MILITARY MEDICINE, 184, 1/2:e169, 2019

Applying Lean Principles to Reduce Wait Times in a VA

Emergency Department
Anita A. Vashi, MD, MPH*; Farnoosh H. Sheikhi, MS†; Lisa A. Nashton, MHA‡; Jennifer Ellman, MSN,
RN§; Priya Rajagopal, MD§; Steven M. Asch, MD, MPH*†

ABSTRACT Introduction: We describe the use of Lean quality improvement methodologies at a Veterans Affairs
(VA) medical facility to redesign Emergency Department (ED) front-end operations and improve ED flow, specifically

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

to reduce time from Veteran arrival to provider evaluation. Materials and Methods: The intervention, a Rapid Process
Improvement Workshop (RPIW), took place during January 2014 at the VA Palo Alto Health Care System
(VAPAHCS). Key changes made as a result of the RPIW included standardizing and streamlining evaluation and
hand-off processes, better-delineating roles for RNs and MDs, more efficiently utilizing beds and improving team com-
munication. We collected 13 months of pre-intervention and 13 months of post-intervention data. The primary outcome
was the change in “Door to Doctor” time between the pre-intervention and post-intervention periods at VAPAHCS
compared with contemporaneous national control facility sites. Secondary outcomes included the change in “Door to
Triage” time and the rate at which patients left without being seen (LWBS). Data analyses were performed using a
regression-adjusted difference-in-differences approach. This was a quality improvement project and the institutional
review board determined that this project does not meet the definition of human subject research. Results: Overall,
“Door to Doctor” time at VAPAHCS decreased 12.6 minutes after the intervention, compared to 3.7 minutes in the
control sites. Regression-adjusted difference-in-differences estimates for “Door to Doctor” time and “Door to Triage”
time showed a significant reduction at VAPAHCS compared with control sites (8.9 minutes and 5.0 minutes, respec-
tively), during the same time period (standard error = 3.5 min; p = 0.01 and standard error = 1.7 min; p = 0.004,
respectively). Regression-adjusted difference-in-differences estimates for LWBS rates showed that LWBS did not sig-
nificantly change at VAPAHCS compared with control sites (0.1% vs. 0.3%, p = 0.8). Conclusions: Using Lean princi-
ples, VAPAHCS was able to improve Veteran flow in the ED. Use of Lean methods foster interdisciplinary teams and
problem-solving across departments and are one approach VA EDs can use to address systemic factors and contribu-
tors to ED crowding and improve care for Veterans. Future study should incorporate additional measures of quality to
determine the effect of Lean on Veteran outcomes and should evaluate the long-term sustainability of the
improvement.

INTRODUCTION EDs are not exempt from these trends, and several reports
have found that ED crowding and shortages of staffing and
Problem Description
beds, along with an absence of a diversion policy, has nega-
In a 2006 report, the Institute of Medicine’s Committee on the tively impacted quality care provided to Veterans.5–9
Future of Emergency Care in the United States Health System
reported on the national crisis of Emergency Department (ED)
crowding.1 Since the Institute of Medicine’s landmark publication, Importance
Hospital-Based Emergency Care: At the Breaking Point, EDs Timely and effective care in hospital EDs are essential for
continue to face growing problems with crowding, delays, good patient outcomes.10–12 Delays before receiving care in
and cost containment.2 In particular, patient wait times and the ED can reduce the quality of care and increase risks and
flow through the ED have come under close scrutiny.3,4 VA discomfort for patients with serious illnesses or injuries.10–13
The length of time patients wait to see a provider in the ED
*Center for Innovation to Implementation, Palo Alto Veterans Affairs is also an important driver of patient satisfaction.14–16 In
Health Care System, Palo Alto, CA 94025. response, many EDs are now communicating estimated wait
†Division of Primary Care and Population Health, Stanford University times to the general public. The Centers for Medicare and
School of Medicine, Stanford, CA 94305.
Medicaid Services is now reporting ED wait time measures
‡William Jennings Bryan Dorn VA Medical Center, Columbia, SC
29209.
in an attempt to hold hospitals publicly accountable for the
§Palo Alto Veterans Affairs Health Care System, Palo Alto, CA 94304. speed and efficiency of their EDs by showing comparison
The views expressed are solely those of the authors and do not reflect wait times in each community and state.
the official policy or position of the Department of Veterans Affairs or the A 2005 joint report of the National Academy of Engineering
U.S. Government.
and the Institute of Medicine observed that the healthcare sector
doi: 10.1093/milmed/usy165
Published by Oxford University Press on behalf of the Association of has been slow to use systems engineering tools and information
Military Surgeons of the United States 2018. This work is written by (a) US and communication technologies to improve the quality, safety,
Government employee(s) and is in the public domain in the US. and efficiency of its services.17 One promising management

