Secure (Research) Data Desktop Stuart C. Ray, MD y, Director, - PowerPoint PPT Presentation

Secure (Research) Data Desktop Stuart C. Ray, MD y, Director, Infectious Diseases Fellowship Training Program Professor of Medicine and Oncology Johns Hopkins Medical Institutions

Secure (Research) Data Desktop Secure (Research) Data Desktop • Secure staging area for enterprise data Secure staging area for enterprise data – Access – Analysis Analysis • Guiding principles – Utility – Community – Economy – Security

Growing Population of Data Requestors Data requests Data requests Experienced Data Managers

Goals Goals • Implement data security policies relevant to Implement data security policies relevant to research data • Promote good data management practices • Promote good data management practices • Promote good data analysis practices • Promote collaborative analysis • Carrot (rather than stick) approach to adoption

Realities Realities • There is no one ‐ size ‐ fits ‐ all solution for data There is no one size fits all solution for data • Specification in development • Initial funding imminent i i l f di i i • J ‐ CHiP pilot has been well ‐ received, instructive

Virtual Machines ‐ Utility Virtual Machines Utility • Access via thin client/browser Access via thin client/browser • State maintenance • Rapid deployment id d l • Integrated backup • Customizable • Streamlined data storage approval likely

Virtual Machines ‐ Community Virtual Machines Community • Shared context Shared context – Access – User experience User experience – Applications • Data sharing with reduced risk D t h i ith d d i k • Multi ‐ level support (IT, biostats, peers)

Virtual Machines ‐ Economy Virtual Machines Economy • Scale to cost ~$20/month (clinical version Scale to cost $20/month (clinical version ~$12) • Convenient • Convenient – Access via thin client/browser – State maintenance S i • Standardizable

Virtual Machines ‐ Security Virtual Machines Security • Standardized base configuration Standardized base configuration • Software/port controls • Audit logs di l • (Streamlined data storage approval likely)

Use cases – high ‐ level examples Use cases high level examples • A post ‐ doc fellow performs a pilot study using A post doc fellow performs a pilot study using JHM Enterprise clinical data • A trainee analyzes data from a medium sized • A trainee analyzes data from a medium ‐ sized formal research study • A PI is participating in a multi ‐ center study is A PI i i i i i l i d i provided with a large dataset to analyze

Needs Needs • Use cases for SRDD: highlighting strengths Use cases for SRDD: highlighting strengths, risks, etc • Specification of tools: biostatistical GIS • Specification of tools: biostatistical, GIS, visualization, reporting, collaboration, etc • Policies: authorization, authentication, P li i h i i h i i auditing, etc • Cost recovery? y