"The World of Clinical Information Models" 30-jan-2014 Michael van der Zel Clinical Information Systems Architect
CSI?
Techniek letters; chinees, arabisch 你好 أهل وسهل Structure technische samenstelling Syntax zin samenstelling, leesrichting Semantics woordbetekenis Context doel en aannames voor gebruik Proces afspraken over uitvoeren Bron: Bernd Blobel, University of Regensburg CSI Stairway to Heaven
The baseline Infostructure in 2013. Will it fly? How long? With how much effort?
Mission CIMI Mission & Goals Improve the interoperability of healthcare systems through shared implementable clinical information models. Note: implementable means models Goals that can be transformed into software artifacts. Establish and maintain a shared repository of detailed clinical information models Establish appropriate governance, finance, and organizational structure for the group Capture background, context, and discussions relevant to decisions made in creating and using the model Establish a fair and open process for curating the content in the model repository Promote the creation of software tools that translate the models to other commonly used model representation languages (OWL, UML, HL7 MIF, graphic format, HTML, etc.) Promote the creation of software tools that generate implementation artifacts (XML schemas, Java class definitions, A templates, GreenA, etc.)
Betrokken Organisaties Ministry of Health Holdings Singapore, LifeLines (NL), Health Intersections, ONC (USA), NHS Connecting for Health (UK), SMArt, Electronic Record Services, EN13606 Association (EU), B2i Healthcare (AU), Cambio Healthcare Systems, Canada Health Infoway (CA), ISC, GE Healthcare (USA), HL7, IHTSDO, Intermountain Healthcare (US), JP Systems, Kaiser Permanente (US), Mayo Clinic (US), National Institutes of Health (US), Ocean Informatics (UK/AU), OpenEHR, Results4Care (EU), South Korea Yonsei University, Tolven, Veterans Health Administration (VA USA),... N.B. Amerika, Europa, Australie, Azie, en de rest?
m ean in g : T EXT ta rg e t: EHR_URI type : T EXT l in ks LOCATABLE 0..* p arti ci p a ti o n s ca teg o ry: CODEABL E_T EXT l an gua g e : CODE_ PHRASE territory: CO DE_ PHRASE conte nt 0..* a rch etyp e _ id : ARCHET YPE_ID tem p l ate _i d : T EM PL AT E_ID [0..1] rm _ ve rsi o n: Strin g PARTICIPATION 0..* COMPOSITION CIMI ARCHETYPED arch e typ e _de tai l s arche typ e _n od e_ i d : Stri ng na m e: T EXT ui d : UID_ BASED_ ID [0..1] p arty PARTY_PROXY fu ncti on : T EXT m od e : CO DEABLE_ T EXT ti m e : DAT E_T IM E [0..1 ] 1..1 CONTENT_ITEM CEM 0..* i te m s 1..* d ata CIMI RM 1.0.2 nul l _ fl avor: CODEABL E_T EXT [0..1 ] i te m s SECTION ENTRY l a n gu age : CODE_ PHRASE val u e ITEM DATA_VALUE 0..1 1..* Standard Terminologies CLUSTER structure_ typ e: CODE_ PHRASE [0..1] ELEMENT CEMs DCMs A Templates openehr Archetypes CEN Archetypes Repository of Shared Models in a Single Language LRA Models CMETs, HMDs RMIMs Initial Loading of Repository V2 LRA V2 XML HTML V3 XML FHIR UML openehr Archetype Translators CEN Archetype A OWL SOA Payload Thanks to Stan Huff, Intermountain Healthcare LINK
Imported Intermountain Healthcare Klinische Bouwsteen «roo tco nce p t» Bloodpressure? «roo tco nce pt» BloodPressurePanel S o u rce : n l.nictiz.b lo o dp re ssure -v0.10 1 Imported OpenEHR Archetype S o u rce CE B l oo d P re ssu re P an e l.xm l with M o de l V alu e S e ts (20 1 1-fe b -1 3 ).xls «rootcon ce p t» Blood Pressure S o u rce h ttp://d cm.ne hta.o rg.au /ckm / o p en E HR-E HRO B S E RVA T ION.b ody_ we ig ht.v1.xm l 1 B lo e dd ru kmeetw a a rd e «d a ta» Bov endruk «q ua l ifier,e nu m e ra... Meetlocatie «en u m» Rech terb o ve n a rm Li n ke rb o ven a rm Rech terd i j Li n ke rd ij Rech terp o ls Li n ke rp o ls V in ger Rech tere n kel Li n ke re n ke l 1 B lo e dd ru kmeetw a a rd e «d a ta» Onderdruk «q u alifie r,enu m e ra... Meetmethode «en um» Ni et-in vasie f In va si e f 0..1 0..1 PQ «d a ta» SystolicBloodPressureMeas «q u alifie r,e nu m e ra... BodyLocationPrecoord? «q u alifie r,enu m e ra ti... MethodDev ice? PQ PQ «da ta» Diastolic B lo e dd ru kmeetw a a rd e? PQ «d a ta,d eri va ti o n» GemiddeldeArterieleDruk «d a ta» MeanArterialPressureMeas «sta te,e n u m era tio n» Body Position 0..1 «q uali fi er,e n um era... BodyPosition «e n u m» Hea d o f b ed ra ise d Lyin g o n l eft side Lyin g o n rig ht sid e P ro ne S ea te d S ta nd in g S up in e «en u m» S ta and Li g g end Zi tte nd A ch tero verle u n end La te raa l L ig ge nd T re nde le n b urg Noticeable differences: Systolic & Diastolic cardinality Labels. With SCT code bindings. Valuesets. Noticeable differences: No distinction between state & qualifier. Labels. Code binding external. Cardinality always? Valuesets.? «q ua l ifier,e num e ra... Location of measurement «enu m» In take M eth o d De vice In traven o us Ca th ete r T ube «d a ta» DiastolicBloodPressureMeas PQ «da ta» Systolic PQ «da ta» Mean Arterial Pressure «q ua li fi er,e num e ra... Method «e n u m» A uscu lta tio n In va sive M ach in e P alp a ti o n «en u m» Fi n g er In tra -arte rial Le ft ankl e Le ft arm Le ft thi g h Le ft wrist Rig ht a n kl e Rig ht a rm Rig ht thi g h Rig ht wrist T oe?? «sta te» Position «en u m» Lyi ng Lyi ng wi th ti lt to l e ft Recl ini n g S ittin g S ta ndi ng Noticeable differences: Cardinality always? Labels. No code bindings. Valuesets. Some (models) are more equal than others George Orwell
Thanks to Linda Bird, MOH Holdings, Singnapore Diagnosis Example
e b u C l e c i d t o n a M l m a e c s i n i l Iso C m e l b o Concept r P
Clinical Model Cubes
Granularity Medical Condition (e.g. Diabetes Record) Questionnaire Apgar Score Propensity to Adverse Reaction Full BP Systolic Bp Number of Data Elements
Wearables & Implantables H Where information meets biology blood flow heart rate vascular pressure adrenal secretion oxygen intake digestion bone growth circulation excretion heart regulation maintenance of tissue moisture thyroid secretion nutrient absorption salivary secretion carbon dioxide removal ocular transmission cochlear transmission H WebSeries https://www.youtube.com/watch?v=otlbzq4zv_i cellular repair immune system response detoxification protein synthesis homeostasis regulation insulin production nervous system response respiration
All models are wrong, some are useful George Box @ m.van.der.zel AT umcg.nl