MILITARY MEDICINE, Vol. 184, January/February 2019 e169

 Applying Lean Principles to Reduce Wait Times

approach to achieving these aims is the Lean method. Often waiting times to be seen by a provider and (2) evaluate if the
termed “Lean thinking,” this quality improvement philosophy is intervention achieved its stated aim using quasi-experimental
a bundle of concepts, methods, and tools that grew from the controls and a difference-in-differences approach.
Toyota Production System and is used widely in the manufactur-
ing industry. The Lean process evaluates operations step by step
METHODS
to identify waste and inefficiency, and then creates new solu-
tions to improve operations, remove waste, increase efficiency, Context
and reduce expenses. There has been an increased interest in The VAPAHCS ED serves a Veteran population of 85,000,
implementing Lean in the healthcare sector,18 including EDs;19 spread throughout two inpatient facilities and eight outpa-
in a 2009 survey of US hospitals, of the 53% of hospitals tient clinics; the FY14 census of the VAPAHCS ED was

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

reporting having implemented some form of Lean, 60% approximately 20,000. During this study, the ED consisted
reported implementing Lean in the ED.20 VHA leadership of 12 acute beds, 4 Fast Track beds, and 2 treatment rooms
has called for the advancement of the VA healthcare system that could be used as examination areas during overflow
by using Lean Six Sigma, and in 2011, in an effort to improve periods.
the quality and efficiency of Veteran healthcare, the national
Veterans Engineering Resource Center initiated the Lean VAPAHCS Lean Process
Enterprise Transformation program to promote Lean princi-
ED Value Stream Scoping
ples and strategies in 10 VA medical facilities.
The VAPAHCS ED began scoping their value stream for the
ED in February 2013. The executive sponsor for the ED Value
Available Knowledge Stream was the Deputy Chief of Staff, and the two Value
Process optimization in the ED has been well studied using a Stream Owners were the Chief and the Nurse Manager of
variety of quality improvement methods. Several studies have the ED. The Executive Sponsor and Value Stream Owners
analyzed specific interventions such as adding a physician to teamed with nursing leadership and were facilitated by a Value
triage or limiting inpatient boarding on ED flow.21–26 The Stream Process Improvement Facilitator, Process Improvement
most commonly measured outcomes of process optimiza- Coach, and Lean Consultant to form the guiding coalition for the
tion include length of stay (LOS),21–25,27 time to provider,21,24,28 ED Value Stream. Target states and associated metrics were
left without completed assessment,21,24 and left without being defined and aligned with VAPAHCS’s four strategic prior-
seen (LWBS).22,27 Such studies most often a priori select and ity areas: people, access, quality, and safety.
implement an intervention that is evidence-based. Others have
used multidisciplinary team-based approaches to achieve ED Value Stream Analysis Event
improvements. A Value Stream Analysis event was facilitated during April
Lean tools and methods are becoming a popular way to iden- 30 through May 2, 2013. Participants included the Deputy
tify process improvements and design interventions to improve Chief of Staff, Chief of the ED, Nurse Manager of the ED,
ED flow.19,23,27–40 However, few have rigorously evaluated its Assistant Nurse Manager of the ED, Chief of Environmental
use in the ED and even fewer have examined Lean in the VA Management Service, Chief of Inpatient Mental Health, Chief
setting.40–42 Of studies to date, most have been conducted in sin- of Laboratory Services, Chief of Radiology, Medical Support
gle centers, report only pre–post measures at best, and often lack Assistant Supervisor, Chief of Admitting and Eligibility, a
longitudinal data and use of comparison groups. When adequate general surgeon, the Director of Quality Improvement for
comparison groups are used, the evidence is mixed at best. A Inpatient Medicine, and the head of bed management. All
recent large multi-center controlled study of Lean-based interven- agreed that the focus would be on Veteran flow through the
tions in Canada found that while there were reductions in ED ED, guided by metrics. The outputs of the event included
LOS among the 36 hospitals that participated in a Lean program, several process improvement project charters that aligned
similar reductions were also observed among the 63 matched VAPAHCS priority areas with value stream metrics and target
control hospitals over the same period.29 One possible explana- goals. For example, for the priority area “People,” the metric
tion for these conflicting results is the inconsistent application of chosen was “Percentage of staff actively engaged in improve-
the Lean principles. ment work” with a goal of 85%. In the priority area “Quality,”
the metric chosen was “Door to Doctor time” with a goal of
Goals of Investigation less than 20 minutes. Metrics and target goals were based
As part of a strongly supported VA Lean transformational on group consensus.
effort, the VA Palo Alto Health Care System (VAPAHCS)
ED prioritized flow redesign. The objective of our study is Intervention: “Door to Doctor” Rapid Process
to (1) describe the extent and depth to which Lean methodol- Improvement Workshop
ogies are used of the use of Lean methodologies to rapidly Following the Value Stream Analysis event, a Rapid Process
redesign ED front-end operations that aimed at reducing Improvement Workshop (RPIW), facilitated by the Value

e170 MILITARY MEDICINE, Vol. 184, January/February 2019

 Applying Lean Principles to Reduce Wait Times

Stream Process Improvement Facilitator, occurred January Final RPIW Recommendations

