Grants to Enable Diagnostic Excellence
Utility of Predictive Systems to identify Inpatient Diagnostic Errors: The UPSIDE Study
Principal Investigator: Andrew Auerbach, University of California, San Francisco
- To determine the incidence of diagnostic errors among patients who die in hospital or are transferred to the intensive care unit 2 days or more after admission to a general medicine service through a structured, standardized adjudication process of patient records.
- To combine adjudication data with data from Vizient to determine which specific factors contribute to risks for diagnostic errors, and to use risk estimates to calculate incidence and impact of factors contributing to those errors.
- To create machine learning models that can be used to retrospectively identify patients in whom a diagnostic error was likely to have taken place.
Answering the call to engage patients and families in the diagnostic process: A new patient-centered approach using health information transparency to identify diagnostic breakdowns in ambulatory care
Principal Investigator: Sigal Bell, Beth Israel Deaconess Medical Center
- (a) To establish a new patient-centered framework co-designed with patients/families and care partners to measure and categorize patient-reported diagnostic breakdowns and (b) apply this new analytic tool to establish the incidence, types, and contributing factors to patient-reported diagnostic breakdowns in ambulatory care using two large and unique existing databases.
- To develop and implement a new electronic health record portal-based method enabling chronically ill patients and their families to (a) contribute to the visit note and diagnostic process and (b) identify and report diagnostic breakdowns using existing electronic health record data.
- To assess the use and impact of this method on the diagnostic process measuring safety (incidence/types of patient-reported diagnostic breakdowns among chronically ill patients), implementation, and stakeholder experience outcomes.
An expert-guided machine-learning approach to estimate the incidence, risk and harms associated with diagnostic delays for infectious diseases
Principal Investigator: Philip Polgreen, University of Iowa
- To determine the incidence of diagnostic delays for a wide range of infectious diseases.
- To identify the risk factors associated with diagnostic delays for infectious diseases that are frequently delayed or have serious outcomes.
- To estimate the impact of diagnostic delays in terms of healthcare costs and mortality.
Application of a Machine Learning to Enhance e-Triggers to Detect and Learn from Diagnostic Safety Events
Principal Investigator: Hardeep Singh, Baylor College of Medicine
- To develop, refine, test, and apply the Safer Dx Trigger Tools Framework to enable detection, measurement, and learning from diagnostic errors in diverse emergency department settings and calculate the frequency of diagnostic errors in the emergency department based on these e-triggers and describe the burden of preventable diagnostic harm.
- To explore machine learning techniques that yield robust, accurate models to predict diagnostic errors using electronic health record-enriched data derived from expert-labeled patient records containing diagnostic errors (from project aim 1, above).