319 Variability in Patient Care Process Measures Across 7 Massachusetts Academic Emergency Departments

Abstract

Listen

Translate article

Study Objectives: To identify areas of variability and consistency in clinical operations and practice parameters across academic emergency departments (EDs). Methods: A retrospective cross-sectional descriptive review of Massachusetts EDs where emergency medicine residents spend 1 month of required training time. Of the 13 eligible, 7 became participants by identifying a site PI, obtaining local institutional review board approval, and collecting the requested data. Seven domains of data were collected for 2011 including: patients, care space and disposition; clinical staffing; triage type; lab and radiology testing; process improvement trials; documentation and billing; acute coronary syndrome and congestive heart failure clinical practice. We have reported averages and 1 standard of deviation to describe central tendency and variability. Results: Average ED daily volume (177 +/− 79) and space (3.4 +/− 1.4 patients per sq ft per yr) are quite variable across facilities. Attending physician staffing has low variability (2.5 +/− 0.5 hrs per day) while nursing staffing is more variable (1.9 +/−0.8 nursing hrs per patients seen). When PAs and NPs are included with residents as physician extenders, affiliate site attending physicians with their extenders see fewer patients per hour than those at primary residency sites (0.8 patients per hr +/− 0.2 versus 1.5 +/− 0.4). However, affiliate site EDs bill a higher percentage of level 5 APC charts (36%) compared to the primary sites (24%), which are also the higher volume receiving centers for high acuity patients. Most EDs use some degree of RN triage (87%), but 43% have a combination model using some degree of direct to room or physician triage. EDs with observation units did not have shorter ED length of stay (LOS) for admitted patients as compared to those without (6.5 +/−0.8 hrs versus 5.0, +/−0.5hrs). ED LOS for discharged patients demonstrates relatively low variability (3.9 +/− 0.5 hrs). In the past 10 years, 85.7% of EDs have worked on patient flow, triage and inpt admission processes. 71.4% have addressed care-space use, fast track areas, lab testing services, and consultations. Only 57.1% have worked on radiology testing services. The greatest variability in payer mix occurs amongst private insurance (28.9%, +/−25.6) and self-pay patients (25.5%, +/−22.3), with much less variability amongst Medicare (24.2% +/−11.7) and Medicaid (21.5% +/−12.0) patients. All EDs use either a Troponin T or an ultrasensitive Troponin I. The average interval for 2 troponin assays as part of an acute coronary syndrome evaluation was 6hrs. All discharged a maximum of 10% of patients with a primary diagnosis of congestive heart failure from the ED. Conclusion: As academic EDs cope with increasing volume and patient complexity, a variety of innovative approaches have been taken to improve throughput and care quality. However, throughput to discharge seems relatively homogeneous. Larger EDs are managing with fewer resources and a poorer income generation scenario. The area of least process improvement innovation was radiology testing. Regarding cardiovascular care, there is potential practice and throughput improvement for patients being evaluated for acute coronary syndrome with a more evidence-based use of cardiac enzymes by using intervals shorter than 6 hours, and with congestive heart failure patients where less than 10% are discharged.