13–17, 2014, to focus specifically on the “Door to Doctor Four specific changes were made to systems of care and
Time” metric. The participants of this 5-day RPIW included Veteran flow. First, the triage and rooming process was
the ED Nurse Manager, Assistant Chief of the ED, a medical restructured primarily to handle high-volume periods (11:30 a.
support assistant, two ED physicians, and two triage/charge m.–8 p.m.). During peak flow hours, an additional nurse was
nurses from the ED. The RPIW began with process map- assigned to triage duties to expedite rooming of Veterans.
ping, wherein the current process steps were diagrammed Utilization of nursing order protocols by the triage nurse was
and validated. The time required for each step was estimated encouraged when appropriate. A “pull to full” strategy was
or measured when possible using process observation and/or employed so that when rooms were available, Veterans were
electronic data. Participants also identified opportunities and brought back immediately and triaged at bedside by the primary

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

ideas for improving the current state map (Fig. 1). After nurse. Once roomed, all Veterans were changed into gowns,
mapping the current state, participants solicited feedback via and saline lock placement was deferred until during or follow-
informal, unstructured interviews from Veterans, staff in the ing a physician evaluation. The second change was made to
ED, and staff from services that support the ED. This feed- staffing roles, detailed in Table II, which were more clearly
back helped validate the current state map as well as high- defined to reduce confusion and rework. The third change was
light some possible root causes for delays in “Door to Doctor” to standardize the huddle schedule and include all ED staff
evaluation time. in the huddles, which greatly improved communication
Next, bottlenecks, waste, and other process problems were between staff and helped identify bottlenecks in Veteran
identified and root causes of those problems were sought flow. Communication was further improved by utilizing
(Table I). After brainstorming and target-state mapping of the electronic tracking system and comment field to guide huddle
possible improvements, participants proposed process rede- discussions and communicate Veteran care plans. Finally, the
sign recommendations. Changes were adjusted in an iterative last significant change was the use of a proactive ED bed
way using the plan-do-study-act cycle. The final redesign rec- management system, where stable, roomed Veterans not requir-
ommendations are detailed below and in Table II. ing active therapy or monitoring were moved back to the ED
Finally, participants developed standard work documents waiting room after initial evaluation to wait for pending test
to train staff in the new processes, communicated results of results, consultations, etc. This created space within the ED
the RPIW to impacted staff, and created implementation and to begin workups on new Veterans. Additionally, stable, ambu-
audit plans. These plans were also highlighted in a visual latory Veterans were sent to outpatient radiology and phlebot-
management board located in the ED. On the final day of the omy with a priority pass that ensured timely imaging and blood
event, recommendations and lessons learned were presented draws while freeing nurses to perform other tasks and spend
to VAPAHCS executive leadership and management. more time with more critically ill Veterans.

FIGURE 1. Current state map used to identify problem areas in processes.

MILITARY MEDICINE, Vol. 184, January/February 2019 e171

 Applying Lean Principles to Reduce Wait Times

TABLE I. Root Cause Analysis During RPIW

Identified Problems Root Cause Analysis

Delay from Veteran check-in to nursing triage During peak hours, volume too great for one triage nurse to handle
No alternate processes during high-volume periods
Non-value added documentation and travel by RNs between triages
Delay from triage to room placement No empty beds available to room Veteran
Admitted ED Veterans boarding and waiting for inpatient beds
No protocol for using ED hallways or overflow exam rooms
Some beds occupied by Veterans not requiring monitoring or treatment and only waiting for results
Following triage, Veterans returned to waiting room even if open bed available
Triage limited to triage intake area and often a bottleneck

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

Nurse assistant frequently pulled from triage to assist in other areas
Delay from rooming to RN evaluation RN not informed they have been assigned a new Veteran
RN busy with other Veterans
RN occupied with non-value added documentation
Delay from rooming to MD evaluation Veterans are not undressed and ready for examination
Nurse placing saline lock
No standard for Veteran assignment. MD self-assigns
Constant MD interruptions (ambulance, phone calls, signing EKGs)
MD occupied with documentation

RN, registered nurse; EKG, electrocardiogram; MD, medical doctor.

TABLE II. Final Recommendations Following RPIW

Triage
Restructure triage
– “Pull to full”: Veterans pulled directly into a room when available (minimize waiting room stay)
– Number of triage nurses increased from 1 to 2 during high-volume periods (11:30a-8p)
– Nursing triage protocols: triage nurse initiates workup by placing orders if ED is busy
Standard rooming and prep
– All Veterans change to gown prior to provider evaluation
– Saline lock placed after or during provider evaluation
– Primary nurse encouraged to utilize nursing order protocols when appropriate
Staffing roles
Defined nursing roles
– Charge nurse serves as Flow Manger: initiates huddles, communicates regularly with flow physician, reallocates and redirects resources as needed
– Resource nurse: responsible for overflow Veterans in waiting room, hallway and overflow helps with Veteran flow as needed
– Primary nurse: helps triage Veterans at bedside
– Fast track nurse: helps triage fast track Veterans at bedside in Fast track area
– Nursing assistant in triage: rooms and preps all Veterans, vitals, and EKGs
Defined physician roles
– Flow physician: takes all EKG reads, greets all ambulances, communicates with charge RN throughout shift
– Communication physician: fields all physician phone calls for non-specific needs (i.e., clinic transfers)
Communication
– Huddle schedule standardized (set times for day, afternoon and evenings shifts and during change of shift)
– All ED staff (nurses, physicians, medical support assistants and administrator on duty) included in huddles
– Veteran status (likely admit versus discharge) reviewed for all Veterans during huddles
– Patent flow bottlenecks identified during huddles
– Availability of inpatient and ICU beds reviewed during huddles
– Electronic tracking board used to facilitate huddles
– Comment field on electronic track board used as a communication tool for care team
Proactive ED bed management
– Stable Veterans not actively receiving monitoring or therapy (i.e., no IV, nebulizer, pain medications, safe to ambulate, etc.) moved back to ED
waiting room after MD evaluation to wait for test results, social work, etc
– Ambulatory, low-acuity Veterans provided with “fast pass” to receive blood draw and imaging in outpatient center (located down the hallway
from the ED)

