Skip to main content

We’ve updated our Terms & Conditions and Privacy Policy. By using this site, you agree to these terms.

Precision Medicine

Title: Developing and Validating EHR-Integrated Readmission Risk Prediction Models for Hospitalized Patients with Diabetes

Site PI: H. Lester Kirchner, PhD (PI: Daniel Rubin, Temple University)
Funder: NIDDK
Grant Number: 5R01DK122073

Dates: 08/21/2020 – 06/30/2025

Hospitalized patients with diabetes are at higher risk of readmission within 30 days than patients without diabetes, and >1 million readmissions occur among diabetes patients in the US annually. Certain interventions can reduce readmission risk, but applying these interventions widely is cost prohibitive. One approach for improving the efficiency of interventions that reduce readmission risk is to target high-risk patients. We previously published a model, the Diabetes Early Readmission Risk Indicator (DERRI), that predicts the risk of all-cause 30-day readmission of patients with diabetes. The DERRI, however, has modest predictive accuracy (C-statistic 0.63- 0.69), and requires manual data input. Recently, we demonstrated that adding variables to the DERRI substantially improves predictive accuracy (DERRIplus, C-statistic 0.82). However, using this larger model to predict readmission risk based on manual input of data would be too labor intensive for clinical settings. Indeed, most readmission risk prediction models are limited by the trade-off between accuracy and ease of use; lack of translation to a tool that integrates with clinical workflow; modest accuracy; lack of validation; and dependence on data only available after hospital discharge. The objectives of the current proposal are: 1) To develop more accurate all-cause unplanned 30-day readmission risk prediction models using electronic health record (EHR) data of patients with diabetes (eDERRI); 2) To translate the models to an automated, EHR-based tool that predicts % readmission risk of hospitalized patients; and 3) To prospectively validate the eDERRI models and tool. To develop the models, we will leverage data from the PaTH Clinical Data Research Network (CDRN), a multi-center, 40-plus hospital member of the National Patient-Centered Clinical Research Network (PCORnet).


Title: Rational Integration of Clinical Sequencing (RISE)

MPI: Josh Peterson, Vanderbilt University; Jing Hao, Geisinger; David Veenstra, University of Washington
Funder: NIH/NHGRI
Grant number: 5 R01 HG 009694
Dates: 8/1/2017 – 7/31/2021 (NCE to 07/31/2022)

The advent of clinical genome sequencing to identify patients at risk for serious diseases and to tailor treatments promises to greatly improve health outcomes and provide a foundation for the delivery of Precision Medicine. However, even as laboratory methods to perform sequencing become highly efficient, uncertainty around the optimal breadth and economic value of sequencing as well as ambiguity around which individuals should be tested presents a critical barrier to wider use. As we rapidly approach an era of inexpensive sequencing, new approaches to quantify and optimize the economic and clinical value of genome-tailored care are needed. For the Rational Integration of Sequencing (RISE) project, we propose to develop a Discrete Event Simulation (DES) to estimate the average clinical efficacy and cost-effectiveness of prospectively acquiring sequence data across a diverse patient population. The simulation will leverage literature-based estimates of clinical outcome rates, costs, and utilities combined with clinical exome and pharmocogenomic implementation program data describing how results are returned and reacted to within clinical care. The first Aim will develop a conceptual framework and computational infrastructure to understand the drivers of economic value in genomic screening. The second Aim will externally validate the RISE model using real-world use data. The third Aim will assess the cost effectiveness of genomic screening scenarios, identify key drivers of value and inform research priorities in genomic screening.

