026.1.0, as the other SNOMED CT supplements do.
The guide is now built against UZ Core 0.8.0, which carries breaking changes of its own. Three of them reach this guide: the laboratory request in form 130 now conforms to UZ Core ServiceRequest, whose canonical moved from https://dhp.uz/fhir/core/StructureDefinition/uz-core-servicerequest-laboratory to https://dhp.uz/fhir/core/StructureDefinition/uz-core-servicerequest; code on UZ Core PractitionerRole is now mandatory, so every PractitionerRole exchanged through this guide has to carry a position code; and the coverage type codes have been renumbered, so dtsj-treated-case on a Coverage is now covtp-0001-00001. The UZ Core changelog lists the rest.
The local breast finding code system, its SNOMED CT supplement and the ConceptMap between them have been removed. Breast mammography findings previously admitted the whole of SNOMED CT plus one local code for macrocalcifications; it now enumerates the nine SNOMED CT concepts that are actually reported. Findings recorded under https://terminology.dhp.uz/fhir/integrations/CodeSystem/screening-breast-finding-cs must be restated in SNOMED CT, and a SNOMED CT concept outside those nine no longer validates.
The home page now sets out how versions are identified: artifacts carry the version of the guide, its major and minor version follow the UZ Core release the guide is built against, and translation supplements carry the version of the code system they supplement instead.
First published release of the DHP Integrations IG. Numbering starts at 0.7.0 so that the guide's version matches the UZ Core release it is built against. There is no earlier release to compare against, so this entry describes what the guide contains.
Four Ministry of Health forms are profiled as FHIR Documents: Form 011 hemodialysis session, Form 066 hospital discharge statistical card, Form 066-1 psychiatric and narcological discharge statistical card and Form 130 laboratory test results. Each ships an example document Bundle and a mapping page tying every field of the paper form to its FHIR element: Form 011, Form 066, Form 066-1 and Form 130.
Document categories and identifiers covers how documents are classified, through the document category code system carrying the Ministry of Health form numbers, and how official form numbers and template numbers are recorded alongside per-instance identifiers. Clinical document lifecycle sets out the Composition statuses in use and the transitions between them.
UZ Core Encounter 066 adds an admission origin extension recording where the patient came from before admission. The diagnosis role code system covers the roles these forms distinguish, from referral and admission diagnoses through main, competing, concomitant and background diagnoses to the death-related roles.
Tuberculosis data from DHIS is exchanged as individual resources rather than as documents, mapped field by field on the tuberculosis integration page. DHIS Condition carries the diagnosis, DHIS Episode of Care groups a patient's diagnosis, treatment and follow-up, DHIS Specimen covers material collected for laboratory testing, and DHIS Observation carries test results, specialised into sputum smear microscopy with acid-fast bacilli load grading, chest X-ray findings and HIV co-infection results.
The codes the source system uses are published as code systems - conditions, test codes, components, results, methods, specimen types, specimen features and drug susceptibility - and each is mapped onto a standard so that receivers need not depend on the local codes: diagnoses, results, specimen types and specimen features to SNOMED CT, diagnostic tests and drug susceptibility components to LOINC.
The screening integration page maps the cervical cancer screening and early breast cancer detection system to FHIR. Screening Patient and Screening Practitioner build on the UZ Core profiles; Screening ServiceRequest the referrals for investigation; Screening Observation the results and biometric measurements; Screening DiagnosticReport the final mammography, cytology and colposcopy reports; and Screening DocumentReference the consent forms and protocols.
Findings are coded with BI-RADS, cervical cytology, colposcopy, HPV, transformation zone, mammography finding, breast density, image quality and breast quadrant codes, the last reached through the breast quadrant extension that localises a finding. ConceptMaps translate breast findings, cervical cytology, colposcopy and HPV results to SNOMED CT, and the programme's own practitioner roles to the national roles.
Sick leave documents how a certificate of incapacity for work is represented. Sick Leave CarePlan is the case itself, Sick Leave Observation carries the attributes that do not fit on the CarePlan, and Sick Leave RelatedPerson, with its birthdate extension, covers certificates issued for family care. Extensions record the workflow status and its history of active periods, the diagnosis use type, and the related person a family care certificate is issued for.
The supporting terminology covers the category of certificate (ordinary, child care, students, and issue while intoxicated), the observation components a case records, its conclusions including disability group determinations, the reason for the case, whether the patient is urban or rural, the local statuses a case moves through, regimen violations, the place of incident, the delay between onset and presentation, the degree of relation for family care, and local encounter classes. Certificate numbers have their own naming system.
Nine of the code systems are supplements: they add Uzbek and Russian designations to codes defined elsewhere rather than defining new concepts. They supplement SNOMED CT for tuberculosis body sites, screening body sites, diagnostic procedures, screening observation codes, observation types and breast findings; LOINC for laboratory tests; v3-ActCode for encounter class; and the FHIR request status codes for care plan status.
The SNOMED CT supplements are a temporary solution. A proper translation of SNOMED CT into Uzbek is underway, and once it is available these supplements will be retired in favour of it.
Diagnoses that are not constrained to a narrower list are bound to the full ICD-10 value set. The local observation code system holds the one concept, number of previous hospitalizations, that no standard covers.
Every area ships realistic examples: the four document Bundles, and the patients, practitioners, organizations, encounters, observations, service requests, specimens and reports they reference. Two example questionnaires are included, a complaints intake template for gynecologists with an example response, and a hepatitis B and C treatment questionnaire.