EKG, electrocardiogram; ICU, intensive care unit; IV, intravenous; RN, registered nurse; MD, medical doctor.

e172 MILITARY MEDICINE, Vol. 184, January/February 2019

 Applying Lean Principles to Reduce Wait Times

Outcome Measures was created by merging all the aggregated data sets together
The primary outcome was the change in “Door to Doctor” longitudinally.
time between the pre-intervention and post-intervention peri- To identify control groups, we considered all other VHA
ods in VAPAHCS versus control facility sites. “Door to facilities with an ED that reported data during at least 1
Doctor” is defined as the median time in minutes between study month (n = 115). Because the EDIS was undergoing a
check-in and the first assignment of a provider for all ED national roll-out with variable implementation just prior to
Veterans seen during the time period specified.1 the study period, only sites with good EDIS adoption were
Secondary outcome measures include the change in used. Adoption was defined using a nationally collected and
“Door to Triage” time, change in LOS for admitted and dis- reported measure of accuracy (0–100%) that calculates the
charged Veterans in the ED, and change in LWBS rate. percentage of visits where the name of the initially assigned

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

“Door to Triage” time is defined as the median time in min- provider matches the name of the person entering the initial
utes between Veteran check-in and assessment by the triage provider assignment. This measure is the best available proxy
nurse. LOS is defined as the median time elapsed in minutes for full adoption of EDIS and thus ensures higher accuracy in
between Veteran check-in and departure from the ED, and waiting time data; as such, it was recommended by VA EDIS
was measured for both Veterans admitted to the hospital and implementation leaders to ensure high quality waiting time
those discharged from the ED. LWBS occurs when a data. VAPAHCS had an average of 90% accuracy over the
Veteran leaves the ED prior to being assigned a provider. 2-year study period. Therefore, facility sites with low percen-
This measure was reported as percent of total visits for the tages of accuracy (<90%) were excluded (n = 104), leaving
time period specified where the disposition is indicated as 11 control sites (Table III).
“left without being seen”2 at each facility per month.
Analysis Plan
For each outcome measure of interest, daily medians (based
Study Design and Methods on individual Veteran encounters) were averaged over monthly
The intervention (RPIW) at VAPAHCS took place during intervals for each facility. Thus, for a given outcome measure,
January 2014, and facility-level data were collected over 2 over the 13-month pre- and 13-month post-periods, there were
years. We considered 13 months (January/2013–January 26 measurement points for the intervention site (n = 1) and
2014) before the intervention as the pre-intervention period 286 measurement points for the control sites (n = 11).
and the 13 months after the intervention (February 2014– Data analyses were performed on the primary and second-
February 2015) as the post-intervention period. ary outcomes using a regression-adjusted difference-in-
The data sets were extracted retrospectively from VHA differences approach. This approach measures an intervention
Support Service Center reports. These reports provide aggre- effect while accounting for any pre-intervention differences
gated data at the facility level with information about ED between the intervention and comparison sites. The differ-
workload counts, flow performance measures, and EDIS ence between the intervention and control groups is measured
adoption metrics. We have used aggregated facility-level before the intervention (pre) and following the intervention
data to compare the operational performance measures at (post). Therefore, in order to calculate the difference-in-
VAPAHCS to other comparable VHA facilities. In order to differences, the difference between the intervention and con-
study the changes before and after the intervention, we trol groups before the intervention (pre) is subtracted from
extracted the facility-level data from EDIS each month for 2 the difference between intervention and control groups after
years with all related ED characteristics. An analytic data set the intervention (post). The absolute differences between the

TABLE III. ED Characteristics by Facility Site (January 2013–February 2015)

Facility Region EDIS Accuracy (%) Complexity Level Average Yearly ED Volume # of ED Beds Average Daily Provider Hours
Intervention Site West 90.0 1a 18,086 16 45.7
Control A Northeast 95.1 1c 30,321 17 32
Control B South 95.0 1a 34,241 19 70.7
Control C Midwest 94.1 1a 21,246 18 52.1
Control D South 93.1 1a 25,115 28 63.7
Control E South 93.0 1c 22,059 19 53.1
Control F South 92.9 1a 17,764 39 103.1
Control G Midwest 92.1 1c 22,840 10 41.7
Control H West 92.0 1a 13,605 24 43.4
Control I Northeast 91.7 1c 11,315 15 42.6
Control J South 90.5 1a 16,555 16 61.3
Control K West 90.0 1b 26,845 14 45.7

EDIS, Emergency Department Integration Software.

