The anatomy of health data Dr Heather Leslie @omowizard
Anatomy (Greek anatomē, “dissection”) is the branch of biology concerned with the study of the structure of organisms and their parts. Health data – any data "related to health conditions, reproductive outcomes, causes of death, and quality of life" for an individual or population. We have a responsibility to question the structure & design rigour of our atomic health data Is Is our health data fi fit for purpose?
Single purpose – project or application Broad reuse Intent Short term 10, 20, 50 years Software engineers Business analysts Who? Clinicians Health Informaticians data Convert paper questionnaire forms Source? Replicate paper records design Careful requirements gathering ??? Persistent, summary data Type Event-based Questionnaire Clinical advisory board Broad range of professions Review Broad range of health domains Broad range of geographical locations
Single purpose – project or application Broad reuse Intent Short term 10, 20, 50 years Software engineers Business analysts Who? Clinicians Health Informaticians data Convert paper questionnaire forms Source? Replicate paper records design Careful requirements gathering ??? Persistent, summary data Type Event-based Questionnaire Clinical advisory board Broad range of professions Review Broad range of health domains Broad range of geographical locations
• SINGLE organisation • 15 years of questionnaire data • 7 different diseases • 83 data dictionaries • Each disease Smoking • Diagnosis • Family history example • Social history • Lifestyle • Pregnancy • Lab results • Measurements • Mental state • 4 different data managers over the years, tweaking…
• Cigarette smoking? • Tobacco smoking? • All smoking? • Tobacco use? • Smoking • Non-smoking eg snus Intent? • Typical use • Episodes of use • Actual use • On a specified day/date • Average in the past 5 years • Aged 20-30
Overview • Summary/cumulative • Pattern of tobacco smoking • Record once, updated with new versions
Diary • Measurement • Recording of tobacco smoking & associated context • Recorded as discrete separate events
Value added questionnaires Questionnaire Standardised pattern Context/definitions are implied Context/definitions are explicit • Ever smoked? Yes/No Ever smoked? • Ever smoked regularly for at least a year? Yes/No Never/Current/Former ± Definition of ‘Smoker’ Ever smoked 7 cigarettes? Yes/No Ever smoked cigarettes? Never/Current/Former ± Definition of ‘Smoker’ Do you regularly smoke cigarettes? Yes/No Cigarette smoking pattern? Daily/non-daily • Average number of cigarettes per day? Typical use? /day; /week '1- 4’; '5 - 14’; '15 - 24’; '25 - 39’; '40+’. OR • Amount of cigarettes? /day, /week, /month + Interval of time (start date/end date); Average
Chaotic Proprietary > < Systematic Open standard Fragmented Opaque > < Coherent Transparent
Silos are unsustainable
Silos are unsustainable
Time for an alternative approach?
Open atomic data standards Change of focus Benefits • Data, not applications • Common, shared • Health data • Standardise atomic data • EHR/EMR; Research; Registries; Reporting; • National clinical data dictionary Population health • Independent of vendors/systems • Clinical Decision Support • Complements existing standards • Semantic queries • Bootstrap new development • Exchange whole health record • Road map for existing vendors/projects • Aggregation/Analysis • Platform ecosystems, not silos • Big Data • Plug and play applications • Data sharing • AI & Precision medicine • … • …
www.openEHR.org/CKM
ISO 13606-2
Little data is the newest revolution! • Coherent, coordinated, connected health data • Open standards • Designed strategically • Clinically verified as 'fit for purpose’ • Independent of any single clinical system, vendor or project. • Realise the value of our health data; stop battling the chaos; harness it to make a difference
Dr Heather Leslie Atomica Informatics heather.leslie@atomicainformatics.com Contact @omowizard / @atomicainfo openEHR Foundation @openEHR / @clinicalmodels
Recommend
More recommend
