Interventions to Improve the Timeliness of Emergency Care

Authors

  • Daniel Handel MD,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Stephen Epstein MD,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Rahul Khare MD,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Denise Abernethy,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Kevin Klauer DO,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Randy Pilgrim MD,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Olan Soremekun MD,

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author
  • Osman Sayan MD

    1. From Oregon Health & Science University (DH), Portland, OR; Beth Israel Deaconess Hospital, Harvard University (SE), Boston, MA; Northwestern University (RK), Chicago, IL; the Emergency Nurses Association (DA), Des Plaines, IL; Emergency Medicine Physicians (KK), Canton, OH; The Schumacher Group (RP), Lafayette, LA; the University of Pennsylvania (OSo), Philadelphia, PA; and Columbia University (OSa), New York, NY.
    Search for more papers by this author

  • Breakout participants: Hany Atallah, Joseph Bushra, Sarang Deo, Ru Ding, Stephen Epstein, Gary Gaddis, Eric Goldlust, David Griffen, Daniel Handel, Thomas Riisgaard Hansen, Keith Kocher, Erik Kulstad, Howard Ovens, Randy Pilgrim, Charles Reese, Phillip Rice, Osman Sayan, Sandra Schneider, Steven Shane, Cynthia Singh, Inger Sondergaard, Olan Soremekun, Susan Stern, Arjun Venkatesh, Abel Wakai, Michael Ward, Lalena Yarris, and Andrew Zinkel.

  • Funding for this conference was made possible (in part) by 1R13HS020139-01 from the Agency for Healthcare Research and Quality (AHRQ). The views expressed in written conference materials or publications and by speakers and moderators do not necessarily reflect the official policies of the Department of Health and Human Services, nor does mention of trade names, commercial practices, or organizations imply endorsement by the U.S. Government. This issue of Academic Emergency Medicine is funded by the Robert Wood Johnson Foundation.

  • The authors have no relevant financial information or potential conflicts of interest to disclose.

  • Supervising Editor: James, Miner, MD.

Address for correspondence and reprints: Daniel Handel, MD; e-mail: handeld@ohsu.edu.

Abstract

ACADEMIC EMERGENCY MEDICINE 2011; 18:1295–1302 © 2011 by the Society for Academic Emergency Medicine

Abstract

With a persistent trend of increasing emergency department (ED) volumes every year, services are intensifying. Thus, improving the timeliness of delivering emergency care should be a primary focus, both from an operational and from a research perspective. Much has been published on factors associated with delays in emergency care, and the next phase in this area of research will focus on exploring interventions to improve the timeliness of care. On June 1, 2011, Academic Emergency Medicine held a consensus conference titled “Interventions to Assure Quality in the Emergency Department.” This article summarizes the findings of the breakout session that investigated interventions to improve the timeliness of emergency care. This article will explore the background on the concept of timeliness of emergency care, the current state of interventions that have been implemented to improve timeliness, and specific questions as a framework for a future research agenda.

Timeliness of care has been a focus of emergency departments (EDs) for over 25 years. The first publication to explore this was a time and motion study that examined wait times at different phases of the patient visit, in particular, time spent in triage, the treatment room, and waiting for discharge.1 This study helped to identify and correct ED inefficiencies. As defined by the authors of this article, timeliness in the context of emergency medicine (EM) is the time it takes to get a patient to an emergency provider to initiate the delivery of emergency care. The Institute of Medicine defines timely care as “reducing waits and sometimes harmful delays for both those who receive, and those who give care.”2

On June 1, 2011, Academic Emergency Medicine held a consensus conference titled “Interventions to Assure Quality in the Emergency Department.” This article summarizes the findings of the breakout sessions that investigated interventions to improve the timeliness of emergency care. Based on the feedback of the participants in the session, the metrics in Table 1 have been proposed to measure current state of operations and future efforts to improve timeliness in the ED setting.

Table 1. 
Metrics
  1. CT = computed tomography; CVA = cardiovascular accident; ECG = electrocardiogram; STEMI = ST-segment elevation myocardial infarction.