MILITARY MEDICINE, Vol. 184, January/February 2019 e173

 Applying Lean Principles to Reduce Wait Times

intervention site and the control sites are not important. It is RESULTS
the difference-in-differences, or the differences in the changes Table III summarizes the characteristics of EDs in our study.
over time that are subjected to analysis. This means that our The study included 1 intervention site (VAPAHCS) and 11
statistical methods remove any potentially unobserved con- control sites. Among the control sites, there were 143 mea-
founding differences in the intervention and control sites that surements in the baseline period and 143 measurements in
are fixed over time, apart from any that are simultaneous with the post-intervention period. Overall, “Door to Doctor” time
the intervention. at VAPAHCS decreased 12.6 minutes after the intervention,
Because a simple pre–post comparison of intervention from 34.7 to 22.1, compared with 3.7 minutes from 37.2
group and control groups ignores the underlying assumptions to 33.5 in the control sites (Fig. 2). Regression-adjusted
of longitudinally structured data, a linear mixed-effect model difference-in-difference estimates for “Door to Doctor”
time showed a statistically significant reduction in “Door

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

was utilized to examine this difference in outcome measures
between the pre-intervention and post-intervention periods at to Doctor” time at VAPAHCS compared to control sites
VAPAHCS versus control sites. This model allowed us to (8.9 minutes), during the same time period (standard error
account for within-class correlation between sites and across = 3.5 min; p = 0.01) (Table IV).
time. Our analysis assessed for interaction between post- In the secondary analysis, we examined the “Door to
intervention period and intervention site and examined the Triage” time at VAPAHCS. Overall, “Door to Triage” time at
adjusted p-value for this term at 95% significance level. VAPAHCS decreased 6.3 minutes after the intervention from
All statistical analyses were conducted using R statistical 16.4 to 10.1, compared with 1.3 minutes from 11.0 to 9.7 in
software, version 0.98.50 (R lme4 package was used to fit the the control sites (Fig. 3). Regression-adjusted difference-in-
model and R ggplot2 package was used for data visualization). difference estimates for “Door to Triage” time showed a sta-
tistically significant reduction in “Door to Triage” time at
VAPAHCS compared with control sites (5.0 minutes), during
Ethical Considerations the same time period (p = 0.004) (Table IV). LOS for admit-
This systematic, data-guided project was designed to bring ted and discharged Veterans at VAPAHCS decreased 42.2
about an immediate improvement in a local setting and was minutes and 16.9 minutes, respectively, after the interven-
considered a quality improvement project. Additionally, the tion, compared with 1.8 and 5.2 minutes at the control sites.
Institutional Review Board determined that this project does Regression-adjusted difference-in-difference estimates for
not meet the definition of human subject research. LOS for both admitted and discharged Veterans showed a

FIGURE 2. Change in “Door to Doc” over time in the intervention and control sites before and after the intervention.

e174 MILITARY MEDICINE, Vol. 184, January/February 2019

 Applying Lean Principles to Reduce Wait Times

statistically significant reduction in LOS at VAPAHCS 2.3% for control sites). Regression-adjusted difference-in-
compared with control sites (40.4 and 11.7 minutes, respec- difference estimates for LWBS rates showed LWBS did not
tively), during the same time period (p = 0.002 and 0.04, significantly change at VAPAHCS compared with control
respectively) (Table IV). sites, during the same time period (p = 0.8) (Table IV).
We also compared changes in the LWBS rate between the
intervention and control sites. The average LWBS rate during
the pre-intervention period for VAPAHCS and control sites DISCUSSION
was 0.8% and 2.0%, respectively. In the post-intervention To improve wait times in the ED, VAPAHCS adopted Lean
period, rates were quite similar (0.9% for VAPAHCS and principles in redesigning ED front-end operations. A RPIW
identified wasted steps and changes that would more efficiently

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

TABLE IV. Changes in Waiting Time, ED LOS, and LWBS Rate allocate existing resources. These included standardizing and
from Pre- to Post-periods streamlining work processes, clearly delineating roles for RNs
and MDs, efficiently utilizing beds and improving team com-
Facility Site Pre Post Diff DiD p-Value of the DiD
munication. Our findings suggest that the intervention was suc-
Changes in Door to Doc time (minute) cessful in reducing “Door to Doctor” time relative to similar
VAPAHCS 34.7 22.1 −12.6
controls throughout the VA. Previous studies have shown that
Control sites 37.2 33.5 −3.7 −8.9 0.01
Changes in Door to Triage time (minute) by following Lean, EDs may observe reductions in lengths of
VAPAHCS 16.4 10.1 −6.3 stay and waiting times however, these EDs have been limited
Control sites 11.0 9.7 −1.3 −5 0.004 to pre- post- data reported without use of controls, without sta-
Changes in ED LOS for admitted Veterans (minute) tistical testing to test pre–post differences, and without numeric
VAPAHCS 398.7 356.5 −42.2
data to support reported changes.19,43,31,30 To our knowledge,
Control sites 346.9 345.1 −1.8 −40.4 0.002
Changes in ED LOS for discharged Veterans (minute) this is the first study to detail redesign efforts using Lean and
VAPAHCS 163.2 146.3 −16.9 evaluating changes achieved at the intervention site using con-
Control sites 161.8 156.6 −5.2 −11.7 0.04 trol groups and rigorous methodology.
Changes in LWBS rate (%) Key success factors cited for this program include the fact
VAPAHCS 0.8 0.9 0.1
that, like other effective Lean interventions, flow was priori-
Control sites 2.0 2.3 0.3 −0.2 0.8
tized ahead of efficiency. Further, it was not large or resource
DiD, difference-in-differences. intensive breakthroughs or modifications that led to success

FIGURE 3. Change in “Door to Triage” over time in the intervention site and control sites before and after the intervention.

