Project description
Making the most of data from electronic health records
Although highly valuable information can be extracted from electronic health records (EHRs), these remain unexploitable because they are unstructured and written in natural language. The EU-funded SAVANA project allows healthcare professionals to generate real-world evidence, make new discoveries, create personalised medicine and evaluate health outcomes. To achieve that, it will create a tool that uses natural language processing to extract data from massive amounts of EHRs’ clinical narratives. The new tool will satisfy the requirements of hospital ethics committees, national health services regulations and pharmaceutical industry policies and is addressed to managers, hospitals and researchers.
Objective
In the last twenty years, the average return on R&D expenditure in the pharma industry has dropped from almost 18% to 3.7%. Moreover, annual funding for biomedical research has more than doubled while new drugs approvals have declined by one third. There is a wide consensus that the main cause of this problem is the exhaustion of a model intended to develop ‘broad indications’ and the need for a new ‘precision medicine’ model. We simply do not know enough about the underlying disease mechanisms involved, and more research is required to develop better disease classifications, which will enable a more targeted development approach for drugs and therapies.
Electronic Health Records (EHRs) has been used for more than ten years in most developed countries, and they gather now exhaustive clinical information of millions of patients. Leveraging EHRs could accelerate clinical research, and improve healthcare quality.
However, in order to uncover unknown disease models from EHRs, precision medicine requires massive research studies on thousands of patients (often in several countries). Currently there is no tool capable of: 1) automating the extraction of data from EHRs, and also, solving the privacy concerns raised by EHRs.
SAVANA RESEARCH uses Natural Language Processing to extract data from massive amounts of EHRs’ clinical narratives. It has the following advantages intended to make a leap in clinical research efficiency: 1) It uses only de-identified clinical records and ensures state of the art technologies to protect data privacy; 2) It is capable of decoding ten times more EHRs in half of the time; 3) It is capable of identifying 100 times more variables from EHRs; 4) And it costs 40% less.
The application of NLP to healthcare is a fast-growing market that is expected to reach 2.65 billion by 2021, by growing at a CAGR of 20.8%. SAVANA RESEARCH’s target markets are primary Europe and North America, which together comprises 75% of all clinical trials worldwide.
Fields of science
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
- natural sciencescomputer and information sciencesdata sciencebig data
- natural sciencescomputer and information sciencesdata sciencenatural language processing
- natural sciencescomputer and information sciencessoftwaresoftware development
- medical and health scienceshealth sciencespersonalized medicine
- medical and health sciencesclinical medicinehepatology
Programme(s)
Funding Scheme
SME-2 - SME instrument phase 2Coordinator
28016 Madrid
Spain
The organization defined itself as SME (small and medium-sized enterprise) at the time the Grant Agreement was signed.