•Time from arrival to triage
•Time from arrival to being placed in an ED bed
•Time from arrival to emergency provider
•Time from arrival to ECG for chest pain
•Time from arrival to initiation of antibiotics for pneumonia
•Door to balloon time for STEMI (goal < 90 minutes)
•Door to initiation of thrombolytics for acute ischemic CVAs
•Door to interpretation of head CT scan results to rule out ischemic vs. hemorrhagic CVAs (goal within 45 minutes of arrival)
•Time to reporting of labs
•Time to reporting of radiology tests
•Data ready time (when all relevant data becomes available)
•Data ready to decision time
•Median time from ED arrival to ED departure for admitted ED patients
•Median time from ED arrival to ED departure for discharged ED patients
•Median time from ED arrival to disposition decision
•Median time from ED admit decision to departure time
•Time from bed request to bed assignment
•Admit decision time to ED departure time for admitted patients
•Median time from ED arrival to ED departure for discharged ED patients
•Patients who left before being seen (%) by an emergency provider or before completing treatment

A review of the literature shows that a wide variety of ED metrics have been used in an attempt to make improvements in efficiency and quality of care. Yet, which metrics to measure and how to measure them is still a work in progress.3–5 The most significant results were adopted by the Joint Commission as CORE measures, striving to assure quality care for all ED patients with pneumonia (door to antibiotic < 6 hours), ST-segment elevation myocardial infarction (STEMI; door to balloon < 90 minutes), chest pain (door to electrocardiogram [ECG] < 10 minutes), and stroke (door to computed tomography results < 45 minutes) who present to the ED.3 Welch et al.3 defined time stamps and time intervals for ED visits, including arrival to provider, provider to data ready, data ready to decision time, triage interval, laboratory interval, and length of stay (LOS) for admissions and discharges, to name a few.

A consensus statement, Definitions for Consistent Emergency Department Metrics, was released in July 2011 with broad endorsement. The definitions/time stamps include arrival time, offload time, triage time, admission time, ED LOS, etc.6

Many of the timeliness metrics are also used as quality performance measures. According to growing evidence that the timeliness of emergency care is associated with quality of care, there is internal and external motivation to improve ED operations.3 Having accurate and measurable metrics is critical in planning efficiency or quality improvement projects. Patients and families often correlate length of time spent in the ED with quality.

Interventions to Improve Timeliness of Care

We will divide the numerous interventions by process, target, and modality/resource (Table 2).

Table 2. 
ED-based Interventions
ModelModality
PersonnelProtocolsTechnology
  1. ACS = acute coronary syndromes; CVA = cerebrovascular accident; ESI = Emergency Severity Index; POC = point of care; PNA = pneumonia; RFID = radio frequency identification.

Input• Provider in triage/team triage
• Two track triage for ESILevel 3s
• Nurse-driven advanced order sets
• Bedside registration
• Pull-to-full protocol/bedside triage
• Prehospital data transmission
• Self-registration kiosk
Throughput• Dedicated ED transporters
• ED phlebotomist
• Physician order sets (ACS, PNA, CVA, sepsis, etc.)
• Consultation protocols
• RFID
• POC testing
• Telemedicine consultation
Output• Exit nurse (aide with transfer of admissions to floor)
• Discharge nurse
• Navigator program
• Electronic transmission of nurse floor report
• Full capacity protocol
 

Input Process

Target: Prehospital Setting

Modality: Technology.  Some EDs have taken advantage of technology to assist in the delivery of timely care. They have focused on prehospital as well as pretriage interventions. Examples include prehospital transmission of data and telephonic triage from home.

Prehospital data transmission (PHDT) has been shown to improve timeliness in the care of patients with acute STEMI. The transmission of prehospital ECGs directly to the receiving ED has been shown to shorten door-to-open artery times by allowing the activation of the interventional cardiology team even before the ED arrival of the patient.7

Other applications of PHDT include the transmission of information from personal physicians, clinics, or outside health care facilities such as rehabilitation centers, nursing homes, or transferring hospitals. This transmission can occur via telephone, computer, or fax. The delivery of this type of information, prior to the arrival of the patient, can improve the timeliness of care by allowing the proper triage of the patient and mobilization of resources prior to the patient’s arrival. The early availability of patient data could aid in the early initiation of protocols for disease states such as myocardial infarction, stroke, and sepsis. The time savings could be that which is required to triage the patient and review the history of present illness and past history after arrival; all of this could be obtained prior to arrival. No studies were found that analyze the exact time savings of a formal prehospital data transmission system, except in the case of STEMIs.