MILITARY MEDICINE, Vol. 184, January/February 2019 e175

 Applying Lean Principles to Reduce Wait Times

but rather multiple small process enhancements unique to for the ED value stream at the intervention site was to improve
local people, processes, and environment. Leadership buy-in, efficiency from the time of ED provider evaluation to the
creating standard work surrounding the new changes, and patient’s exit from the ED.
providing visual reminders throughout the department helped Some limitations must be considered when interpreting the
ensure sustained results. Continued improvement activities results. First, data of this study may not be generalizable to
like daily huddles and adoption of a Methods, Equipment, other EDs, especially outside VA settings. However, ED
Supplies, and Staffing (MESS) board provided daily stability in crowding and Lean quality improvement are widespread and
the work environment. Such methods prioritized assessing and our findings should inform these efforts. Second, because mul-
providing area readiness with emphasis on transparency of tiple interventions were implemented simultaneously, it is chal-
information, accountability, and follow through on quick hits lenging to determine precisely which specific component was

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

and larger issues. In addition to daily huddles, other efforts such the most impactful. However, it is important to note that there
as on-going weekly VSA meetings, staff meetings, and email were no other significant changes to the ED system during the
follow-ups promoted outstanding communication between time period studied. For example, there were no increases in
nursing, clerical, and physician staff across various shifts. ED or hospital beds, staffing, equipment, or other resources.
Ultimately, the improvement in flow for this value stream Third, though ED staff and all participants were unaware of
motivated staff to participate in future RPIWs. the data collection or analysis, they could not be blinded to the
Important lessons were also learned during informal inter- intervention. While the successes of the RPIW period may be
views. Veterans reported frequently coming to the ED because partially attributed to the Hawthorne effect, we believe its overall
they felt the ED had “great” and “timely” service, but also effect has been attenuated given the lengthy follow up period.
because they had challenges accessing care elsewhere in Fourth, we attempted to minimize differences in data qual-
the healthcare system. Frontline physicians noted that lack ity and capture between sites by selecting sites with good
of examination space was one of the biggest bottlenecks in EDIS adoption. Doing so should bias our results toward the
examining Veterans and resulted in increasing “Door to null. We acknowledge, however, that there may be other
Doctor” time. Communication with other service departments important differences between sites that we were not able
revealed ED staff could utilize outpatient lab, pharmacy and to capture.
radiology services for ambulatory, low-acuity cases, freeing Additionally, we studied a limited number of metrics asso-
nurses to spend more time with sicker Veterans. This feed- ciated with ED flow, and other contributing factors such as
back was crucial to developing the final interventions. inpatient census and acuity were not included. However, we
Many factors are known to influence ED flow and LOS did look at other unintended consequences. Finally, we chose
and have been categorized as input, throughput, and output the best proxy available to study door-to-provider time, based
factors.44,45 This process improvement project focused on on when a physician assigns himself or herself to a Veteran.
throughput factors, arguably the most modifiable factors from While this may not be the exact time a physician sees the
the ED’s vantage point. Because these components are con- Veteran, the same criterion was used consistently and trends in
trolled, in large part, by the ED, ED leaders can design and the data would contain the same amount of systematic error.
implement improvements in these areas, as demonstrated here. Future study should incorporate additional measures of qual-
In fact, our findings suggest that restructuring triage, among ity to determine the effect of Lean on Veteran outcomes and
other process changes, had the most significant impact on the should evaluate the long-term sustainability of the improvement.
“Door to Triage” time, and thus the “Door to Doctor” time. The next challenge is to go beyond the application of Lean
However, many other factors, such as inadequate staffing tools and projects in discrete clinical areas and to develop a
levels, poor communication with laboratory and imaging ser- Lean culture of continuous learning and improvement. Such
vices, and restricted availability or access to inpatient beds for culture change is critical to a healthcare organization’s suc-
admitted patients, are controlled by stakeholders outside the cess in moving from short-term performance improvements
ED. In fact, one of the most common reasons cited for ED to sustained, highly reliable, evidence-based improvements
crowding is the inability to transfer admitted patients from the that ultimately lead to transformation across the organization.
ED to an inpatient bed.46,47 While the decrease in overall
LOS in this study was largely driven by front-end operations,
this intervention was not designed to address output factors REFERENCES
and highlights future opportunity to further improve ED LOS. 1. Institute of Medicine (IOM): Hospital-Based Emergency Care: At the
Though emergency physicians and providers often bear the Breaking Point. Washington, DC, National Academy Press, 2006.
brunt of the ED crowding problem, they commonly cite hav- 2. Kellermann AL.: Crisis in the Emergency Department. N Engl J Med
ing little power to address its causes, since many of those 2006; 355(13): 1300–1303.
3. Horwitz LI, Green J, Bradley EH.: US Emergency Department performance
causes are beyond their control. One solution to this problem
on wait time and length of visit. Ann Emerg Med 2010; 55(2): 133–141.
is the use of Lean methods, which encourage interdisciplinary 4. Horwitz LI, Bradley EH.: Percentage of US Emergency Department
teams and problem-solving across departments. Consequently, patients seen within the recommended triage time: 1997 to 2006. Arch
at the conclusion of this study, the next Lean RPIW selected Intern Med 2009; 169(20): 1857–1865.

