Welch, Shari J. MD
Nineteen influential members of the emergency medicine community with specific interests and expertise in performance improvement came together at the Performance Measures and Benchmarking Summit in February to craft a document that standardizes the vocabulary and terminology used in quality improvement. Last month, I reviewed the ED cohort scheme and general definitions for the emergency department, including arrival time, physician disposition time, and ED length of stay, to name just a few.
There are compelling reasons that emergency practitioners should standardize the language, terminology, and implementation of performance measures and benchmarking practices, not the least of which is regulatory burden. The Joint Commission on Accreditation of Healthcare Organizations is now pursuing clinical quality improvement data in the form of core measures. Any facility without an infrastructure to track these data risks its accreditation. The measures are likely to be under double scrutiny as the Centers for Medicare & Medicaid Services launches its Pay for Performance (P4P) initiative, which will reward hospitals that perform better along those same parameters. (A Comprehensive Review of Development and Testing for National Implementation of Hospital Core Measures. www.jcaho.org/pms, March 2005; Amer Med News. Sept. 5, 2005.)
The Joint Commission also just levied additional regulatory burdens in the form of the so-called “flow standards.” (Meeting the JCAHO Patient Flow Standard, Urgent Matters Regional Conference, Oct. 27, 2005.) If an ED wants to maintain its credentials and be reimbursed maximally, data tracking and following measures of quality are imperatives. It is essential that further regulatory requirements use parameters that come from within the specialty and that emergency medicine experts who understand the nuances of data collection and analysis lead these endeavors.
Standardization will make improvement efforts more meaningful and point the specialty in the right direction
The Performance Measures and Benchmarking Summit was charged with formulating definitions for basic ED operations, drafting a consensus statement on benchmarking in emergency medicine, developing a comprehensive set of benchmarks for patient flow and operations that also could be used for operational quality, forming a simple cohort scheme for categorizing EDs for benchmarking, and disseminating and publishing the results. To continue last month's list of definitions, let's start with performance and time measures for discharged patients, which include door-to-doctor time, the time difference in minutes between arrival time and MD/LIP (licensed independent practitioner) contact with the patient, and doctor-to-discharge time, the time difference in minutes between MD/LIP contact with the patient and discharge time. The ED LOS for discharged patients is the time difference in minutes between arrival time and discharge time.
For admitted patients, the door-to-doctor time is the time difference in minutes between arrival time and MD/LIP contact with the patient; the doctor-to-decision-to-admit time is the time difference in minutes between MD/LIP contact with the patient and the decision to admit is made. The decision to left ED time is the time difference in minutes between the decision to admit and the physical departure of the patient from the ED treatment area. The ED LOS for admitted patients is the time difference in minutes between arrival time and physical departure of the patient from the ED treatment area (sum of door-to-doctor time plus doctor-to-decision-to-admit time plus decision to left ED time). Daily boarding hours are calculated by the sum of patients calculated in the decision to left ED time minus 120 minutes for each boarder divided by 60 for all boarders in a 24-hour period.
For transferred patients, the door-to-doctor time is the difference in hours between arrival time and MD/LIP contact with the patient, and the doctor-to-decision-to-transfer time is the difference in hours between MD/LIP contact with the patient and the decision to transfer is made. The decision-to-transfer-to-transfer-accepted time is the difference in hours between the decision to transfer the patient and the acceptance of the transfer. The transfer-accepted-to-left-ED time is the difference in hours between the acceptance of transfer and the physical departure of the patient from the ED treatment area.
The ED LOS for transferred patients is the time difference in hours between arrival time and physical departure of the patient from the ED treatment area (sum of door-to-doctor time plus doctor-to-decision-to-transfer time plus the decision-to-transfer time-to-transfer-accepted time plus transfer accepted to left ED time).
To define the performance and proportion measures for patients who left before they were supposed to, the language needs to be tied to specific predictable events in every patient encounter. EMTALA has defined the medical screening exam (MSE) as a defining event in emergency care. The causes of a patient's unofficial departure and the actions taken vary and warrant monitoring. Our group recommends that the key performance indicators for patients leaving before the provider deemed treatment is complete should be referred to as:
Patients leaving before the medical screening exam refers to any patient who leaves the ED before initiation of the MSE. It is expressed as a rate of occurrences per 100 visits. Calculate the interval from when the patient is recognized as an encounter in the ED to the time that the patient departed or the medical screening exam was initiated.