Focusing on a reduction in ED volume, Midland Memorial Hospital (Texas) employed a telephonic triage service to assist patients in their decision to use the ED for care of their current medical issue. Registered nurses staffed a triage telephone line 24 hours a day, 7 days a week. Algorithms based on established care protocols were used. Depending on the patient’s condition, they may be directed to the ED or to other health care delivery options. The program resulted in a decline in nonurgent ED patient volume, a reduction in door-to-doctor time from 30 minutes to 15–18 minutes, a reduction in total ED LOS from 180–200 minutes down to 140–150 minutes, a decrease in the left without being seen (LWBS) rate from 14% to 1%, and higher Press-Ganey patient satisfaction scores.8

Target: ED Registration/Triage

Modality: Technology.  Some hospitals have implemented new registration strategies in an attempt to reduce wait times for ED patients. Strategies have included the use of protocols and technologies to help with the often time-consuming process of registering a patient. Success has been claimed with the use of protocols that streamline the initial registration to obtaining the most basic data to identify a patient (“quick registration”), followed by a more detailed registration, often at the bedside, often while the patient is waiting for a provider or after tests and/or treatment has been initiated.9 This parallel process of registration eliminates the time from registration to initiation of care and therefore improves timeliness. Cape Canaveral Hospital implemented bedside registration and saw an 85% reduction in triage to treatment room time, despite a 15% rise in patient volume over the same period.10

Kiosk technology, prevalent in other industries, has made its way to health care. EDs have begun using computerized kiosks in lieu of sign-in lists as patients enter ED waiting rooms. Using these kiosks allows patients to convey more information to the triage nurses than was previously transmitted by the simple sign-in list. The kiosk can be programmed with software that pretriages the patient by asking certain discerning questions. In this way, the triage staff can be alerted to the potential seriously ill patient in the waiting room. In addition, the kiosk technology can be used after triage, for those “less sick” patients, to gather demographic and insurance information as well as history of present illness, past medical history, medication lists, and allergies. Data from the kiosk can allow ED management to gauge waiting room census on a minute-to-minute basis and thus accurately employ resources.11 No formal studies were found on the effects of kiosk technology on timeliness of care in the ED.

Modality: Personnel.  One approach to improve the timeliness of care is to initiate care as proximate to the patient’s arrival in the ED as possible. Strategies to employ the reassignment of personnel resources include bedside/registration triage (BRT), team triage (TT), and two-track triage (TTT).

Bedside registration/triage uses a strategy of maximizing resource utilization and parallel processing to improve timeliness. Using a “pull-to-full” strategy, patients are immediately brought back to open chairs or stretchers within the patient care area of the ED. Registration and triage are performed at the bedside by registration staff with mobile computer workstations and by the primary nurse caregivers, respectively. If capacity exists in the ED care areas, this strategy provides the following benefits: 1) elimination of the bottleneck that can occur with the triage process; 2) allowing the primary nurse to have a better understanding of the patient’s condition; 3) creating one less patient handoff, reducing the likelihood or error; and 4) reducing one nurse encounter, thereby streamlining the patient entry process. BRT was implemented at JFK Medical Center (Edison, NJ) with an estimated reduction in door-to-triage time from 25 to 17 minutes, an improvement in door-to-physician time from 50 to 38 minutes, and a reduction in door-to-discharge time of 1.25 hours (3 hours down to 1.75 hours).9

Team triage involves the allocation of a physician, nurse, paramedics, ED technicians, or other staffing configurations, to perform triage as a team. Patients determined to require urgent or emergent care (ESI Levels 1 and 2 and some Level 3s) are immediately brought to a stretcher in the acute care area of the ED. Nonurgent cases are cared for by the TT team. The complete history and physical examination are completed in triage or the TT area. Either the patient is either immediately discharged or, should they be necessary, diagnostics are ordered and performed. The patient either waits in the waiting room or in chairs (not stretchers) in the TT area. Implemented in the Vanderbilt University Medical Center ED, the results of the TT program were: most patients were seen within 1 minute of arrival, a decrease in the percentage of LWBS from 5% to less than 1%, and a marked increase in patient satisfaction.12,13

