discusion 9
username of me
ALHEChapter20_EResearch.pptxChapter 20: eResearch
John Sharp
Learning Objectives
After reviewing the presentation, viewers should be able to:
Describe the scope of eResearch and Clinical Research Informatics within the clinical research workflow
Describe the use of EHR data in various phases of research including research originating from EHR data
Conceptualize how informatics tools can be utilized in recruiting subjects for clinical research
Detail how informatics supports the ongoing management of clinical trials
Review the new trends in big data, real-time analytics and data mining
Definition
eResearch: use of information technology to support research
In the past few years we have witnessed the shift from paper-based research to almost completely electronic
Major contributing factors: adoption of electronic medical records and electronic research platforms
There is no doubt that health informatics and specifically eResearch will have a major impact on evidence based medicine in the future
The new field to study eResearch is Clinical Research Informatics
Preparatory to Research
Electronic retrieval of information:
PubMed
National Library of Medicine
Google Scholar
Google Books
ClinicalTrials.gov (WHO for international)
Research collaboration networks have seen significant growth in recent years. Research networks are typically web-based applications which include features such as a personal profile, opportunities to connect with others with similar interests and the ability to post status updates
Preparatory to Research
Research Collaborative Networks Tools
Vivo: An open source tool developed at Cornell University
Harvard Profiles Catalyst: An open source community of over 130 member institutions with built-in network analysis and data visualization tool
SciVal Experts: Commercial solution to find research funding and measure benchmarks
EHR Recruiting: ability to evaluate adequate pools of patients to be recruited into the study. This requires a clinical data repository from EHR data with a query tool to search de-identified clinical information. By modifying inclusion and exclusion criteria, a researcher can find the appropriate cohort for recruitment based on a reasonable recruitment rate
Electronic grant process: researchers can search for grant opportunities and grant submission is now common for government and civilian agencies
Preparatory to Research
Study Initiation
Volunteer recruitment on the Internet
ResearchMatch: matches patients seeking clinical trials and researchers seeking volunteers
TrialX: permits volunteer to search clinical trials from ClinicalTrials.gov
Social network: example, ArmyOfWomen
EHR can be used to find cohorts of eligible patients and create patient contact lists (alerts) for recruitment. Clinical trial alerts can be embedded within EHR based on diagnoses, lab tests or other patient characteristics. Alert would typically remind provider that patient may be eligible for a clinical trial and who to contact
ResearchMatch Program
Study Management and Data Management
Clinical trial management systems (CTMS)
Manage the planning, preparation, performance, and reporting of clinical trials
Budget management, study calendar of patient visits, and creating electronic case report forms (eCRFs)
Examples of CTMSs
Research Electronic Data Capture (REDCap) by Vanderbilt
OpenClinica
REDCap Program
EHRs and Clinical Trials
Integration is rarely available within commercial EHRs
Data from EHRs can be exported and then imported into study data management systems
Design a variety of study types: epidemiologic research
Identification of risk factors
Comparative effectiveness research
Challenges with EHR Data
Data that is not routinely collected in EHRs can be collected with “smart forms”
Collection of research data using medical devices (e.g. EKG). Many organizations are integrating device data with EHRs
Patient Reported Outcomes (PROs): is the term used to denote health data that is provided by the patient through a system of reporting. This data might be collected with a tablet and inputted into the EHR
Data Management Systems for FDA Regulated Studies
Regulation 21 CFR Part 11
Selecting a system compatible with regulatory requirements
Significant validation tests must be developed and executed
Commercial programs may assist: PhaseForward and Oracle Clinical
Open source programs such as OpenClinica can help
Interfaces and Query Tools
Clinical data repositories to support research are commonplace and based on EHR data:
i2b2: Informatics for Integrating Biology and the Bedside
TrialViz (UK)
STRIDE (Stanford University)
Challenges: gain regular access to source clinical systems and preservation of semantics across systems during aggregation process
Natural language processing of unstructured EHR data is critical
Health Information Organizations are also a rich resource for research
Big data means big research tools such as Hadoop. The Apache Hadoop software library “is a framework that allows for the distributed processing of large data sets across clusters of computers using a simple programming model”
Research can come from voluminous image data
Research can also arise from genomic information integrated with EHR data
Interfaces and Query Tools
Data Analysis
With tools like The R Project for Statistical Computing, an open source statistical package, there is the potential for integration of the statistical package with the data repository
REDCap provides access to their API (Application Programming Interface) to connect directly to statistical programs. SAS also provides for integration of patient data from a variety of sources with tools for data cleaning, standardization and exploration
Data visualization is a new and evolving field to assist research
Real time analytics is the provision of analyzed data relatively instantly to support decision making. IBM’s Watson is the best example we have today
eResearch has become an almost paperless process
There is a need for Clinical Research Informaticists
EHR data is the largest source of research data today
There are obstacles in using all EHR data for research because much of it is unstructured
Informatics tools can be used for patient recruiting and management of research
Big data is the results of EHRs, imaging and genomics so that researchers must have tools to analyze these huge data sets
Conclusions
