The role of Big Data in digital healthcare transformation
- industry Highlights
- 5 years ago
The value of the health treatment
When visiting their doctor or going to the hospital, people tend to have faith that the health care professionals involved are giving them the appropriate treatment.
However, since the average healthy individual spends no more than ten minutes in a health facility, it is unlikely that a physician will manage to assess the state of their overall health, apart from emergency cases. The “Evidence-Based Medicine” (EBM) approach which physicians usually apply prescribes drugs or selects treatment based on proven success in clinical research, but even this has its limitations.
Electronic Health Records (EHRs) technologies have been on the market for some time. However, with the data stored across a number of different systems and formats, they are not designed in a logical way. Still, a staggering 80% of medical and clinical patient information comprises unstructured data, such as written physician notes, consultant notes, radiology notes, pathology results, hospital discharge notes, etc. This begs the question, what will happen in 2020 when healthcare data is expected to reach 25,000 petabytes, a 50-fold increase from 2012 levels?
Too many patients, very little time
There is an urgent need to develop new, scalable and expandable big data infrastructure and analytical methods that can enable healthcare providers to access knowledge for the individual patient, yielding better and faster decisions and outcomes. This is not possible without the development of the “personalized medicine” concept: a framework that leverages patient EHRs and OMICS (primarily genomics) data to facilitate clinical decision making that is predictive, personalized, preventive and participatory.
With the personalized medicine approach, clinicians are able to:
- Validate medical treatments and responses to therapies. Thiss allows them to predict the possible side effects, and detect adverse events to treatments, based on the genetic make up for each individual patient in comparison to other similar patients.
- Describe better targeted therapies for individuals through easily determining which drug(s) will work better with each individual patient, as opposed to adopting an empirically driven approach of trial-and-error.
- Make better decisions on risk prediction and prevention focus, rather than disease management. The collection of genetic information from previous family tree testing can be useful in identifying possible future diseases that can either be avoided, or adequately controlled.
It’s impossible to talk about personalized medicine without recognizing its deep integration with Big Data. Focusing on the great volume of diagnostic images, genetic test results and biometric information raises several challenges related to data integration, processing and analytics, visualization and interpretation that need to be addressed.
the need to integrate collected data – Clinical information (e.g. medical diagnosis, medical images, patient histories) and biological data (e.g. gene, protein sequences, functions, biological process and pathways) have diverse formats and are generated from different sources. It is very important to focus on developing tools and techniques to make better sense of these data, and use the information they provide for further knowledge discovery.
Quality of data, standardizing data and terminology- This typically requires extending existing data libraries or the development of domain-specific ontologies.
Volume of data, huge amounts of data itself does not solve problems- Data should be summarized in a meaningful way in order to properly deliver the information, knowledge and, ultimately, wisdom that help decision making.
Data velocity, healthcare data is changing and evolving, and will continue to do so- These rapid changes in the data pose a significant challenge to offering fast searching, browsing, and analysis of real-time content. Users should have: (1) the ability to filter, prioritize and rank the data (relevant to the domain, or use case); (2) the ability to quickly process and digest data; and (3) the ability to draw on, evolve, and hone in on relevant background knowledge.
Besides the challenges posed by technology and infrastructure when using smart data to enable personalized medicine, there are several other administrative challenges. Healthcare organizations and policy makers should promote a fundamental shift in their decision making, and embrace a ‘brave new world’ that promotes data sharing with appropriate security and privacy protection. New policy guidelines should also be created in order to enable faster digital transformation.
Thanks to its previous expertise in healthcare, and the renowned capabilities of the SAP HANA Cloud Platform, Sqilline is now working on developing a Breast Cancer Risk Assessment Tool that will evaluate data of people with a family history of breast cancer, combining various data sources, e.g. electronic cancer registries. The overall objective of the tool is to provide opportunities for the effective prevention and treatment of breast cancer.