Leveraging Big Data to Balance New Key Performance Indicators in Emergency Physician Management Networks

Abstract

Managing emergency physicians is a complex task and has increasingly intensified with the recent consolidation of many emergency departments (EDs). Large‐scale physician groups are facing challenges in resource deployment and performance evaluation. To objectively evaluate physicians across facilities, we leverage big data from an emergency physician management network and propose data‐driven metrics using a large‐scale database consisting of 84 hospitals, 1,079 physicians, and 10,615,879 patient visits in 14 states over 600,000 clinical shifts from 2010 to 2014. To ensure physicians are fairly evaluated and compensated within diverse facilities, we propose an index system and use clustering to help identify factors which might impact physician performance. The proposed indices benchmark physicians from the perspectives of revenue potential, patient volume, patient complexity, and patient experience by controlling for exogenous factors at the facility level. We empirically show the volume and complexity indices are key elements of the revenue potential index, and use two‐stage least squares regression to relate volume and complexity and uncover their drivers. Revenue potential and patient experience are found to be positively correlated, which suggests productive physicians are often liked by their patients. Through implementing the proposed evaluation system, administrators can better manage and incentivize physicians and provide directions for performance improvement, while controlling for location idiosyncrasies. The proposed framework can also be adapted to non‐medical professional settings such as value chains, where employees often provide services in various profit‐ and cost‐centers.

Definitions and factors associated with emergency physician productivity: a scoping review

“…In our review, three included studies measured productivity using such indices [20][21][22]. A few published modelling studies have also developed similar productivity indices based on theoretical data [23][24][25]. Their use and implications in clinical practice remains to be studied.…”

Hospital Employee Performance Evaluation Based on Knowledge Map

“…Moreover, the performance evaluation scheme of medical staff established by them can reflect the investment of labor value and meet the requirements of hospital personnel management in the new medical reform. (Foster et al 2018) used the big data method to analyze the factors affecting the work performance of hospital doctors, including income potential, number of patients, disease complexity, patient experience, etc., and concluded that the number of patients and basic complexity are the key factors determining the income potential, and the income potential is also positively related to the patient experience. (Ider et al 2011) discussed the game between performance appraisal and nosocomial infection report in Mongolian hospitals, and found that hospital employees would minimize the number of nosocomial infection cases reported as much as possible due to the penalty measures in performance appraisal, so they suggested replacing the original index of nosocomial infection prevention and control measures with the original index of nosocomial infection rate.…”

OM Forum—Supply Chain Thinking in Healthcare: Lessons and Outlooks

“…The surge in data availability, in conjunction with growing computer power, has allowed healthcare analytics tools, such as AI and machine learning, to play an expanding role in the advancement of healthcare. For example, machine learning is now used to inform diagnosis (Miotto et al 2018), make predictions (Finlay 2018), develop prescriptive treatment algorithms (Champagne et al 2018, Jameson andLongo 2015a), reduce readmissions (Liu et al 2018, Queenan et al 2019, and objectively evaluate physicians (Foster et al 2018) (see Guha and Kumar (2018) for an overview of how big data analytics has been applied in the healthcare domain and for a roadmap of future research). Moreover, health information exchanges, facilitated by technologies like blockchain (Babich and Hilary 2019), may lead to more efficient hSCs by minimizing transaction costs and wastes (e.g., fraud, counterfeits).…”