Patient Flow E-Newsletter
Volume 3, Issue 4
Thursday, December 21, 2006
Success in the emergency department (ED) has always been viewed as saving lives. Over the last decade, however, other measures of success have become important. Hospital leaders have increasingly looked to ED managers to ensure that timely, efficient, cost-effective and high-quality care is provided, and that patients are satisfied. Now, after years of discussion, performance measures are becoming commonplace, and in some cases, mandatory.
Increasingly, we're not just competing with ourselves to improve. For the first time ever, we have computer systems and databases in place which make cross-institutional comparisons possible. More and more, we're seeing benchmarking from institution to institution as a way for hospital leaders to gauge how their organization stacks up among its competitors locally, and its peers nationwide. Most importantly, these comparisons are starting to have real effects on a hospital's reputation, market share and even reimbursement. Thus, while the move towards benchmarking performance measures is a positive change, we need to be cautious about how we use the data.
For those of us on the front lines of emergency medicine, all of these comparisons should raise a red flag. Since the current performance measurement process between hospitals does not reliably compare 'apples to apples,' it is quite likely that some of this benchmarking will result in flawed conclusions and clouded decision making. The source of this inaccuracy lies with the use of inconsistent standards, data definitions and methods of measurement.
Take arrival times as an example. One institution may define 'arrival time' as the moment a patient walks through the ED's door, while its competitor within the market may define it as the point when a patient is first seen by a triage nurse. Yet another may start the clock of 'arrival time' from the moment when a patient's name and other data are entered into the hospital's computer system.
To address these types of inconsistencies, the Emergency Department Benchmarking Alliance (EDBA) facilitated a gathering of some of the country's top leaders in this area. We were fortunate to have with us representatives from the American College of Emergency Physicians, Emergency Nurses Association, ED Practice Management Association, Institute for Healthcare Improvement, Urgent Matters, and more. Our group met with the following goals:
The group settled on definitions for about two dozen concepts related to ED performance, from active acuity level and daily boarding hours, to physician disposition time and decision to transfer. Many of the concepts were related to time and interval measurements, including door to doctor time, laboratory turnaround time, conversion time, discharge time, ED length of stay and others. We also addressed a number of categorical or proportional variables, such as daily census, percentage of Patients Who Left Without Treatment and others.
We knew from the beginning, however, that it wouldn't be as simple as just determining the standard definitions. We recognized that you can't implement the same standards for every hospital or medical center in the country without regard to its size, patient volume and community setting.
For example, very large EDs (e.g. >65,000 annual visits) commonly have a 'left without receiving treatment (LWOT)' rate of three or four percent. While not ideal, this rate is more or less 'normal.' By contrast, most small community hospital ED's (e.g. 10-15,000 visits) have a much lower LWOT rate, often less than one percent. In a smaller ED, a LWOT rate of three to four percent would be quite unusual and represent a serious concern worthy of special attention.
After much deliberation, we settled on a cohort scheme that breaks hospitals into one of four categories based on annual ED volumes: Less than 10,000; 10,000 - 29,999; 30,000 - 49,999; and 50,000 or more visits. Second, an acuity function is applied that takes into account an ED's trauma designation, admission rate and whether the hospital has transplant services.
The next step is generating acceptance and adoption nationwide, but the key is to start out slowly. While most of the country's largest institutions have established measurements and could immediately begin adopting the standards, it may be challenging for some smaller institutions that do not currently have the mechanisms in place to collect and measure this data.
Hospital and emergency medicine leaders at these smaller institutions should keep in mind that they don't need to implement all of these measures at once. They can start with arrivals and length of stay. Once they have implemented measurements for these two, then they can turn to looking at things like radiology, lab and admission turnaround times.
Ultimately, we hope to see a general acceptance of these performance measures and the cohort scheme, especially by the American Hospital Association, CMS and other regulatory or certifying bodies. If that occurs, we could legitimately create very large, sophisticated databases across institutions nationwide. These databases could be used for large-scale operations management research, and the lessons learned could help all of us to manage the very significant volume and quality challenges facing us today.
Ideally, when my colleagues and I try to compare Christiana Care Health System's numbers to other large, urban teaching hospitals with Level 1 trauma centers, we could compare ourselves to 300 or more similar institutions across the country, all using common measurement standards. Only by comparing 'apples to apples,' so to speak can we truly gauge our success in the ED and determine how to best improve upon our performance for our patients.
Charles Reese, M.D.
Chairman, Department of Medicine
Christiana Care Health System
Wilmington, DE
Chair, Emergency Department Benchmarking Alliance (EDBA)