e176 MILITARY MEDICINE, Vol. 184, January/February 2019

 Applying Lean Principles to Reduce Wait Times

5. Department of Veterans Affairs Office of Inspector General. Inadequate 23. White BA, Brown DF, Sinclair J, et al: Supplemented Triage and Rapid
Staffing and Poor Patient Flow in the Emergency Department VA Treatment (START) improves performance measures in the Emergency
Maryland Health Care System Baltimore, Maryland. Washington, DC; Department. J Emerg Med 2012; 42(3): 322–8.
2013 Available at https://www.va.gov/oig/; accessed March 1, 2018. 24. Rogg JG, White BA, Biddinger PD, Chang Y, Brown DF.: A long-term
6. Department of Veterans Affairs Office of Inspector General. Emergency analysis of physician triage screening in the Emergency Department.
Department Delays Memphis VA Medical Center Memphis, TN.; 2012. Acad Emerg Med 2013; 20(4): 374–80.
Available at https://www.va.gov/oig/; accessed March 1, 2018. 25. White BA, Chang Y, Grabowski BG, Brown DFM.: Using lean-
7. Department of Veterans Affairs Office of Inspector General. Emergency based systems engineering to increase capacity in the Emergency
Department Concerns Central Alabama VA Health Care System Department. West J Emerg Med 2014; 15(7): 770–6. doi:10.5811/
Montgomery, Alabama. Washington, DC; 2016. Available at https:// westjem.2014.8.21272.
www.va.gov/oig/; accessed March 1, 2018. 26. Powell ES, Khare RK, Venkatesh AK, Van Roo BD, Adams JG,
8. Department of Veterans Affairs, Department of Veterans Affairs Office Reinhardt G.: The relationship between inpatient discharge timing