Two-track triage uses components of the above two processes. It uses the capacity of the acute care areas of the ED for triage of the more acutely ill patients and involves the allocation of a physician to triage. In the TTT model, a nurse performs a “quick look” triage rather than a full triage. Patients are divided into “less sick” and “sicker” subgroups. The primary nurse to whom that patient is assigned immediately transfers the “sicker” patient onto a stretcher in the acute care ED, and a full triage is performed. The “less sick” patient is assessed by a nurse/physician team. This provides a second chance to decide if the patient should really be in the “sicker” group. If not, care proceeds similarly to the nonurgent cases in the TT model above. Banner Health implemented TTT in eight of their EDs in Phoenix and Colorado. The process was called “door-to-doc.” Average time to be seen by a doctor fell by 58% from 117 to 49 minutes, the number of LWBS fell by 76% from 7.1 to 1.7%, the ED LOS fell by 14% from 310 to 268 minutes, and the ED patient volume increased by 1% from 110,400 to 111,503 patients.14

Modality: Protocols.  The implementation of nurse-driven advanced order sets in triage has been used as a means of initiating the assessment and treatment of patients with certain complaints. The ability to initiate care from triage allows EDs to use parallel processing. As patients are being registered and are waiting to see a provider, diagnostics (laboratory testing and/or imaging studies) that are felt to be needed can be completed, thereby shortening the overall LOS. Parallel processing, the performance of processes simultaneously so that wait times run concurrently rather than in series, is an excellent strategy for time savings.

In 2003, in response to ED crowding, Gwinnet Medical Center (Lawrenceville, GA) implemented the use of nurse-driven advanced order sets in their triage. In Urgent Matters, they reported an improvement in patient satisfaction scores as well as a qualitative improvement in throughput times.15 It is not clear that any rigorous methodology was used to definitively show the benefits of this intervention. A systematic review supported that nurse-driven orders can reduce ED LOS.16

Input/Throughput Process

Target: ED

Time that is spent searching for personnel, patients, and equipment negatively affects the timeliness of care. After assessing a patient, a triage nurse may need to find an empty room and stretcher for that patient. Equipment such as a wheelchair or an ECG machine might be necessary. The acute care ED nurse who will be assigned to that patient must be located for a report to be given. The time spent in the performance of all of these processes is reduced with the use of radiofrequency identification (RFID) technology. RFID tags in personnel badges, patient ID wristbands, and equipment are localized by sensors installed throughout the ED and even other hospital departments (e.g., radiology). This information is sent to tracking software that conveniently displays the data on a computer station, allowing the triage nurse to quickly identify vacant rooms, empty stretchers, an ECG machine or even the accepting ED nurse.

Albert Einstein Medical Center (Philadelphia) installed RFID technology into their ED in 2003 in an effort to streamline patient flow, improve ED throughput, and create a more accurate minute-to-minute “big picture” patient flow map for their administrators. The program resulted in: 1) a reduction in ED LOS from 9 hours (2002) to 3.5 hours (2008; even with a 24% increase in overall ED volume); 2) a reduction in the LWBS rate from 5% (2002) to 0.5% (2007); 3) an 89% reduction in ambulance diversion; 4) an increase in Press-Ganey patient satisfaction scores of 15% to 20%; and 5) an improvement in core measures compliance (4-hour delivery of antibiotics for pneumonia patients increased to 94%).17

Throughput Process

Target: ED

Modality: Personnel.  The recording of health care information is a necessary, but time-consuming, process in the delivery of health care. A recent focus of the U.S. federal government and stimulus spending package, the implementation of electronic health record (EHR) systems has improved the legibility, portability, and completeness of health care information but often, at least initially, at the expense of valuable time. Some EDs have sought strategies to improve provider efficiency and timeliness, increasing the amount of provider time spent at the bedside and decreasing the amount of time spent in front of a computer workstation documenting information.

