Health policy/brief research reportExogenous Predictors of National Performance Measures for Emergency Department Crowding
Introduction
According to the Institute of Medicine, emergency department (ED) crowding is a major public health problem1 and is associated with patient safety issues, poorer quality, and negative outcomes.2 The National Quality Forum has endorsed several crowding measures, including median ED length of stay (separated by admitted and discharged patients), waiting times, decision to admit to admission times, and rates of left without being seen.3 Current plans by the Centers for Medicare & Medicaid Services are to offer higher payments for hospitals who report some of these measures in 2013.4 Under current specifications, there are no recommended standards for any of the measures and no plan to stratify measures by ED volume or case mix.3, 4 ED performance will likely be measured and compared without consideration of factors out of the control of an ED (ie, exogenous variables).
Basic assumptions of performance measurement are that hospitals can be compared directly through objective metrics and that the information should be useful to consumers to compare quality and useful to payers to reward better performance. However, this is challenging because differences between patient populations affect the performance measures themselves.5 Although some differences may be difficult to measure, one solution is risk adjustment, which accounts for observable differences in risk factors (eg, illness severity). Extending this logic to ED throughput, factors outside of a hospital's control may affect performance. For example, a rural ED with 8 beds and 8,000 annual visits is operationally different from an inner-city teaching hospital with 60 beds and 120,000 annual visits. Maintaining the same performance (ie, throughput and system responsiveness) at the teaching facility could be more challenging because of system complexity and demand variability. Therefore, one approach to benchmarking is to compare “like” facilities by stratifying data with exogenous variables such as case mix or annual volume.
We used the National Hospital Ambulatory Medical Care Survey (NHAMCS) to assess the association between exogenous ED-level variables and performance on the National Quality Forum–approved measures. Our goals were to find exogenous ED-level factors to help with benchmarking and public reporting. We sought to create a simple stratification system for ED throughput measures that could be easily understandable by consumers.
Section snippets
Study Design and Setting
We conducted a study of the 2008 and 2009 NHAMCS to explore facility-level predictors of 4 measures of ED throughput proposed by the National Quality Forum. These include ED length of visit for admitted and discharged patients, waiting time, and rate of left without being seen. The decision to admit to departure from the ED (ie, boarding) was excluded because data on it were not collected in 2008 and greater than 20% of the data were missing in 2009.
Data Collection and Processing
Data came from the 2008 to 2009 NHAMCS, which
Results
In 2008 to 2009, median waiting time was 35 minutes (95% confidence interval [CI] 26 to 43 minutes), median length of visit for patients treated but not admitted was 131 minutes (95% CI 121 to 142 minutes), median length of visit for patients admitted was 244 minutes (95% CI 218 to 270 minutes), and rate of left without being seen was 1.3% (95% CI 0.9% to 1.8%) (Table 1). The majority of exogenous variables demonstrated significant bivariate associations with performance measures. For measures
Limitations
We analyzed only 2 years' worth of data. Using different years may yield different results because NHAMCS studies have demonstrated lower performance over time.10, 11 It is also possible that our a priori chosen variables that were available in NHAMCS could have excluded other potentially useful variables. It is also possible that other variables unavailable in NHAMCS (eg, trauma center status, pediatric hospitals) could be helpful in risk stratification. We were unable to study the time from
Discussion
Several exogenous variables were associated with large differences in performance across 4 ED crowding measures approved by National Quality Forum. Although ED visit volume was the strongest exogenous predictor of performance, several other variables were also important. Therefore, developing a simple, sensible stratification system to compare hospital performance would not adequately characterize the myriad exogenous factors that affect performance. We had initially hypothesized that ED volume
References (13)
The future of emergency care in the United States health system
Acad Emerg Med
(2006)- et al.
The effect of emergency department crowding on clinically oriented outcomes
Acad Emerg Med
(2009) National Quality Forum Endorses Measures to Address Care Coordination and Efficiency in Hospital Emergency Departments
CMS-1504-FC hospital outpatient prospective payment system and CY 2011 payment rates. November 2, 2010. Hospital outpatient regulations and notices
The risks of risk adjustment
JAMA
(1997)- et al.
Plan and operation of the National Hospital Ambulatory Medical SurveySeries 1: programs and collection procedures
Vital Health Stat 1
(1994)
Cited by (50)
Fair Play: Application of Normalized Scoring to Emergency Department Throughput Quality Measures in a National Registry
2021, Annals of Emergency MedicineFreestanding emergency departments in Texas do not alleviate congestion in hospital-based emergency departments
2020, American Journal of Emergency MedicineAssociation between adopting emergency department crowding interventions and emergency departments' core performance measures
2020, American Journal of Emergency MedicineCitation Excerpt :Using these questions, the primary independent variables are whether or not the hospital adopted each of 20 ED crowding interventions. It is likely that factors other than an ED intervention may affect ED performance measures [25,27,28]. Therefore, we controlled for patient-level (e.g., race), hospital-level (e.g., size), and temporal (e.g., time of ED admission) confounders.
Predictors of Delayed Emergency Department Throughput Among Blunt Trauma Patients
2020, Journal of Surgical ResearchEmergency Department Crowding and Time at the Bedside: A Wearable Technology Feasibility Study
2018, Journal of Emergency NursingHospital and Community Characteristics Associated With Pediatric Direct Admission to Hospital
2018, Academic Pediatrics
Please see page 294 for the Editor's Capsule Summary of this article.
Supervising editor: Donald M. Yealy, MD
Author contributions: JMP and SLD conceived this study and drafted the article. SLD and TH conducted the statistical analysis. TH provided critical revisions to the article. JMP takes responsibility for the paper as a whole.
Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist.
The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
A podcast for this article is available at www.annemergmed.com.
Publication date: Available online May 23, 2012.