The aim of SEMCARE is to build a semantic data platform to support clinical trials. The platform will identify patient cohorts based on patient-level criteria (e.g. age, gender, diagnosis, symptoms, lab results) scattered in heterogeneous clinical data. As most of patient-level data is unstructured, language technologies are necessary to extract and exploit the relevant data. Our platform combines the power of full text search with text analytics and semantic web technologies for a hybrid semantic full-text search. SEMCARE addresses the ever growing need to exploit medical data from clinical trials and for monitoring and improving healthcare delivery. The platform has the potential to provide for a more efficient, scalable method of patient recruitment for clinical trials. SEMCARE will also be capable to detect undiagnosed patients with rare diseases based on sign and symptom combinations. This has the potential to speed up the research on this group of diseases, which constitute a highly attractive market for pharmaceutical companies. The SEMCARE consortium includes players from the demand and supply sides. Hospitals in three different European countries serve as pilot sites and have particular interests in applying SEMCARE in their clinical research activities. The companies Averbis GmbH and SYNAPSE have a clear interest in the exploitation of the results.
4. BACKGROUND
Healthcare needs have been changing
•There is a requirement for new, safer, more effective medicines in areas of changing medical needs
•With the pressure on healthcare budgets there is a focus on best practice care
Source: EHR4CR
5. CHALLENGE
Less drugs come to market1…
…despite more investments in research
1. EFPIA. The Pharmaceutical Industry in Figures
Source: EHR4CR
6. INCREASING BURDEN ON CLINICAL TRIALS
Protocol Design
•Trials have become increasingly complicated
•Advances in healthcare, such as personalised medicine, require more patient information
Patient Recruitment
•Larger on average and require more participants.
•Recruitment has become more difficult and expensive
1999
2005
% Change
Total Procedures per Trial
24
35
46%
Length of Clinical Trials
96
158
65%
Participant Enrolment Rate
460
780
70%
Staff work burden (in effort units)
21
35
67%
Source: PhRMA Report 2010
7. OBVIOUS INEFFICIENCIES IN CLINICAL DATA MANAGEMENT
Patient Data Care
EHR Records
Clinical Trial Research Database
Electronic Data capture for Clinical Trials
8. THERE IS A NEED TO BRIDGE THE GAP
HOSPITAL
TRUSTED CLOUD
STUDY PORTAL
ETL Anonymization
Text Mining
Data
Warehouse
Data Mining
Anonymized Text
Annotations
Structured Data
9. AVERBIS HEALTHCARE ANALYTICS
Integration of heterogeneous data in various input formats
Gain insights in structured and unstructured data
Anonymization of unstructured data for research purposes
Easy and powerful interfaces
for experts and novices
Averbis Healthcare Analytics is a clinical discovery platform to analyze Electronic Health Records
Open infrastructure, easy 3rd party integration (e.g. Qlikview)
10. AVERBIS HEALTHCARE ANALYTICS
Averbis Healthcare Analytics is a clinical discovery platform
to analyze Electronic Health Records
3rd party applications
QlikView
I2B2
Transmart
Terminology Management
Text Mining
Search & Analytics
Services
ETL, Data Integration
Data Privacy, Anonymization
Data Storage
11. AVERBIS HEALTHCARE ANALYTICS
Averbis Healthcare Analytics is a clinical discovery platform
to analyze Electronic Health Records
3rd party applications QlikView I2B2 Transmart
Terminology Management
Text Mining
Search & Analytics
Services
ETL, Data Integration
Data Privacy, Anonymization
Data Storage
12. Coding & Billing
Quality Management
AVERBIS HEALTHCARE ANALYTICS
Predictive Analytics
Diagnosis Support
Feasibility Studies
Patient Recruitment
Semantic Interoperability
Hypothesis Validation
Commercial Insights for Pharma
Healthcare analytics support various requirements of clinical care, clinical research, administration and governance
Clinical Care
Research
Pharma
13. Coding & Billing
Quality Management
AVERBIS HEALTHCARE ANALYTICS
Predictive Analytics
Diagnosis Support
Feasibility Studies
Patient Recruitment
Semantic Interoperability
Hypothesis Validation
Commercial Insights for Pharma
Healthcare analytics support various requirements of clinical care, clinical research, administration and governance
Clinical Care
Research
Pharma
14. DIAGNOSIS SUPPORT
Show me all patients with typical Morbus Pompe symptoms, but without a Morbus Pompe diagnosis
16. Coding & Billing
Quality Management
AVERBIS HEALTHCARE ANALYTICS
Predictive Analytics
Diagnosis Support
Feasibility Studies
Patient Recruitment
Semantic Interoperability
Hypothesis Validation
Commercial Insights for Pharma
Healthcare analytics support various requirements of clinical care, clinical research, administration and governance
Clinical Care
Research
Pharma
19. OUR GOAL IN SEMCARE
19
Focus Overcoming barriers that limit access to EHRs for research Delivering a platform to re-use electronic health records
Use-Case
Detecting patients
• in the area of cardiology
• with undiagnosed (rare) diseases
• for clinical studies Timeline 2-years project, 1,5 Mio EUR budget
In cooperation with a broad network of pharma companies
20. For further questions, please contact:
Dr. Philipp Daumke
+ 49 (0)761 203 97690
philipp.daumke@averbis.com