April 13, 2017

Research

Current Major Grants & Projects | Current Collaborations | Completed Grants & Collaborations

Current Major Grants and Projects

Identifying Adverse Drug Events in Ambulatory Care Clinical Notes
IBM-BWH Collaborative Research Project
Zhou L (PI/Project Lead) 11/1/2020–10/31/2022
The goal is to investigate, characterize and annotate adverse drug events (ADEs) documented in ambulatory care clinical notes and to assess the ability of NLP and machine learning to identify ADEs.
Investigating sex differences in subjective cognitive decline across patients captured in electronic health records
The Program for Interdisciplinary Neuroscience (PIN) and the Women’s Brain Initiative (WBI), BWH
Wang L (PI) 03/1/2020–02/28/2021
The goal is to apply informatics techniques to clinical data in the EHR to identify patients with subjective cognitive decline (SCD), and to investigate gender differences in clinical characteristics and outcomes across patients with SCD.
Clinical Informatics to Advance Epidemiology and Pharmacogenetics of Serious Cutaneous Adverse Drug Reactions
NIH-NIAID — 1R01AI150295
Zhou L (PI) 09/16/2019–08/31/2023
The goal is to leverage advanced informatics and longitudinal EHR data to identify severe cutaneous adverse reactions (SCAR) cases to study genetic and epidemiological risk factors for SCAR related to commonly used drugs such as antibiotics.
Improving Allergy Documentation and Clinical Decision Support in the EHR
Agency for Healthcare Research and Quality (AHRQ) — R01HS025375
Zhou L (PI) 05/07/2018–04/30/2022
The goal of this study is to improve allergy documentation by developing a comprehensive and interactive value set for reaction mechanism and type, developing an innovative allergy reconciliation module within the EHR, and redesigning drug allergy alerting mechanisms using big data analytics and up-to-date evidence.

Current Collaborations

Evaluating and Improving the Accuracy of ICD-Coded Hospital Data Systems in Estimating the Incidence of Nonfatal Firearm Injuries by Intent Type
National Collaborative on Gun Violence Research (NCGVR)
Miller, M (Northeastern University); Goralnick, E 09/1/2020–08/31/2022
Developing Scalable Algorithms to Incorporate Unstructured Electronic Health Records for Causal Inference based on Real-world Data
NIH/NLM 1R01LM013204
Lin J (PI) 06/1/2020–03/31/2025
Risk Factors for Psychosis and Mania with Prescription Amphetamine Use
NIH/NIMH 1R01MH122427
Moran L (PI; McLean Hospital) 06/10/2020–04/30/2024
Assessing the effectiveness of oral anticoagulants in patients with atrial fibrillation at high risk of underutilization due to dementia, recurrent falls, or poor anticoagulation quality
NIH-NIA R01AG063381
Lin J (PI) 05/01/2019–04/30/2021
From Sepsis Prognosis Prediction to Tailored Clinical Practice: A Deep Learning Approach
CRICO/RMF
Dykes P (PI) 07/01/2019–06/30/2021
Promoting follow-up of abnormal cancer screening tests using population-based systems to support stepped care multilevel intervention
NIH-NCI U01CA225451
Haas J (PI; Mass General Hospital) 04/01/2018–09/30/2022
Communicating Narrative Concerns Entered by RNs (CONCERN): Clinical Decision Support Communication for Risky Patient States
NINR — 1R01NR016941
Collins S (PI; Columbia University) 04/06/2017–03/31/2022
Data in the nurses’ notes could be used by computers and Apps to tell the care team that a patient is not doing well and that they should act more quickly. This project will build an App that makes it easier for the care team to see and understand that data and act quickly to save patients.