One solution is the use of scribes. Scribes are individuals with some knowledge of health care and health care terminology who accompany the provider during the interview and examination of a patient. The scribe, rather than the provider, enters the information gathered into the EHR. This allows the provider more time with patients and allows the provider to see more patients per shift. Several studies have found an improvement in throughput and relative value units (RVUs) generated per hour by providers.18,19

Following the implementation of an EHR in the ED and a perception of increased time spent at the computer, the leadership at the John F. Kennedy Medical Center (Edison, NJ) hired scribes to assist in the documentation of the patient encounter. Although not rigorously studied, they report an increase in patient satisfaction after implementing the use of scribes.20

Modality: Technology.  Another strategy to improve provider timeliness and efficiency in this age of EHRs is the use of voice recognition software (VRS) to allow for real-time, seamless dictation of charts. However, no studies were found that document the impact of VRS on such factors as door-to-provider times, ED LOS, or LWBS rates.

Modality: Protocols.  In the care of ED patients, the use of consultation services is a common event. Unfortunately, the process is fraught with time delays: 1) time to phone/page the consultant, 2) time for consultant to reply, 3) time for consultant to arrive to see patient, and 4) time for consultant to make recommendations or provide care. The use of consultants can lead to sizable delays in patient care, and some EDs have focused on streamlining this process. Strategies include the establishing of: 1) consultation guidelines that predetermine the illnesses/conditions that each consultative service covers, 2) agreed upon standards for the timing of consultations (e.g. how long is too long), and 3) specific mechanisms for consultation in the ED (e.g., dedicated consultation clerk to allow accountability).

Due to the various methods for initiating consultation (e.g., Web page, e-mail, telephone, pager system, cell phone), it is often difficult to gather data on consultation metrics. A centralized system for consultation in the ED would make this easier. The ED at Stony Brook University Medical Center (New York, NY) implemented such a program and realized many benefits, including a decrease in ED LOS of 1 hour in patients receiving a consult.21

Modality: Personnel and Technology.  The multiple steps involved in the care of an ED patient can cumulatively add to the length of time the patient spends in the ED. For the patient requiring the analysis of a blood specimen, the collection of that blood, as well as its processing, can be responsible for a significant amount of the patient’s LOS. EDs have looked to decrease the time to collection and processing time of blood specimens and have focused on reducing the laboratory turnaround time for the results.

Emory Healthcare (Atlanta, GA) implemented a program using on-call or dedicated ED phlebotomists during their peak volume hours. The phlebotomists collect the blood specimens, label them, and ensure their delivery to the hospital laboratory. They track and follow-up on the results for the providers. As a result there was: 1) a decrease in laboratory turnaround times from 165 to 119 minutes, 2) shorter door-to-physician times (reduction of 24 minutes), 3) lower contamination rates (decrease from 5% to 1.5%), and 4) an improvement in patient satisfaction scores. Unfortunately, there was no measurable effect on ED patient LOS.22

Output Process

Target: Hospital

Modality: Protocols.  One strategy uses the full capacity of the hospital to care for admitted patients, pending inpatient bed placement. The Full Capacity Protocol (Hallway Bed Protocol) calls for the use of free space on medical and surgical floors (e.g., hallways, solaria, lounges) to board patients without formal bed locations. Devised and implemented at Stony Brook University Hospital, the use of the protocol was associated with: 1) lower hospital LOS for admitted patients (6.2 to 5.4 days), 2) higher patient satisfaction (survey), and 3) no negative patient outcomes from inpatient unit boarding.23,24

Hospital bed availability is influenced by many variables. One strategy for improving bed availability, and thus the access to these beds by admitted patients boarding in the ED, is to limit this variability and create a more even and predictable flow of patients to the inpatient beds. Due to a concentration of elective vascular surgery admissions at Boston Medical Center on Tuesdays, Wednesdays, and Thursdays, it was difficult for admitted ED patients to obtain inpatient beds on those days. An initiative was undertaken to even out the distribution of elective vascular surgery cases (surgical and elective schedule smoothing) over the five weekdays. After this, the amount of time admitted ED patients waited for inpatient beds decreased from 3 to 2.2 hours.25 Other work was reviewed at the same institution that found a significant amount of critical care diversion times were associated with scheduled admissions.26

A study at Northwestern University’s Feinberg School of Medicine used a cross-sectional computer modeling analysis to study the effect of early inpatient discharges on the holding of admitted patients in the ED. At baseline, their ED has a total of 77.0 hours per day of admitted patient boarding. A shift of the peak inpatient discharge time to four hours earlier completely eliminated ED boarding, while the discharge of 75% of all inpatients by noon as well as the discharge of all inpatients between 8am and 4pm both were associated with a decrease in the total boarding hours to 3.0.27