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

of Inspector General. Emergency Department Staffing and Patient and Emergency Department boarding. J Emerg Med 2012; 42(2):
Safety Issues VA San Diego Healthcare System San Diego, California. 186–96. 16.
Washington, DC; 2014. Available at https://www.va.gov/oig/; accessed 27. Ng D, Vail G, Thomas S, Schmidt N: Applying the Lean principles of
March 1, 2018. the Toyota Production System to reduce wait times in the Emergency
9. Department of Veterans Affairs, Department of Veterans Affairs Office Department. CJEM 2010; 12(1): 50–7.
of Inspector General. Access and Quality of Care Concerns, PVAHCS, 28. A Structured Approach to Transforming a Large Public Hospital
Phoenix, AZ and Minneapolis VAHCS, Minneapolis, MN. Washington, Emergency Department via Lean Methodologies - Naik - 2011 - Journal
DC; 2016. Available at https://www.va.gov/oig/; accessed March 1, for Healthcare Quality - Wiley Online Library. Available at http://
2018. onlinelibrary.wiley.com/doi/10.1111/j.1945-1474.2011.00181.x/full;
10. Bernstein SL, Aronsky D, Duseja R, et al: The effect of Emergency accessed October 13, 2017.
Department crowding on clinically oriented outcomes. Acad Emerg 29. Vermeulen MJ, Stukel TA, Guttmann A, et al: Evaluation of an
Med 2009; 16(1): 1–10. Emergency Department lean process improvement program to reduce
11. Stang AS, Crotts J, Johnson DW, Hartling L, Guttmann A.: Crowding length of stay. Ann Emerg Med 2014; 64(5): 427–38.
measures associated with the quality of Emergency Department care: a 30. Dickson EW, Singh S, Cheung DS, Wyatt CC, Nugent AS.:
systematic review. Acad Emerg Med Off J Soc Acad Emerg Med 2015; Application of lean manufacturing techniques in the Emergency
22(6): 643–56. doi:10.1111/acem.12682. Department. J Emerg Med 2009; 37(2): 177–82.
12. Wu D, Zhou X, Ye L, Gan J, Zhang M.: Emergency department crowd- 31. Dickson EW, Anguelov Z, Vetterick D, Eller A, Singh S.: Use of Lean
ing and the performance of damage control resuscitation in major in the Emergency Department: a case series of 4 hospitals. Ann Emerg
trauma patients with hemorrhagic shock. Acad Emerg Med 2015; 22(8): Med 2009; 54(4): 504–10.
915–21. 32. Chan H, Lo S, Lee L, et al: Lean techniques for the improvement of
13. Schull MJ, Vermeulen M, Slaughter G, Morrison L, Daly P.: patients’ flow in Emergency Department. World J Emerg Med 2014; 5
Emergency department crowding and thrombolysis delays in acute myo- (1): 24–8.
cardial infarction. Ann Emerg Med 2004; 44(6): 577–85. 33. Decker WW, Stead LG.: Application of lean thinking in health care: a
14. Spaite DW, Bartholomeaux F, Guisto J, et al: Rapid process redesign in role in Emergency Departments globally. Int J Emerg Med 2008; 1(3):
a university-based Emergency Department: decreasing waiting time 161–2.
intervals and improving patient satisfaction. Ann Emerg Med 2002; 39 34. King DL, Ben-Tovim DI, Bassham J.: Redesigning Emergency
(2): 168–77. Department patient flows: application of Lean Thinking to health care.
15. Pines JM, Iyer S, Disbot M, Hollander JE, Shofer FS, Datner EM.: The Emerg Med Australas 2006; 18(4): 391–7.
effect of Emergency Department crowding on patient satisfaction for 35. Kulkarni RG.: Going lean in the emergency department: a strategy for
admitted patients. Acad Emerg Med 2008; 15(9): 825–31. addressing emergency department overcrowding. Medscape Gen Med
16. Rodi SW, Grau MV, Orsini CM.: Evaluation of a fast track unit: align- 2007; 9(4): 58.
ment of resources and demand results in improved satisfaction and 36. Murrell KL, Offerman SR, Kauffman MB.: Applying lean: implementa-
decreased length of stay for Emergency Department patients. Qual tion of a rapid triage and treatment system. West J Emerg Med 2011;
Manage Health Care 2006; 15(3): 163–70. 12(2): 184.
17. National Academy of Engineering (NAE) and Institute of Medicine 37. Hitti EA, El-Eid GR, Tamim H, Saleh R, Saliba M, Naffaa L.:
(IOM): Building a Better Delivery System: A New Engineering/Health Improving emergency department radiology transportation time: a suc-
Care Partnership. Washington, DC, National Academies Press (US), 2005. cessful implementation of lean methodology. BMC Health Serv Res
18. Mazzocato P, Savage C, Brommels M, Aronsson H, Thor J: Lean think- 2017; 17(1): 625.
ing in healthcare: a realist review of the literature. Qual Saf Health Care 38. Jacobson GH, McCoin NS, Lescallette R, Russ S, Slovis CM.: Kaizen:
2010; 19(5): 376–82. a method of process improvement in the emergency department. Acad
19. Holden RJ.: Lean thinking in emergency departments: a critical review. Emerg Med 2009; 16(12): 1341–9.
Ann Emerg Med 2011; 57(3): 265–78. 39. El Sayed MJ, El-Eid GR, Saliba M, Jabbour R, Hitti EA.: Improving
20. Hospitals See Benefits of Lean and Six Sigma. ASQ. 9 March 2009. emergency department door to doctor time and process reliability.
http://asq.org/qualitynews/qnt/execute/displaySetup?newsID=5843. Medicine (Baltimore) 2015; 94: 42.
21. Holroyd BR, Bullard MJ, Latoszek K, et al: Impact of a triage liai- 40. Bucci S, De Belvis AG, Marventano S, et al: Emergency department
son physician on Emergency Department overcrowding and through- crowding and hospital bed shortage: is Lean a smart answer? A sys-
put: a randomized controlled trial. Acad Emerg Med 2007; 14(8): tematic review. Eur Rev Med Pharmacol Sci 2016; 20(20):
702–8. 4209–19.
22. Han JH, France DJ, Levin SR, Jones ID, Storrow AB, Aronsky D.: The 41. Gayed B, Black S, Daggy J, Munshi IA.: Redesigning a joint replace-
effect of physician triage on Emergency Department length of stay. ment program using Lean Six Sigma in a Veterans Affairs hospital.
J Emerg Med 2010; 39(2): 227–33. JAMA Surg 2013; 148(11): 1050–6.

MILITARY MEDICINE, Vol. 184, January/February 2019 e177

 Applying Lean Principles to Reduce Wait Times

42. Pocha C.: Lean Six Sigma in health care and the challenge of imple- 45. Asplin BR, Magid DJ, Rhodes KV, Solberg LI, Lurie N, Camargo CA.:
mentation of Six Sigma methodologies at a Veterans Affairs Medical A conceptual model of emergency department crowding. Ann Emerg
Center. Qual Manag Healthc 2010; 19(4): 312–8. Med 2003; 42(2): 173–80.
43. Holden RJ.: Lean thinking in emergency departments: a critical review. 46. Forster AJ, Stiell I, Wells G, Lee AJ, Van Walraven C.: The effect of
Ann Emerg Med 2011; 57(3): 265–78. hospital occupancy on emergency department length of stay and patient
44. Solberg LI, Asplin BR, Weinick RM, Magid DJ.: Emergency depart- disposition. Acad Emerg Med 2003; 10(2): 127–33.
ment crowding: consensus development of potential measures. Ann 47. Chan L, Reilly KM, Salluzzo RF.: Variables that affect patient throughput times
Emerg Med 2003; 42(6): 824–34. in an academic emergency department. Am J Med Qual 1997; 12(4): 183–6.

Downloaded from https://academic.oup.com/milmed/article-abstract/184/1-2/e169/5051964 by guest on 17 June 2020

e178 MILITARY MEDICINE, Vol. 184, January/February 2019