Completed Grants and Collaborations

Finding Needles in the Haystack: Using Clinical Data and Machine Learning to Identify Patients for Early Palliative Care Interventions
Partners Healthcare Innovation
Zhou L (PI) 05/01/2019–04/30/2020
The goal of this study is to explore advanced machine learning methods to predict short-term mortality using a large volume of detailed electronic health records (EHRs) and then incorporate those tools into ongoing palliative care projects at BWH.
Computerized Support for Malpractice Auditing and Coding
CRICO/RMF
Zhou L (PI) 07/01/2018–12/30/2019
Similar-cAses Finder for Risk Reduction – the SAFRR system
CRICO/RMF
Zhou L (PI) 07/01/2018–06/30/2020
Assessing Allergy Safety During Electronic Health Record Transitions
CRICO/RMF
Kimberly Blumenthal (PI; Mass General Hospital) 07/01/2018–06/30/2020
NLP to Improve Accuracy and Quality of Dictated Medical Documents
Agency for Healthcare Research and Quality (AHRQ) — R01HS024264
Zhou L (PI) 09/30/2015–09/29/2019
The goal is to study the natural of errors in clinical documents generated by speech recognition (SR) technology during the traditional transcription services and the front-end provider documentation, and develop innovative methods based on NLP to detect and correct these errors. We also investigate provider adoption, use, productivity, and satisfaction of SR technology across different clinical settings. For this project, Dr. Zhou’s research team is collaborating with researchers from University of Colorado and Brandeis University.
Encoding and Processing Patient Allergy Information in EHRs
Agency for Healthcare Research and Quality (AHRQ) — R01HS022728
Zhou L (PI) 09/30/2013–9/30/2018
The goal of this study is to 1) study allergy documentation and clinical decision support in a large EHR data repository over two decades, 2) conduct detailed analyses on standard terminologies and allergy data to build a comprehensive knowledge base for representing allergy information and 3) design, develop, and evaluate a natural language processing (NLP) module for extracting and encoding free-text allergy information, and to measure the feasibility and efficiency of the proposed NLP system for the new process of allergy reconciliation.
Novel electronic health record phenotyping of LGBTQ intersectional identities and associated health disparities using NLP and machine learning approaches
Harvard Catalyst Health Disparities Research Program and President and Fellows of Harvard College — 2016D010012
Zhou L (PI) 07/01/2017–06/30/2018
The goal of this study is to address lesbian, gay, bisexual, transgender, and queer (LGBTQ) health disparities through clinical and translational research. We aim to describe and measure the extent to which sexual orientation and gender identity was documented in the EHR among BWH primary care patients and to evaluate differences in healthcare utilization, morbidity, and mortality between and among LGBTQ- and non-LGBTQ-identifying patients.
NLP to Identify and Rank Clinically Relevant Information for EHRs in the Emergency Department
Agency for Healthcare Research and Quality (AHRQ) — 1R21HS024541
Goss FR (PI) 09/01/2016–8/31/2018
The goal of this study is to develop and evaluate a sophisticated NLP tool to automatically identify and rank clinically relevant information from EHRs that providers rely upon to make medical decisions for their patients.
Stratified Regression Models for Case-Only Studies
Patient-Centered Outcomes Research Institute (PCORI) — HSRP20163120
Mittleman M (PI) 2016–2019
The goal is to develop analytic techniques to 1) estimate measures of absolute effect in case-only studies including number needed to treat or harm, 2) to evaluate the presence of heterogeneity of treatment effects on the additive scale, and 3) to address time-varying confounding in case-only studies using weights for the probability of exposure.
Automated Knowledge Extraction from Free-text Malpractice Claims Data to Enhance Coding & Analytics
Controlled Risk Insurance Company and Risk Management Foundation (CRICO)
Zhou L (PI) 07/1/2015–06/30/2017
The goal is to apply innovative health information technology, including NLP and machine learning, to enable faster, more accurate, and better quality data coding and analytics, which will likely result in quicker feedback to organizational stakeholders and practitioners. We have built a prototype application that can conduct automated coding and assist with auditing tasks.
Big Data in Healthcare: the Impact on Healthcare Quality, Cost and Access in China
Harvard-China Fund
Zhou L (Co-Director) 01/01/2016–12/31/2016
Our goal was to organize a big data symposium to bring together researchers, educators and practitioners from different fields and promote strong connections between academia, industry, government and the nonprofit sector to facilitate the transformation of data from knowledge to action.
Extracting Information from Unstructured Clinical Data for Enhancing Risk-identification and Stratification
Partners Siemens Research Council (PSRC)
Zhou L (PI) 05/01/2014–07/30/2015
The goal was to use free-text data to make inferences about the likelihood of readmission of patients to reduce readmissions by enabling providers to target coordinated care to the most vulnerable patients.
Integration of an NLP-based Application to Support Medication Management
Agency for Healthcare Research and Quality (AHRQ) — 1R21HS021544-01
Zhou L (PI) 07/01/2012–06/30/2014
The goal was to develop novel methods and a system using NLP and other technologies to facilitate the medication reconciliation process in the ambulatory setting.
Identifying and Integrating Family History Information from EHRs Using an NLP-based Approach
Partners Siemens Research Council (PSRC)
Zhou L (PI) 4/1/2012–3/31/2013
The goal was to study family history information in EHRs, develop a NLP tool to extract this information from clinical documents, investigate standards for representing family histories, and identify gaps.
Novel Methods for Improving Concepts for Structured Clinical Documentation
Partners Siemens Research Council (PSRC)
Zhou L (PI) 10/1/2011–3/31/2013
The goal of this study was to develop a NLP-based approach to support clinical documentation requirements and best practices by maximizing concept and template adoption among clinician users, reducing duplicates, and linking to standards.
Advancing Clinical Decision Support
HHS Contact # HHSP23320095649WC; Task Order HHSP23337009T ONC ARRA Contract
Bell D and Middleton B (Project Directors) 5/1/2010–11/30/2012
The overall goal of this project was to accelerate the successful implementation and effective use of CDS interventions to facilitate evidence based clinical practice and meaningful use of health IT. It is funded under the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009. Dr. Zhou led a subtask which aimed to outline parameters, structural elements, formats and standards necessary to support the core elements of a practical, scalable CDS sharing service.
Automatic Mapping of Local Medication Terms to RxNorm Concepts
Partners Siemens Research Council (PSRC)
Zhou L (PI) 10/1/2010–9/30/2011
The goal was to analyze the structure and semantics of two terminologies, Partners Master Drug Dictionary (MDD) and RxNorm, and develop automated methods and a tool to map MDD to RxNorm.
Improving Medication Lists Using Temporal Reasoning and Clinical Texts
Agency for Healthcare Research and Quality (AHRQ) — 1R03HS018288-01
Zhou L (PI) 10/1/2009–9/30/2011
The goal was to develop and test methods and tools in the field of NLP and temporal reasoning to facilitate the use of clinical narratives to improve the accuracy and completeness of medication lists.
Using Temporal Reasoning to Augment Clinical Decision Support for Medication Administration
Partners Siemens Research Council (PSRC)
Zhou L (PI) 10/1/2009–8/31/2011
The goal was to use temporal information to enhance computerized decision support and improving patient safety using electronic Medication Administration Records (eMAR).