Modality: Facility.  Another strategy looks to create inpatient beds earlier in the day by moving stable, ambulatory inpatients scheduled for discharge to inpatient discharge units. In 2002, The Regional Medical Center at Memphis, faced with a severe ED boarding issue, created a “discharge resource room” on the ground floor of the hospital for inpatients pending discharge. This resource was available from 8am to 10pm daily. Sixty percent of all hospital discharges were moved through this unit, and the average total ED throughput time decreased from approximately 9 hours to 5 hours over a 1-year period.28

Workshop Facilitation Method

The “Interventions to Safeguard Timeliness of Care” breakout session was held in the afternoon, following the morning of workshops by content experts in various aspects of crowding research. Sample research questions were drafted prior to the session from the planning committee to stimulate discussion of the participants. The participants were divided into five groups, which rotated to a discussion leader on each research question to provide feedback. Each group rotated through the research questions once and discussed further areas in that particular area. At the end of the session, each participant was provided three dots with which to vote for what they felt were the most important questions. The top four questions were included for further discussion in this article.

Research Questions

Research Question 1: How Does the Use of Advanced Treatment Protocols and ED Order Set Utilization Improve Timeliness of Care?

The use of predefined order sets by both physicians and nurses might improve the timeliness of emergency care, but at the potential cost of efficiency. While the use of these order sets is likely worthwhile for time-sensitive conditions, full cost–benefit analyses will be necessary to determine for which conditions these order sets might be most beneficial.

Examples of advanced treatment protocols exist largely in the context of advanced triage systems that attempt to improve the efficiency of waiting times by using that time for laboratory, radiology, and other ancillary services expected to be used during the course of care.29 The order sets are based upon diagnosis-driven clinical pathways. To the extent that advance treatment protocols empower nursing staff to initiate ED care, such protocols also might provide a powerful morale boost. However, physicians must also see these protocols as efficient to provide positive feedback to the nurses initiating these order sets.

Physician order sets have been developed to improve compliance with various quality metrics,30,31 but may also play an important role in the improvement of timeliness.

Specific research questions include:

  • 1Provider variability: while overall the use of such order sets might result in certain efficiencies of time, participants expressed concern over provider variability in the use of order sets.
  • 2Order set efficacy: which order sets improve timeliness of care the most? Do order sets for certain conditions perform better than others? Are there characteristics that identify well-performing order sets?
  • 3Timeliness vs. efficiency: what are the costs in efficiency, if any, sacrificed to improve timeliness?
  • 4Outcomes: does the use of order sets lead to demonstrable improvement in timeliness measures, such as overall ED LOS? Does the use of order sets lead to improved compliance with core measures of quality (e.g., aspirin in acute myocardial infarction)?

Research Question 2: Comparing ED Interventions (Split Patient Flow Modeling With ESI 3 Patients and an Internal Waiting Room, Provider in Triage, etc.): A Multicenter Trial Is Needed Examining Their Effects on ED LOS, LWBS rates, and a Cost–Benefit Analysis

Creating new models of patient flow through the ED has been demonstrated to improve certain timeliness metrics, but often at increased cost (e.g., placing an additional physician at the point of triage).32–34 To date, there have been few if any comparative effectiveness studies of various interventions to ED patient flow. As the heterogeneity of EDs presents certain challenges (e.g., some interventions will be more effective in larger EDs than in smaller EDs), simulation models might be an effective way to make preliminary studies of these interventions. Another possibility for study would be to use large existing networks of EDs, such as the ED Benchmarking Alliance.

One key intervention point identified was one bottleneck, defined as the time between available results and provider action upon those results. Formal identification, evaluation, and measurement of bottlenecks might better inform ED interventions. In particular, identification of human factors in bottlenecks, often not accurately identified, might yield additional opportunities for improvements in timeliness. Outcome measures will need to gauge the cost effectiveness of interventions, not only of timeliness (e.g., ED LOS), but also upon patient satisfaction and the effectiveness (quality of medical care) of the care provided.