Patients leaving after the medical screening exam refers to any patient who leaves the ED after his MSE, but before the provider documented treatment as complete. It is expressed as a rate of occurrences per 100 visits. Calculate the interval from the time of the initiation of the MSE to the time that the patient departed.
Patients leaving against medical advice are patients recognized by the institution and leaving after interaction with the ED staff but before the ED encounter has officially ended. This differs from patients leaving after the medical screening exam because it includes documentation of patient competence, discussion about risks and benefits, and completion or refusal to complete documentation confirming the intent to leave against the recommendation of medical care staff.
The definition of a complaint should be standardized to include all spontaneous concerns about service delivery in the ED, written or verbal, brought to the attention of ED leaders. Separate categories of service concerns should be those identified during a survey process or during the billing process. Complaints are typically counted as one complaint per communication, and are tracked in rates per 1000 ED visits.
The reasons, methods, and responses to diversion of patients from the ED are felt to be widely variable and inconsistent. Regardless of the root cause, the reason an organization's members choose to divert a patient from the ED is that in their judgment they are unable to provide appropriate services to that patient at that time. Quantifying the number of hours an ED maintains diversion status provides an indicator of frequency within that facility without indication of the cause or any way to compare with like facilities.
ED diversion is an occurrence communicated to the community and emergency medical service providers indicating that resources in a hospital are compromised (due to relative shortages of staff, equipment, or beds). It is a request for patients being transported by EMS to be taken to another hospital for service. There may be specific inclusion or exclusion groups of patients, according to local EMS protocol. The diversion occurrences are tracked by the number of hours per time period when that request has been made.
To measure emergency department patient flow, this group is recommending benchmarking key process indicators for hospital capacity and throughput by breaking the time measures down into fractiles. From there, outliers based on cohort groups can be identified and root causes of barriers to patient flow may be identified and targeted for improvement. For instance, using medians instead of averages is generally preferred for comparing facilities with one another. Most interval data are not normal, and averages can be misleading. Medians are far less sensitive to extreme outlier values, and are more appropriate for comparing nonparametric sets.
As an example, arrange all the lengths of stay from a given period of time from lowest to highest value and identify the middle value, that is, the median length of stay for all ED patients (50th percentile). If a similar function is done in multiple facilities, then a second array of values can be created with its own median and other percentile values. Similar exercises can be done for almost all of the performance measures noted above.
A given facility can then understand where it stands compared with other facilities within the same cohort. Facilities using this indicator (regardless of region, size, acuity, and specialty) will recognize when they fall outside the 50th percentile of comparatives which might then trigger performance reviews and initiatives.
Capacity is reflected by patient length of stay, but the definition of “extended length of stay” is arbitrary and not useful. In addition to monitoring minutes or hours of the patient stay, consider breaking length of stay into fractiles. As with measures of flow and throughput, this group recommends benchmarking to identify outliers and target root causes of barriers to throughput.
Emergency medicine leaders are increasingly faced with challenges that go beyond the scope of traditional clinical medicine and department staffing. A thorough understanding of quality improvement principles and benchmarking is now necessary for ED leaders to be successful in providing patient-centered care, improving customer satisfaction, and evaluating service initiatives. Correctly treating emergent complaints is no longer the only focus, and emergency physicians also are being asked to provide safe, timely, efficient, and cost-effective care.
The vocabulary, terminology, and methodology of monitoring performance and improvement in the emergency department developed by this group of ED leaders was a bold first step. It was exciting and appropriate that emergency medicine practitioners were the driving force behind it. Though the group readily acknowledges that this is a first draft and will likely undergo many revisions, front-line practitioners should look at their own improvement efforts more closely. This standardization will make everyone's improvement efforts more meaningful, and this work points in the direction that the specialty is likely to be moving over the next few years.
© 2006 Lippincott Williams & Wilkins, Inc.