Published papers:

  • Laney K. Jones, Nan Chen, Dina Hassen, Tracey Klinger, Megan McMinn, Dustin N. Hartzel, David Veenstra, Scott Spencer, Susan R. Snyder, Josh F. Peterson, Victoria Schlieder, Marc S. Williams, Jing Hao. Impact of a population genomic screening program on health behaviors for familial hypercholesterolemia risk. Circulation: Genomic and Precision Medicine, In press
  • Guzauskas GF, Jiang S, Garbett S, Zhou Z, Spencer SJ, Snyder SR, Graves JA, Williams MS, Hao J, Peterson JF, Veenstra DL. Cost-effectiveness of population-wide genomic screening for Lynch syndrome in the United States. Genet Med. 2022 May;24(5):1017-1026. doi: 10.1016/j.gim.2022.01.017. Epub 2022 Feb 25. PMID: 35227606. https://pubmed.ncbi.nlm.nih.gov/35227606/
  • Guzauskas GF, Garbett S, Zhou Z, Spencer SJ, Smith HS, Hao J, Hassen D, Snyder SR, Graves JA, Peterson JF, Williams MS, Veenstra DL. Cost-effectiveness of Population-Wide Genomic Screening for Hereditary Breast and Ovarian Cancer in the United States. JAMA Netw Open. 2020 Oct 1;3(10):e2022874. doi: 10.1001/jamanetworkopen.2020.22874. PMID: 33119106; PMCID: PMC7596578. https://pubmed.ncbi.nlm.nih.gov/33119106/
  • Hao J, Hassen D, Manickam K, Murray MF, Hartzel DN, Hu Y, Liu K, Rahm AK, Williams MS, Lazzeri A, Buchanan A, Sturm A, Snyder SR. Healthcare Utilization and Costs after Receiving a Positive BRCA1/2 Result from a Genomic Screening Program. J Pers Med. 2020 Feb 3;10(1):7. doi: 10.3390/jpm10010007. PMID: 32028596; PMCID: PMC7151600. https://pubmed.ncbi.nlm.nih.gov/32028596/
  • Veenstra DL, Guzauskas G, Peterson J, Hassen DA, Snyder S, Hao J, Williams M. Cost-effectiveness of population genomic screening. Genet Med. 2019 Dec;21(12):2840-2841. doi: 10.1038/s41436-019-0580-4. PMID: 31303645. https://pubmed.ncbi.nlm.nih.gov/31303645/

  • Title: Rational Integration of Polygenic Risk Scores (RIPS)

    MPI: Josh Peterson, Vanderbilt University; Jing Hao, Geisinger; David Veenstra, University of Washington
    Funder: NIH/NHGRI
    Grant number: 1 R01 HG012262-01
    Dates: 03/10/2021 – 02/28/2027

    There has been extraordinary growth in new techniques to predict common, complex disease based on polygenic risk scores (PRS). Without an understanding grounded in evidence, it is unlikely that the clinical use of PRS will propagate from highly specialized applications and environments to become adopted more broadly and provide greater benefit to the US population. Critical challenges include: 1) understanding the impact of clinical PRS for multiple diseases on long-term patient outcomes, 2) identifying risk thresholds for return of results that optimize patient outcomes and provide cost-effective care, 3) understanding how PRS performance across diverse populations may affect existing disparities and subsequent patient outcomes. We propose to address these challenges using decision analytic modeling and by building on our extensive work in this area to create a novel framework capable of assessing PRSs in the context of monogenic and clinical risks. We have already created clinical-economic models to project lifetime clinical impact and cost-effectiveness for population-level genomic screening with return of monogenic disease risks associated with three CDC Tier 1 conditions: hereditary breast and ovarian cancer, Lynch syndrome, and familial hyperlipidemiaAs part of this proposal, titled Rational Integration of Polygenic Risk Scores (RIPS), we will create models to assess the clinical outcomes and economic value of population screening using PRS in realworld settings and applied to large and diverse populations. The Aims of the proposal include 1) to evaluate published and real-world evidence on the clinical value of adding PRS to inform comprehensive genomic risk assessment; 2) to understand the impact of PRS performance and return risk thresholds on incremental clinical benefit and cost effectiveness for breast cancer, atherosclerotic cardiovascular disease, and colorectal cancer, and 3) to develop research priorities for the equitable development and implementation of PRS across underserved and underrepresented populations.