Research Question 3: How Do Hospital-level Solutions, Such as Surgical Schedule Smoothing, Active Bed Management, and Full Capacity Protocols, Affect ED Boarding Hours and Other Measures of Timeliness Such as Door-to-doctor Time, LWBS, and Overall ED LOS?

It is now recognized that ED crowding stems from hospital crowding.35–37 ED crowding is associated with delays in care, resulting from hospital crowding, which ultimately affects the timeliness of the ED.38,39 Therefore, many participants felt that interventions at the hospital level would be of benefit in improving the timeliness of ED care. Numerous interventions at the hospital level have been studied, some extensively,9,10,15,20,21,25,28 but a comprehensive study of the effect of these interventions on a standard set of ED timeliness measures is lacking.

Participants identified a number of interventions lending themselves to more extensive study of timeliness:

  • 1Improving ED throughput.
    •  a. Time guidelines for consultations.
  • 2Easing ED outflows to inpatient units.
    •  a. Full capacity protocol.
    •  b. Minimize ED workups that can be accomplished on inpatient units.
  • 3Reducing competing nonemergent inpatient bed demands.
    •  a. Surgical schedule smoothing.
    •  b. Canceling elective surgeries.
    •  c. Deferring nonemergent admissions and transfers.
  • 4Improved inpatient bed management.
    •  a. Enhanced bed turnover with incentives for nursing and cleaning personnel.
    •  b. Early discharge planning.
    •  c. Discharges prior to noon.
    •  d. “Bed czar” (bed management authority).
  • 5Extended availability of support services (e.g., physical therapy/occupational therapy/social services).
  • 6Code help procedures: cancelation of nonessential meetings/services when the ED reaches predetermined crowding thresholds.

There was also recognition from participants that hospital level changes must have a high-level hospital administrative champion. Such a champion demonstrates that improving ED timeliness is a hospital-wide goal and can help drive the improvement of outcome metrics, such as time from decision to admission or discharge to time of a patient leaving the ED. In academic centers, such a champion is critical to driving changes in standard operating procedures among physician trainees that delay discharges.

Research Question 4: Effect of Point-of-care Testing on Timeliness

Numerous point-of-care (POC) tests are available, and several have proven utility within the ED.40–42 Participants identified two groups of tests that might improve ED timeliness: 1) tests for immediate/life-threatening decision-making and 2) tests for reducing time for patient disposition. Examples of the former include potassium (dysrhythmias), pregnancy (ectopic pregnancy), lactate (sepsis), portable ultrasound (shock), and INR (intracranial hemorrhage). Examples of the latter include cardiac markers (allow for patient disposition to an observation unit), creatinine (for imaging with contrast), and Rh (for pregnant patients).

While studies have demonstrated the effects of these tests on ED timeliness,43–45 few have considered the additional, hidden costs of maintaining POC testing within the ED environment.46 POC testing may require specialized training of ED staff who may be taken away from other duties to maintain POC equipment and materials. Performing POC testing will also take some ED personnel away from other duties. Without careful attention to reimbursement models, it is possible that ED POC testing might become a new ED cost center. At the same time, POC testing might also drive revenues away from a hospital’s pathology department. Furthermore, more testing also might drive more random error, resulting in additional costs for confirmatory testing. Careful consideration of which tests yield the greatest benefits will be important to avoid unnecessary costs.

Reimbursement may also raise issues of equity. EDs with advantageous payer mixes may be able to improve timeliness through POC testing where EDs without such payer mixes cannot.

Specific research questions are: 1) what are the costs–benefits for each POC test considered for ED use and 2) under what circumstances does POC testing benefit the ED and/or the hospital? At what point are improvements in ED timeliness outweighed by other considerations?

Conclusions

Emergency department timeliness is a critical factor effecting operational performance of the complex care delivery processes required to deliver emergency care. Operational performance ultimately affects quality of care, patient safety, and operational and financial efficiency. Although excellent work is being done in this area, the database of literature and depth of knowledge on this topic is underdeveloped. With respect to the anticipated demands associated with implementation of the Affordable Care Act, and the ongoing expectancy of ever-increasing ED volumes, investigation of solutions for operational excellence and timeliness are essential to meet current and future demands. Although much work in ED operational performance is in its infancy, interest and enthusiasm for the implementation of evidence-based solutions is growing rapidly. This consensus conference is a very needed early step to that end.

Ancillary