Nearly 50 million people have surgery each year in the United States, approximately one million develop serious complications, and over 150,000 die within 30 days. Health-care-associated infections (HAIs), the most common complications in surgical patients, place a huge burden on patients, their families, the healthcare system, and society. As there is a clinical and economic imperative to decrease surgical complications, efficient and accurate reporting of HAIs is paramount. Unfortunately, the current model of HAI reporting routinely involves data collection that is inaccurate, burdensome and/or expensive. Achieving accurate and efficient automation of data collection for performance measures, including HAIs, will not only support better local surgical quality improvement, but also will accelerate the transition from process to outcome-based measures suitable for value-based purchasing, new models of physician payment, and public reporting. This proposal attempts to address this challenge by leveraging currently available informatics technologies to automate the data collection from electronic health records for HAIs and other surgical performance measures. We hypothesize that, using advanced informatics methods combined with an iterative process that is informed by multidisciplinary input, data from electronic health records can be mined to develop scientifically valid, cost- effective outcome measures which will compare favorably with those derived from ACS NSQIP data, a clinical registry recognized for its rigorous accuracy. We will test and optimize the algorithms at five unique hospitals in four health systems and two electronic health record vendors to begin to determine the scalability of this approach. The American College of Surgeons (ACS), the primary site, has the content expertise, clinical registry and national influence necessary to drive change in the approach to performance measurement in surgery. We have assembled an accomplished team of national leaders in surgical outcome measurement and data (Clifford Ko, MD, MHS), quality improvement and implementation science (Elizabeth Wick, MD, Peter Pronovost, MD, PhD), informatics (Genevieve Melton, MD, Ph.D., Quintiles), and hospital epidemiology (Trish Perl, MD, MPH). Successful completion this work will impact the field in three main ways: 1) make available a strategy to automate five NQF-endorsed performance measures from the EHR; 2) provide essential knowledge about what is required at the hospital level to automate outcome measurement and the scalability of this approach to multiple hospitals, and; 3) accelerate transition from the use of process to outcome measures for value-based purchasing and physician payment models. The proposed research is novel and timely, and is the next important step on the path to making valid surgical performance measures available to all hospitals and clinicians, and ultimately the public.
Progress has been slow in reducing the number of preventable complications and healthcare-associated infections after surgery. Efficient and accurate reporting of surgical performance and healthcare-associated infections is essential for improvement, but the current reporting model involves data collection that is inaccurate, delayed, burdensome and/or expensive. The proposed study seeks to apply new advanced informatics technology to the electronic health record in order to establish a novel model for timely, accurate, and efficient automated surgical performance and healthcare-associated infection measures that can be used nationally to improve quality and save patient lives.
|Skube, Steven J; Lindemann, Elizabeth A; Arsoniadis, Elliot G et al. (2018) Characterizing Functional Health Status of Surgical Patients in Clinical Notes. AMIA Jt Summits Transl Sci Proc 2017:379-388|
|Hu, Zhen; Melton, Genevieve B; Arsoniadis, Elliot G et al. (2017) Strategies for handling missing clinical data for automated surgical site infection detection from the electronic health record. J Biomed Inform 68:112-120|
|Skube, Steven J; Hu, Zhen; Arsoniadis, Elliot G et al. (2017) Characterizing Surgical Site Infection Signals in Clinical Notes. Stud Health Technol Inform 245:955-959|