Research departments in many healthcare organisations are a very different world to the hospitals running routine clinical operations. As we explore more in this domain in the last years, I have visited a number of research laboratories, mainly engaged in the clinical genomics. That was quite a broad overview, including the US, European and Israeli leading research centres.
As I have covered in previous posts (see Related information, below), we are co-operating with some of the leading research organisations in the world to create a clinical research platform, based on data-lake technologies. It is supposed to enable and streamline the highly data-intensive processes of medical centres and labs and to support numerous regulatory requirements in this area.
Working with laboratory staff is very interesting! They are extremely technology- savvy - not only those who have degrees in bioinformatics, but also micro-biologists and molecular pathologists! At the same time, it is typical for this environment to be less organised and less systematic with respect to how clinical operations are run. The nature of the work dictates it!
‘Fuzzy data discovery’
As a society, we still know very little about the science of human genomes and their impact on our wellbeing. There are very few known gene mutations associated with cancers or cardiovascular diseases. There are very few approved drugs on the market to treat cancer cases. Along with the hi-tech molecular and genome sequencing, there are techniques that involve a lot of fuzzy data discovery through the public and commercial knowledge bases and scientific articles.
Treating physicians are seeking advice from their peers in research departments to find out the solutions beyond the standard treatment protocols. It transforms pure scientific, grants-driven research into translation research which makes precision and personalised medicine work in practice. Going in those directions, labs and their parent clinical centres have to collaborate very closely overcoming challenges of data exchange and consent management.
Creating a roadmap
In order to systematise the data management expectations and design a roadmap for the different types of research centers I tried to classify these organisations by specific aspects of their research scenarios. The following table is not intended to be a complete technology map, but to highlight some similarities and differences between those organisations.
Scientific research organisations
Translational research/Precision medicine
Data Lake – a platform for integrated data sources, big data support (like HADOOP) with high compute and multiparallel processing, elastic to support a growing compute and store demand.
Ingestion toolset – ready to support both phenotype and genotype origin of data, structured and free text public data sources.
Data catalogs - to register and make the data sources known to users and applications.
Data discovery engine – with similarity measurement techniques.
Data governance services - with retention policies, chain of custody and traceability addressing local regulations for sensitive patient data.
On boarding and off boarding workflow management - necessary to support any funded research.
Integrated patient data environment - necessary to process real patient data through the translational applications and generate reportable outcome to support better treatment decisions (enriching clinical decision support) for the EMRs.
Data anonymisation algorithms.
Privacy and informed consent management.
Collaborative analytical workbenches/workspaces.
Application development environment for specific precision medicine algorithms.
Accessibility of precision medicine tools
Big academic hospitals typically have to be engaged in both scientific and translational research - though, ultimately, precision medicine tools should become accessible for medical organisations of any size and maturity level. Technology should allow for gradual involvement, depending on availability of skills and level of ambitions: some would contribute their own application and algorithms, some would benefit from using pre-built tools.
How are you building your research environment? What are the tools that apply to your organisation? As ever, we welcome your thoughts, comments and queries!
Learn how to begin your own journey towards precision medicine at a dedicated session at Momentum 16, Barcelona, October 31 – November 3, 2016. Dr. Long Phe Le of Massachusetts General Hospital will share his experiences of building a clinical research platform in the session, Thursday, November 3. Find out more here.