From little data, big data quality grows
A special presentation on how approaches to getting little data ‘right’ supports better quality when it comes to big data will feature at the Big Data 2014 conference, being held in Melbourne next week.
Presented by leading informaticians Heather Leslie and Hugh Leslie from Ocean Informatics, the 'little data' workshop will provide an introduction to clinical data models and archetypes, computable specifications for a single, discrete clinical concept.
Archetypes describe clinical data such as pulse, blood pressure, symptom, medication order and allergy or adverse reaction, which when combined with terminologies such as SNOMED CT and LOINC can provide a powerful approach to record, store, exchange and use unambiguous clinical data.
Dr Heather Leslie says the development of archetypes is a method to generate the ‘little data’ and ensure it is high quality. “The more systems that utilise a single data pattern, as defined by the clinical archetype, the more potential for interoperable health data,” she says.
The workshop will also cover international approaches to standardisation and collaboration around data models, including the development of the Clinical Information Modelling Initiative (CIMI) and the openEHR movement.
Organised by the Health Informatics Society of Australia (HISA), Big Data 2014 has as its theme 'Big insights: Harnessing the power of health data'. The conference will focus on health outcomes, data visualisation, workforce, data linkage and analytics, genomics and privacy.
Keynote speakers include Angelo Joseph, head of sales engineering at Google Enterprise, who will discuss how insights from Google into big data can offer big opportunities and big thinking.
Big data generated by genomics will be high on the agenda, with presentations from Terry Speed, head of the division of bioinformatics at the Walter and Eliza Hall Institute of Medical Research (WEHI), who will discuss changes in bioinformatics since 1997.
Lindsay Kiley of InterSystems will discuss driving patient benefits in the areas of informatics, genomics and biological system modelling through big health; Denis Bauer of CSIRO will look at personalised, cloud-computed genomics at the health system scale; and Liviu Constantinescu of the Garvan Institute will discuss implementing a clinical genomics infrastructure to sequence 18,000 human genomes per year.
At the hands-on end of the scale, Jason Ferriggi of the Inner East Melbourne Medicare Local will discuss how to extract big data from small sites by creating a primary care data warehouse; and Kate Birch of the University of Melbourne will discuss approaches to data capture at the bedside to combine patient management with research.
How to get doctors interested in big data is the theme of another keynote presentation by Christine Jorm of the Sydney Medical School.
The vexed question of privacy and big data is the theme of a panel discussion on the first day of the conference. HISA will launch a privacy guideline following the panel discussion.
Big Data 2014 will be held on April 3 and 4 at the Pullman Albert Park in Melbourne. For more information on 'little data', archetypes and their practical use in the Northern Territory, see Heather Leslie's article in the April issue of Pulse+IT magazine, out next week.
Posted in Australian eHealth