Surgery

Biography

Biography: Fred W. Peyerl

Abstract

Electronic health record (EHR) databases represent an increasingly valuable resource for biomedical researchers seeking to identify best perioperative practices via retrospective analysis of real-world patient data. EHR databases provide a unique source of aggregated point-of-care data covering all aspects of the patient encounter, thus allowing researchers to investigate clinical and administrative outcomes during and subsequent to the surgical visit. In-depth comparative eff ectiveness studies comparing two or more interventions are facilitated by the availability of admission/discharge, diagnosis, procedure, laboratory, medication, and clinical assessment data. In addition, a wealth of patient demographic and facility data allows investigators to stratify patients into relevant risk groups. A range of validated statistical approaches support quantitative comparisons between interventions. Recently, the Boston Strategic Partners team used data from a large de-identifi ed US EHR database to conduct a retrospective comparative eff ectiveness study examining the impact of intravenous fl uid choice on postoperative outcomes in patients undergoing on-pump coronary artery bypass graft (CABG) and/or valve procedures. Outcomes examined included clinical complications such as pneumonia, acute kidney injury, and electrolyte imbalances, as well as key administrative outcomes such as readmissions following surgery. Th is study identified significant diff erences in outcome rates based on intravenousfluid choice. In light of these outcome diff erences, an economic model was subsequently developed to examine the economic implications of fl uid choice (5% albumin plus crystalloids versus crystalloids alone) in this patient population. Th e results of this comparative eff ectiveness study reveal a key application of EHR databases for the identification of best perioperative practices.