Clinical Data Management: Strategies for unregulated data


Published on

Lightning talk presented at Research Data Access and Preservation 2013 in Baltimore, MD.

Published in: Education
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Society for Clinical Data Management developed and maintains the Good Clinical Data Management Practices guidelines in concordance with the ICH Good Clinical Practices ( for clinical trials. While clinical trials are highly regulated with focused questions and strict guidelines for operation, some common practices can be applied to studies occurring outside a regulated environment.Clinical Data Management strategies: How can they improve data management and sharing for non-clinical research?Unlike data curation, clinical data management (CDM) is a recognized area of expertise and a defined career path. The highly regulated clinical trials environment has produced effective and efficient practices that can be generalized to other areas of research. Good Clinical Practice (GCP) is an international standard developed by the International Conference on Harmonisation that specifies how clinical trials should be conducted and defines the roles and responsibilities of various sponsors, investigators, and monitors. These practices address many of the issues at the core of data curation and sharing. Much academic research is not rigidly structured in the manner of clinical trials. Relevant practices within CDM and GCP must be reinterpreted for non-clinical research so that they can inform general data management, sharing, and preservation practice. This lightning talk will highlight effective strategies from CDM and GCP that promote data integrity, facilitate data preservation and sharing, and facilitate reproducibility of results. -find interesting free fonts (2)-find a good color schemeResources-see Evernote note-
  • Characteristics of clinical data management
  • Characteristics of clinical data management
  • Characteristics of clinical data management
  • The primary goal of CRF design is to collect all the data required by the protocol in such a way that it can be analyzed according to the protocol and statistical analysis plan.
  • Clinical trials have similar data collection issues to social science studies: variety of data types coming in in multiple streams. The GCDMP includes specifications on how to best manage the three types of data streams in clinical trials: CRF, patient reported outcomes, lab data.
  • They aren’t pretty or magical
  • Why would a researcher in a unregulated environment adopt something like a CRF? A CRF is basically a checklist + standardized input of dataChecklist to ensure complete data collectionForm to ensure standardized data collection and entry
  • CRF Book relates the CRF to the protocol – defines what data should be collected and what data must be collected specified by the protocol
  • If everyone had to figure this out for themselves, it would be as variable as the social sciences generally are. Standards in data management free up researchers to focus on the design and analysis, which is what they typically are about anyway.
  • Clinical Data Management: Strategies for unregulated data

    1. 1. Clinical Data Management: Strategies for unregulated data Heather Coates, IUPUI University Library RDAP Summit: April 4, 2013
    2. 2. HIPAA ICH GCP Clinical Trials Clinical Data Management FDA
    3. 3. Regulation  Standard Practice • • • • • Efficiency Efficacy* Safety* Accuracy Confidentiality/Privacy*
    4. 4. • Clear expectations • Standards • Best practices established • Burdensome • Inflexible • Expensive
    5. 5. Data integration Data acquisition Data standards Data review DMP Clinical Data Management Data validation Database design & programming System implementation Coding CRF design
    6. 6. Good Clinical Data Management Practices • 20 areas in 2011 document • General themes – Plan, test, revise, test…implement – All stakeholders involved in design of protocol, data collection tools, data management plan, etc. – Document, document, document – Rule: the bigger the study (sites, data, people), the more planning you need
    7. 7. Good Clinical Data Management Practices • Specify documents required for reproducible research – Organization: SOP – Study: Protocol, Manual of procedures, Data management plan, Statistical analysis plan • Documentation serves practical purposes and benefits the team immediately • Allows specification of roles and responsibilities from the beginning
    8. 8. Good Clinical Data Management Practices Begin with the end in mind OR Produce report-ready output Collect data in a way that allows for efficient data entry, processing, validation, and analysis Enabled by standardized data collection tools (CRF)
    9. 9. Case Report Forms (CRF) • • • • Efficient (concise) Effective (clear) Minimize redundancy Minimize human error – consider completeness, accuracy, legibility, timelin ess • Enables fast data transfer across studies
    10. 10. Raw data Processed data Analysis
    11. 11. Checklist + Form
    12. 12. CRF + Instructions = CRF Book
    13. 13. Why do these strategies work? • Save time and money • Regulated environment – compliance is enforced • Clinical trials are similar in structure and question are fairly narrow in scope BUT!!! • GCDMP provide practical strategies that meet regulatory requirements
    14. 14. References & Resources 1. 2. 3. 4. Society for Clinical Data Management. (2011). Good Clinical Data Management Practices. Washington, D.C. ICH GCP E6. Retrieved from efficacy/efficacy-single/article/good-clinical-practice.html Center for Cancer Research. (nd). Managing Data in Clinical Research. Retrieved from Howard, K. (2005). Data management in clinical trials. Retrieved from e_Study.pdf