Guidelines for Submitting Proposals

Proposals with the following project features will be responsive:

  • Use of either identified or deidentified phenotypic and/or demographic data from the electronic medical record (EMR/RPDR).  If the use of deidentified data is proposed, a large cohort of several thousand subjects would be reasonable. If fully consented or identified data are proposed, the cost and time for data accrual must be reasonably achievable within the funding period.
  • Use of some (or all) of the PCPGM resources when appropriate (click here for a list of resources).  Use of external services and/or resources outside of PCPGM should be sufficiently justified. 
  • Inclusion of weekly project meetings between the PI and the multidisciplinary PCPGM staff team.
  • Clinical trials are excluded from this RFP, but PIs who are conducting an IRB-approved clinical study/clinical trial are eligible to apply for funding to support to project as an ancillary study.

Example Project: Asthma

A subset of asthma patients using inhaled corticosteroids experience frequent ER visits and hospitalizations, demonstrating an apparent refractoriness to the classical therapy for this problem.  Using the RPDR and Natural Language Processing (NLP) tools developed by i2b2, 96,000 patients were identified as diagnosed with asthma.  They were subsequently phenotyped for age, weight, gender, smoking status (via NLP), medical history (via NLP), pulmonary function, co-morbidities and confirmation of the diagnosis of asthma, which left 40,000 that met age and smoking criteria.  A subset of the asthma patients (n=10,000) with frequent ER visits and hospitalizations (non-responders) and an appropriate control group of asthma patients with no ER visits or hospitalizations had DNA samples collected via the Crimson Biospecimen Core and were genotyped to predict if there was a unique genomic signature (or outcome predictor) associated with the steroid non-response phenotype.  The total cost of the project was $200,000.