Crimson Spans Clinical -> Research systems to acquire materials under an IRB. Developed with Daedalus Software Inc. in Cambridge, MA (1) Identify discarded materials (2) process submitted/consented research samples. Study/IRB Registration & management. LIMS system: Manage lab workflow and resulting. Criteria Engine: Identify discarded materials. Billing system: Recoup costs for services. Download of data for banking and study staff
Criteria Engine Picklist (Accession#s) Crimson Holding Tank: 7-30 day rolling window of all clinical accessions Cohort Table Cohort IRB# LIS/APLIS data Flat-text files: Misys, Powerpath, other LIS i2b2 CRC Accessioning Honest Broker Workflow Engine/LIMS
Criteria Engine Picklist (Accession#s) Crimson Holding Tank: 7-30 day rolling window of all clinical accessions Cohort Table LIS/APLIS data Query Study Rule Set Flat-text files: Misys, Powerpath, other LIS i2b2 CRC IRB# Accessioning Honest Broker Workflow Engine/LIMS
Criteria Engine Picklist (Accession#s) Crimson Samples Located Holding Tank: 7-30 day rolling window of all clinical accessions Cohort Table LIS/APLIS data Query Flat-text files: Misys, Powerpath, other LIS i2b2 CRC Accessioning Honest Broker Workflow Engine/LIMS CMV
Criteria Engine Picklist (Accession#s) Crimson Samples Located Workflow Engine/LIMS Holding Tank: 7-30 day rolling window of all clinical accessions Cohort Table Crimson Patient ID (Not MRN#) Subject ID (Study-specific) Crimson Sample ID (Not Acc#) MRN (If consented) i2b2 CRC Sample Shipments Honest Broker Workbench Anon1 Anon2 Anon3 [..] Accessioning CMV
Impact of Crimson’s sample collection on the science.. Go- live February 2007 >15,000 samples handled the first year at BWH. Serving >35 studies Samples/Month, BWH Specimen bank Crimson go-live Feb 2007 MW Serum antibody 1.0 kb 0.5
Impact of Crimson on Sample Throughput Higher throughput at lower cost: “More for less”
Low throughput example: Study1 PI needs ~40 samples+controls to test proof of concept for autoantibody antigen. Samples not available from banks; no access to local clinics, nor staff to perform consent. Given time & effort to set up protocol, consent process, considers costs “infinite”
Study 1: Crimson Collection Query to find appropriate discarded serum samples (anonymzied). Costs of collection: $8-9/sample. ~10 samples/month + 10 controls. Collection completed in 2 months. Grant submitted to NIH on findings.
Medium Throughput Example: Study 8 >1000 Consented patients in study DNA and plasma to measure stable analytes. Daily samples while patients in the hospital. $18/sample for phlebotomy+processing in Specimen Bank + $20+/Sample in study coordinator effort to arrange phlebotomy and processing
Study 8: Crimson collection Coordinators register consented patients in Crimson under their study Crimson query identifies discarded purple tops from clinical labs. Cost: $8/sample + $1-2 coordinator’s time to register patients in Crimson. >60% reduction in sample costs. Estimated 10-20% increase in productivity of study staff and consent rate.
High throughput example: Study 4 Study desires 10,000 samples for epidemiologic analyses Avg. cost/sample for the study: $1,200 $12,000,000 to collect 10K samples Throughput of 5-10 samples/month 120 years to collect 10K with current process.
Study 4: Crimson Collection Forwarded cohorts via i2b2 Avg cost for collection: $8-9/sample Costs for collection of 10K samples: $85,000 Avg throughput: samples/month (1 Crimson node) with 2 Crimson nodes operational. Collection of controls in <1 year Experimental samples in years.
Crimson It works! Successful high-throughput sample collection for many studies. Significant cost-savings to investigators and research enterprise. Enabling communication with i2b2. Multiple Crimson nodes for collection.
BWH PathIT Neil Herring, MBA Diane Steidler Shadan Shafieha BWH Specimen Bank Kristina Lefteri Ashley Grant Partners IRB Judy Scheer, R.N. Libby Hohmann, M.D. Pearl O’Rourke, M.D. Research IS (dev) Diane Keogh Leo Braginsky Natalie Boutin Ana Holzbach Brent Richter Allan Harris Steve Roylance MGH Pathology Patrick Sluss, Ph.D. Chris Loftgren Anand Dighe, M.D. John Gilbertson, M.D. Daedalus Software -US Azita Sharif, CEO Erik Lickerman, M.D. Daedalus Software -PK Nighat Afsheen Khurram Naseem Annas Sheikh i2b2 + MGH LOC Shawn Murphy, M.D. Vivian Gainer Anil Dubey, M.D. Susanne Churchill, Ph.D. Zak Kohane, M.D. Acknowledgements Funding BWH Pathology Partners ResearchIS BWH Pathology Michael Gimbrone, M.D. Andy Onderdonk, Ph.D. Milenko Tanasijevic, M.D.